dirstate: align the dirstate's API to the lower level ones

This conclude the refactoring of this API. We can now finalize the dirstate v2
on disk format.

Differential Revision: https://phab.mercurial-scm.org/D11587

hgext/largefiles/lfutil.py

@@ -1,798 +1,788 b''
 # Copyright 2009-2010 Gregory P. Ward
 # Copyright 2009-2010 Intelerad Medical Systems Incorporated
 # Copyright 2010-2011 Fog Creek Software
 # Copyright 2010-2011 Unity Technologies
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 '''largefiles utility code: must not import other modules in this package.'''
 from __future__ import absolute_import
 
 import contextlib
 import copy
 import os
 import stat
 
 from mercurial.i18n import _
 from mercurial.node import hex
 from mercurial.pycompat import open
 
 from mercurial import (
     dirstate,
     encoding,
     error,
     httpconnection,
     match as matchmod,
     pycompat,
     requirements,
     scmutil,
     sparse,
     util,
     vfs as vfsmod,
 )
 from mercurial.utils import hashutil
 
 shortname = b'.hglf'
 shortnameslash = shortname + b'/'
 longname = b'largefiles'
 
 # -- Private worker functions ------------------------------------------
 
 
 @contextlib.contextmanager
 def lfstatus(repo, value=True):
     oldvalue = getattr(repo, 'lfstatus', False)
     repo.lfstatus = value
     try:
         yield
     finally:
         repo.lfstatus = oldvalue
 
 
 def getminsize(ui, assumelfiles, opt, default=10):
     lfsize = opt
     if not lfsize and assumelfiles:
         lfsize = ui.config(longname, b'minsize', default=default)
     if lfsize:
         try:
             lfsize = float(lfsize)
         except ValueError:
             raise error.Abort(
                 _(b'largefiles: size must be number (not %s)\n') % lfsize
             )
     if lfsize is None:
         raise error.Abort(_(b'minimum size for largefiles must be specified'))
     return lfsize
 
 
 def link(src, dest):
     """Try to create hardlink - if that fails, efficiently make a copy."""
     util.makedirs(os.path.dirname(dest))
     try:
         util.oslink(src, dest)
     except OSError:
         # if hardlinks fail, fallback on atomic copy
         with open(src, b'rb') as srcf, util.atomictempfile(dest) as dstf:
             for chunk in util.filechunkiter(srcf):
                 dstf.write(chunk)
         os.chmod(dest, os.stat(src).st_mode)
 
 
 def usercachepath(ui, hash):
     """Return the correct location in the "global" largefiles cache for a file
     with the given hash.
     This cache is used for sharing of largefiles across repositories - both
     to preserve download bandwidth and storage space."""
     return os.path.join(_usercachedir(ui), hash)
 
 
 def _usercachedir(ui, name=longname):
     '''Return the location of the "global" largefiles cache.'''
     path = ui.configpath(name, b'usercache')
     if path:
         return path
 
     hint = None
 
     if pycompat.iswindows:
         appdata = encoding.environ.get(
             b'LOCALAPPDATA', encoding.environ.get(b'APPDATA')
         )
         if appdata:
             return os.path.join(appdata, name)
 
         hint = _(b"define %s or %s in the environment, or set %s.usercache") % (
             b"LOCALAPPDATA",
             b"APPDATA",
             name,
         )
     elif pycompat.isdarwin:
         home = encoding.environ.get(b'HOME')
         if home:
             return os.path.join(home, b'Library', b'Caches', name)
 
         hint = _(b"define %s in the environment, or set %s.usercache") % (
             b"HOME",
             name,
         )
     elif pycompat.isposix:
         path = encoding.environ.get(b'XDG_CACHE_HOME')
         if path:
             return os.path.join(path, name)
         home = encoding.environ.get(b'HOME')
         if home:
             return os.path.join(home, b'.cache', name)
 
         hint = _(b"define %s or %s in the environment, or set %s.usercache") % (
             b"XDG_CACHE_HOME",
             b"HOME",
             name,
         )
     else:
         raise error.Abort(
             _(b'unknown operating system: %s\n') % pycompat.osname
         )
 
     raise error.Abort(_(b'unknown %s usercache location') % name, hint=hint)
 
 
 def inusercache(ui, hash):
     path = usercachepath(ui, hash)
     return os.path.exists(path)
 
 
 def findfile(repo, hash):
     """Return store path of the largefile with the specified hash.
     As a side effect, the file might be linked from user cache.
     Return None if the file can't be found locally."""
     path, exists = findstorepath(repo, hash)
     if exists:
         repo.ui.note(_(b'found %s in store\n') % hash)
         return path
     elif inusercache(repo.ui, hash):
         repo.ui.note(_(b'found %s in system cache\n') % hash)
         path = storepath(repo, hash)
         link(usercachepath(repo.ui, hash), path)
         return path
     return None
 
 
 class largefilesdirstate(dirstate.dirstate):
     def __getitem__(self, key):
         return super(largefilesdirstate, self).__getitem__(unixpath(key))
 
     def set_tracked(self, f):
         return super(largefilesdirstate, self).set_tracked(unixpath(f))
 
     def set_untracked(self, f):
         return super(largefilesdirstate, self).set_untracked(unixpath(f))
 
     def normal(self, f, parentfiledata=None):
         # not sure if we should pass the `parentfiledata` down or throw it
         # away. So throwing it away to stay on the safe side.
         return super(largefilesdirstate, self).normal(unixpath(f))
 
     def remove(self, f):
         return super(largefilesdirstate, self).remove(unixpath(f))
 
     def add(self, f):
         return super(largefilesdirstate, self).add(unixpath(f))
 
     def drop(self, f):
         return super(largefilesdirstate, self).drop(unixpath(f))
 
     def forget(self, f):
         return super(largefilesdirstate, self).forget(unixpath(f))
 
     def normallookup(self, f):
         return super(largefilesdirstate, self).normallookup(unixpath(f))
 
     def _ignore(self, f):
         return False
 
     def write(self, tr=False):
         # (1) disable PENDING mode always
         #     (lfdirstate isn't yet managed as a part of the transaction)
         # (2) avoid develwarn 'use dirstate.write with ....'
         super(largefilesdirstate, self).write(None)
 
 
 def openlfdirstate(ui, repo, create=True):
     """
     Return a dirstate object that tracks largefiles: i.e. its root is
     the repo root, but it is saved in .hg/largefiles/dirstate.
     """
     vfs = repo.vfs
     lfstoredir = longname
     opener = vfsmod.vfs(vfs.join(lfstoredir))
     use_dirstate_v2 = requirements.DIRSTATE_V2_REQUIREMENT in repo.requirements
     lfdirstate = largefilesdirstate(
         opener,
         ui,
         repo.root,
         repo.dirstate._validate,
         lambda: sparse.matcher(repo),
         repo.nodeconstants,
         use_dirstate_v2,
     )
 
     # If the largefiles dirstate does not exist, populate and create
     # it. This ensures that we create it on the first meaningful
     # largefiles operation in a new clone.
     if create and not vfs.exists(vfs.join(lfstoredir, b'dirstate')):
         matcher = getstandinmatcher(repo)
         standins = repo.dirstate.walk(
             matcher, subrepos=[], unknown=False, ignored=False
         )
 
         if len(standins) > 0:
             vfs.makedirs(lfstoredir)
 
         with lfdirstate.parentchange():
             for standin in standins:
                 lfile = splitstandin(standin)
                 lfdirstate.update_file(
                     lfile, p1_tracked=True, wc_tracked=True, possibly_dirty=True
                 )
     return lfdirstate
 
 
 def lfdirstatestatus(lfdirstate, repo):
     pctx = repo[b'.']
     match = matchmod.always()
     unsure, s = lfdirstate.status(
         match, subrepos=[], ignored=False, clean=False, unknown=False
     )
     modified, clean = s.modified, s.clean
     for lfile in unsure:
         try:
             fctx = pctx[standin(lfile)]
         except LookupError:
             fctx = None
         if not fctx or readasstandin(fctx) != hashfile(repo.wjoin(lfile)):
             modified.append(lfile)
         else:
             clean.append(lfile)
             lfdirstate.set_clean(lfile)
     return s
 
 
 def listlfiles(repo, rev=None, matcher=None):
     """return a list of largefiles in the working copy or the
     specified changeset"""
 
     if matcher is None:
         matcher = getstandinmatcher(repo)
 
     # ignore unknown files in working directory
     return [
         splitstandin(f)
         for f in repo[rev].walk(matcher)
         if rev is not None or repo.dirstate.get_entry(f).any_tracked
     ]
 
 
 def instore(repo, hash, forcelocal=False):
     '''Return true if a largefile with the given hash exists in the store'''
     return os.path.exists(storepath(repo, hash, forcelocal))
 
 
 def storepath(repo, hash, forcelocal=False):
     """Return the correct location in the repository largefiles store for a
     file with the given hash."""
     if not forcelocal and repo.shared():
         return repo.vfs.reljoin(repo.sharedpath, longname, hash)
     return repo.vfs.join(longname, hash)
 
 
 def findstorepath(repo, hash):
     """Search through the local store path(s) to find the file for the given
     hash.  If the file is not found, its path in the primary store is returned.
     The return value is a tuple of (path, exists(path)).
     """
     # For shared repos, the primary store is in the share source.  But for
     # backward compatibility, force a lookup in the local store if it wasn't
     # found in the share source.
     path = storepath(repo, hash, False)
 
     if instore(repo, hash):
         return (path, True)
     elif repo.shared() and instore(repo, hash, True):
         return storepath(repo, hash, True), True
 
     return (path, False)
 
 
 def copyfromcache(repo, hash, filename):
     """Copy the specified largefile from the repo or system cache to
     filename in the repository. Return true on success or false if the
     file was not found in either cache (which should not happened:
     this is meant to be called only after ensuring that the needed
     largefile exists in the cache)."""
     wvfs = repo.wvfs
     path = findfile(repo, hash)
     if path is None:
         return False
     wvfs.makedirs(wvfs.dirname(wvfs.join(filename)))
     # The write may fail before the file is fully written, but we
     # don't use atomic writes in the working copy.
     with open(path, b'rb') as srcfd, wvfs(filename, b'wb') as destfd:
         gothash = copyandhash(util.filechunkiter(srcfd), destfd)
     if gothash != hash:
         repo.ui.warn(
             _(b'%s: data corruption in %s with hash %s\n')
             % (filename, path, gothash)
         )
         wvfs.unlink(filename)
         return False
     return True
 
 
 def copytostore(repo, ctx, file, fstandin):
     wvfs = repo.wvfs
     hash = readasstandin(ctx[fstandin])
     if instore(repo, hash):
         return
     if wvfs.exists(file):
         copytostoreabsolute(repo, wvfs.join(file), hash)
     else:
         repo.ui.warn(
             _(b"%s: largefile %s not available from local store\n")
             % (file, hash)
         )
 
 
 def copyalltostore(repo, node):
     '''Copy all largefiles in a given revision to the store'''
 
     ctx = repo[node]
     for filename in ctx.files():
         realfile = splitstandin(filename)
         if realfile is not None and filename in ctx.manifest():
             copytostore(repo, ctx, realfile, filename)
 
 
 def copytostoreabsolute(repo, file, hash):
     if inusercache(repo.ui, hash):
         link(usercachepath(repo.ui, hash), storepath(repo, hash))
     else:
         util.makedirs(os.path.dirname(storepath(repo, hash)))
         with open(file, b'rb') as srcf:
             with util.atomictempfile(
                 storepath(repo, hash), createmode=repo.store.createmode
             ) as dstf:
                 for chunk in util.filechunkiter(srcf):
                     dstf.write(chunk)
         linktousercache(repo, hash)
 
 
 def linktousercache(repo, hash):
     """Link / copy the largefile with the specified hash from the store
     to the cache."""
     path = usercachepath(repo.ui, hash)
     link(storepath(repo, hash), path)
 
 
 def getstandinmatcher(repo, rmatcher=None):
     '''Return a match object that applies rmatcher to the standin directory'''
     wvfs = repo.wvfs
     standindir = shortname
 
     # no warnings about missing files or directories
     badfn = lambda f, msg: None
 
     if rmatcher and not rmatcher.always():
         pats = [wvfs.join(standindir, pat) for pat in rmatcher.files()]
         if not pats:
             pats = [wvfs.join(standindir)]
         match = scmutil.match(repo[None], pats, badfn=badfn)
     else:
         # no patterns: relative to repo root
         match = scmutil.match(repo[None], [wvfs.join(standindir)], badfn=badfn)
     return match
 
 
 def composestandinmatcher(repo, rmatcher):
     """Return a matcher that accepts standins corresponding to the
     files accepted by rmatcher. Pass the list of files in the matcher
     as the paths specified by the user."""
     smatcher = getstandinmatcher(repo, rmatcher)
     isstandin = smatcher.matchfn
 
     def composedmatchfn(f):
         return isstandin(f) and rmatcher.matchfn(splitstandin(f))
 
     smatcher.matchfn = composedmatchfn
 
     return smatcher
 
 
 def standin(filename):
     """Return the repo-relative path to the standin for the specified big
     file."""
     # Notes:
     # 1) Some callers want an absolute path, but for instance addlargefiles
     #    needs it repo-relative so it can be passed to repo[None].add().  So
     #    leave it up to the caller to use repo.wjoin() to get an absolute path.
     # 2) Join with '/' because that's what dirstate always uses, even on
     #    Windows. Change existing separator to '/' first in case we are
     #    passed filenames from an external source (like the command line).
     return shortnameslash + util.pconvert(filename)
 
 
 def isstandin(filename):
     """Return true if filename is a big file standin. filename must be
     in Mercurial's internal form (slash-separated)."""
     return filename.startswith(shortnameslash)
 
 
 def splitstandin(filename):
     # Split on / because that's what dirstate always uses, even on Windows.
     # Change local separator to / first just in case we are passed filenames
     # from an external source (like the command line).
     bits = util.pconvert(filename).split(b'/', 1)
     if len(bits) == 2 and bits[0] == shortname:
         return bits[1]
     else:
         return None
 
 
 def updatestandin(repo, lfile, standin):
     """Re-calculate hash value of lfile and write it into standin
 
     This assumes that "lfutil.standin(lfile) == standin", for efficiency.
     """
     file = repo.wjoin(lfile)
     if repo.wvfs.exists(lfile):
         hash = hashfile(file)
         executable = getexecutable(file)
         writestandin(repo, standin, hash, executable)
     else:
         raise error.Abort(_(b'%s: file not found!') % lfile)
 
 
 def readasstandin(fctx):
     """read hex hash from given filectx of standin file
 
     This encapsulates how "standin" data is stored into storage layer."""
     return fctx.data().strip()
 
 
 def writestandin(repo, standin, hash, executable):
     '''write hash to <repo.root>/<standin>'''
     repo.wwrite(standin, hash + b'\n', executable and b'x' or b'')
 
 
 def copyandhash(instream, outfile):
     """Read bytes from instream (iterable) and write them to outfile,
     computing the SHA-1 hash of the data along the way. Return the hash."""
     hasher = hashutil.sha1(b'')
     for data in instream:
         hasher.update(data)
         outfile.write(data)
     return hex(hasher.digest())
 
 
 def hashfile(file):
     if not os.path.exists(file):
         return b''
     with open(file, b'rb') as fd:
         return hexsha1(fd)
 
 
 def getexecutable(filename):
     mode = os.stat(filename).st_mode
     return (
         (mode & stat.S_IXUSR)
         and (mode & stat.S_IXGRP)
         and (mode & stat.S_IXOTH)
     )
 
 
 def urljoin(first, second, *arg):
     def join(left, right):
         if not left.endswith(b'/'):
             left += b'/'
         if right.startswith(b'/'):
             right = right[1:]
         return left + right
 
     url = join(first, second)
     for a in arg:
         url = join(url, a)
     return url
 
 
 def hexsha1(fileobj):
     """hexsha1 returns the hex-encoded sha1 sum of the data in the file-like
     object data"""
     h = hashutil.sha1()
     for chunk in util.filechunkiter(fileobj):
         h.update(chunk)
     return hex(h.digest())
 
 
 def httpsendfile(ui, filename):
     return httpconnection.httpsendfile(ui, filename, b'rb')
 
 
 def unixpath(path):
     '''Return a version of path normalized for use with the lfdirstate.'''
     return util.pconvert(os.path.normpath(path))
 
 
 def islfilesrepo(repo):
     '''Return true if the repo is a largefile repo.'''
     if b'largefiles' in repo.requirements and any(
         shortnameslash in f[1] for f in repo.store.datafiles()
     ):
         return True
 
     return any(openlfdirstate(repo.ui, repo, False))
 
 
 class storeprotonotcapable(Exception):
     def __init__(self, storetypes):
         self.storetypes = storetypes
 
 
 def getstandinsstate(repo):
     standins = []
     matcher = getstandinmatcher(repo)
     wctx = repo[None]
     for standin in repo.dirstate.walk(
         matcher, subrepos=[], unknown=False, ignored=False
     ):
         lfile = splitstandin(standin)
         try:
             hash = readasstandin(wctx[standin])
         except IOError:
             hash = None
         standins.append((lfile, hash))
     return standins
 
 
 def synclfdirstate(repo, lfdirstate, lfile, normallookup):
     lfstandin = standin(lfile)
     if lfstandin not in repo.dirstate:
         lfdirstate.update_file(lfile, p1_tracked=False, wc_tracked=False)
     else:
-        stat = repo.dirstate.get_entry(lfstandin)
-        state, mtime = stat.state, stat.mtime
-        if state == b'n':
-            if normallookup or mtime < 0 or not repo.wvfs.exists(lfile):
-                # state 'n' doesn't ensure 'clean' in this case
-                lfdirstate.update_file(
-                    lfile, p1_tracked=True, wc_tracked=True, possibly_dirty=True
-                )
-            else:
-                lfdirstate.update_file(lfile, p1_tracked=True, wc_tracked=True)
-        elif state == b'm':
-            lfdirstate.update_file(
-                lfile, p1_tracked=True, wc_tracked=True, merged=True
-            )
-        elif state == b'r':
-            lfdirstate.update_file(lfile, p1_tracked=True, wc_tracked=False)
-        elif state == b'a':
-            lfdirstate.update_file(lfile, p1_tracked=False, wc_tracked=True)
+        entry = repo.dirstate.get_entry(lfstandin)
+        lfdirstate.update_file(
+            lfile,
+            wc_tracked=entry.tracked,
+            p1_tracked=entry.p1_tracked,
+            p2_info=entry.p2_info,
+            possibly_dirty=True,
+        )
 
 
 def markcommitted(orig, ctx, node):
     repo = ctx.repo()
 
     lfdirstate = openlfdirstate(repo.ui, repo)
     with lfdirstate.parentchange():
         orig(node)
 
         # ATTENTION: "ctx.files()" may differ from "repo[node].files()"
         # because files coming from the 2nd parent are omitted in the latter.
         #
         # The former should be used to get targets of "synclfdirstate",
         # because such files:
         # - are marked as "a" by "patch.patch()" (e.g. via transplant), and
         # - have to be marked as "n" after commit, but
         # - aren't listed in "repo[node].files()"
 
         for f in ctx.files():
             lfile = splitstandin(f)
             if lfile is not None:
                 synclfdirstate(repo, lfdirstate, lfile, False)
     lfdirstate.write()
 
     # As part of committing, copy all of the largefiles into the cache.
     #
     # Using "node" instead of "ctx" implies additional "repo[node]"
     # lookup while copyalltostore(), but can omit redundant check for
     # files comming from the 2nd parent, which should exist in store
     # at merging.
     copyalltostore(repo, node)
 
 
 def getlfilestoupdate(oldstandins, newstandins):
     changedstandins = set(oldstandins).symmetric_difference(set(newstandins))
     filelist = []
     for f in changedstandins:
         if f[0] not in filelist:
             filelist.append(f[0])
     return filelist
 
 
 def getlfilestoupload(repo, missing, addfunc):
     makeprogress = repo.ui.makeprogress
     with makeprogress(
         _(b'finding outgoing largefiles'),
         unit=_(b'revisions'),
         total=len(missing),
     ) as progress:
         for i, n in enumerate(missing):
             progress.update(i)
             parents = [p for p in repo[n].parents() if p != repo.nullid]
 
             with lfstatus(repo, value=False):
                 ctx = repo[n]
 
             files = set(ctx.files())
             if len(parents) == 2:
                 mc = ctx.manifest()
                 mp1 = ctx.p1().manifest()
                 mp2 = ctx.p2().manifest()
                 for f in mp1:
                     if f not in mc:
                         files.add(f)
                 for f in mp2:
                     if f not in mc:
                         files.add(f)
                 for f in mc:
                     if mc[f] != mp1.get(f, None) or mc[f] != mp2.get(f, None):
                         files.add(f)
             for fn in files:
                 if isstandin(fn) and fn in ctx:
                     addfunc(fn, readasstandin(ctx[fn]))
 
 
 def updatestandinsbymatch(repo, match):
     """Update standins in the working directory according to specified match
 
     This returns (possibly modified) ``match`` object to be used for
     subsequent commit process.
     """
 
     ui = repo.ui
 
     # Case 1: user calls commit with no specific files or
     # include/exclude patterns: refresh and commit all files that
     # are "dirty".
     if match is None or match.always():
         # Spend a bit of time here to get a list of files we know
         # are modified so we can compare only against those.
         # It can cost a lot of time (several seconds)
         # otherwise to update all standins if the largefiles are
         # large.
         lfdirstate = openlfdirstate(ui, repo)
         dirtymatch = matchmod.always()
         unsure, s = lfdirstate.status(
             dirtymatch, subrepos=[], ignored=False, clean=False, unknown=False
         )
         modifiedfiles = unsure + s.modified + s.added + s.removed
         lfiles = listlfiles(repo)
         # this only loops through largefiles that exist (not
         # removed/renamed)
         for lfile in lfiles:
             if lfile in modifiedfiles:
                 fstandin = standin(lfile)
                 if repo.wvfs.exists(fstandin):
                     # this handles the case where a rebase is being
                     # performed and the working copy is not updated
                     # yet.
                     if repo.wvfs.exists(lfile):
                         updatestandin(repo, lfile, fstandin)
 
         return match
 
     lfiles = listlfiles(repo)
     match._files = repo._subdirlfs(match.files(), lfiles)
 
     # Case 2: user calls commit with specified patterns: refresh
     # any matching big files.
     smatcher = composestandinmatcher(repo, match)
     standins = repo.dirstate.walk(
         smatcher, subrepos=[], unknown=False, ignored=False
     )
 
     # No matching big files: get out of the way and pass control to
     # the usual commit() method.
     if not standins:
         return match
 
     # Refresh all matching big files.  It's possible that the
     # commit will end up failing, in which case the big files will
     # stay refreshed.  No harm done: the user modified them and
     # asked to commit them, so sooner or later we're going to
     # refresh the standins.  Might as well leave them refreshed.
     lfdirstate = openlfdirstate(ui, repo)
     for fstandin in standins:
         lfile = splitstandin(fstandin)
         if lfdirstate.get_entry(lfile).tracked:
             updatestandin(repo, lfile, fstandin)
 
     # Cook up a new matcher that only matches regular files or
     # standins corresponding to the big files requested by the
     # user.  Have to modify _files to prevent commit() from
     # complaining "not tracked" for big files.
     match = copy.copy(match)
     origmatchfn = match.matchfn
 
     # Check both the list of largefiles and the list of
     # standins because if a largefile was removed, it
     # won't be in the list of largefiles at this point
     match._files += sorted(standins)
 
     actualfiles = []
     for f in match._files:
         fstandin = standin(f)
 
         # For largefiles, only one of the normal and standin should be
         # committed (except if one of them is a remove).  In the case of a
         # standin removal, drop the normal file if it is unknown to dirstate.
         # Thus, skip plain largefile names but keep the standin.
         if f in lfiles or fstandin in standins:
             if not repo.dirstate.get_entry(fstandin).removed:
                 if not repo.dirstate.get_entry(f).removed:
                     continue
             elif not repo.dirstate.get_entry(f).any_tracked:
                 continue
 
         actualfiles.append(f)
     match._files = actualfiles
 
     def matchfn(f):
         if origmatchfn(f):
             return f not in lfiles
         else:
             return f in standins
 
     match.matchfn = matchfn
 
     return match
 
 
 class automatedcommithook(object):
     """Stateful hook to update standins at the 1st commit of resuming
 
     For efficiency, updating standins in the working directory should
     be avoided while automated committing (like rebase, transplant and
     so on), because they should be updated before committing.
 
     But the 1st commit of resuming automated committing (e.g. ``rebase
     --continue``) should update them, because largefiles may be
     modified manually.
     """
 
     def __init__(self, resuming):
         self.resuming = resuming
 
     def __call__(self, repo, match):
         if self.resuming:
             self.resuming = False  # avoids updating at subsequent commits
             return updatestandinsbymatch(repo, match)
         else:
             return match
 
 
 def getstatuswriter(ui, repo, forcibly=None):
     """Return the function to write largefiles specific status out
 
     If ``forcibly`` is ``None``, this returns the last element of
     ``repo._lfstatuswriters`` as "default" writer function.
 
     Otherwise, this returns the function to always write out (or
     ignore if ``not forcibly``) status.
     """
     if forcibly is None and util.safehasattr(repo, b'_largefilesenabled'):
         return repo._lfstatuswriters[-1]
     else:
         if forcibly:
             return ui.status  # forcibly WRITE OUT
         else:
             return lambda *msg, **opts: None  # forcibly IGNORE

mercurial/dirstate.py

@@ -1,1534 +1,1526 b''
 # dirstate.py - working directory tracking for mercurial
 #
 # Copyright 2005-2007 Olivia 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 .pycompat import delattr
 
 from hgdemandimport import tracing
 
 from . import (
     dirstatemap,
     encoding,
     error,
     match as matchmod,
     pathutil,
     policy,
     pycompat,
     scmutil,
     sparse,
     util,
 )
 
 from .interfaces import (
     dirstate as intdirstate,
     util as interfaceutil,
 )
 
 parsers = policy.importmod('parsers')
 rustmod = policy.importrust('dirstate')
 
 SUPPORTS_DIRSTATE_V2 = rustmod is not None
 
 propertycache = util.propertycache
 filecache = scmutil.filecache
 _rangemask = dirstatemap.rangemask
 
 DirstateItem = dirstatemap.DirstateItem
 
 
 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)
 
 
 def requires_parents_change(func):
     def wrap(self, *args, **kwargs):
         if not self.pendingparentchange():
             msg = 'calling `%s` outside of a parentchange context'
             msg %= func.__name__
             raise error.ProgrammingError(msg)
         return func(self, *args, **kwargs)
 
     return wrap
 
 
 def requires_no_parents_change(func):
     def wrap(self, *args, **kwargs):
         if self.pendingparentchange():
             msg = 'calling `%s` inside of a parentchange context'
             msg %= func.__name__
             raise error.ProgrammingError(msg)
         return func(self, *args, **kwargs)
 
     return wrap
 
 
 @interfaceutil.implementer(intdirstate.idirstate)
 class dirstate(object):
     def __init__(
         self,
         opener,
         ui,
         root,
         validate,
         sparsematchfn,
         nodeconstants,
         use_dirstate_v2,
     ):
         """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._use_dirstate_v2 = use_dirstate_v2
         self._nodeconstants = nodeconstants
         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._mapcls = dirstatemap.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
 
     def prefetch_parents(self):
         """make sure the parents are loaded
 
         Used to avoid a race condition.
         """
         self._pl
 
     @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,
             self._nodeconstants,
             self._use_dirstate_v2,
         )
         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 bool(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
 
         XXX The "state" is a bit obscure to be in the "public" API. we should
         consider migrating all user of this to going through the dirstate entry
         instead.
         """
         msg = b"don't use dirstate[file], use dirstate.get_entry(file)"
         util.nouideprecwarn(msg, b'6.1', stacklevel=2)
         entry = self._map.get(key)
         if entry is not None:
             return entry.state
         return b'?'
 
     def get_entry(self, path):
         """return a DirstateItem for the associated path"""
         entry = self._map.get(path)
         if entry is None:
             return DirstateItem()
         return entry
 
     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])
 
     @property
     def in_merge(self):
         """True if a merge is in progress"""
         return self._pl[1] != self._nodeconstants.nullid
 
     def branch(self):
         return encoding.tolocal(self._branch)
 
     def setparents(self, p1, p2=None):
         """Set dirstate parents to p1 and p2.
 
         When moving from two parents to one, "merged" entries a
         adjusted to normal and previous copy records discarded and
         returned by the call.
 
         See localrepo.setparents()
         """
         if p2 is None:
             p2 = self._nodeconstants.nullid
         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
         nullid = self._nodeconstants.nullid
         # True if we need to fold p2 related state back to a linear case
         fold_p2 = oldp2 != nullid and p2 == nullid
         return self._map.setparents(p1, p2, fold_p2=fold_p2)
 
     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._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
         else:
             self._map.copymap.pop(dest, None)
 
     def copied(self, file):
         return self._map.copymap.get(file, None)
 
     def copies(self):
         return self._map.copymap
 
     @requires_no_parents_change
     def set_tracked(self, filename):
         """a "public" method for generic code to mark a file as tracked
 
         This function is to be called outside of "update/merge" case. For
         example by a command like `hg add X`.
 
         return True the file was previously untracked, False otherwise.
         """
         self._dirty = True
         entry = self._map.get(filename)
         if entry is None or not entry.tracked:
             self._check_new_tracked_filename(filename)
         return self._map.set_tracked(filename)
 
     @requires_no_parents_change
     def set_untracked(self, filename):
         """a "public" method for generic code to mark a file as untracked
 
         This function is to be called outside of "update/merge" case. For
         example by a command like `hg remove X`.
 
         return True the file was previously tracked, False otherwise.
         """
         ret = self._map.set_untracked(filename)
         if ret:
             self._dirty = True
         return ret
 
     @requires_no_parents_change
     def set_clean(self, filename, parentfiledata=None):
         """record that the current state of the file on disk is known to be clean"""
         self._dirty = True
         if parentfiledata:
             (mode, size, mtime) = parentfiledata
         else:
             (mode, size, mtime) = self._get_filedata(filename)
         if not self._map[filename].tracked:
             self._check_new_tracked_filename(filename)
         self._map.set_clean(filename, mode, size, mtime)
         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
 
     @requires_no_parents_change
     def set_possibly_dirty(self, filename):
         """record that the current state of the file on disk is unknown"""
         self._dirty = True
         self._map.set_possibly_dirty(filename)
 
     @requires_parents_change
     def update_file_p1(
         self,
         filename,
         p1_tracked,
     ):
         """Set a file as tracked in the parent (or not)
 
         This is to be called when adjust the dirstate to a new parent after an history
         rewriting operation.
 
         It should not be called during a merge (p2 != nullid) and only within
         a `with dirstate.parentchange():` context.
         """
         if self.in_merge:
             msg = b'update_file_reference should not be called when merging'
             raise error.ProgrammingError(msg)
         entry = self._map.get(filename)
         if entry is None:
             wc_tracked = False
         else:
             wc_tracked = entry.tracked
         if not (p1_tracked or wc_tracked):
             # the file is no longer relevant to anyone
             if self._map.get(filename) is not None:
                 self._map.reset_state(filename)
                 self._dirty = True
         elif (not p1_tracked) and wc_tracked:
             if entry is not None and entry.added:
                 return  # avoid dropping copy information (maybe?)
 
         parentfiledata = None
         if wc_tracked and p1_tracked:
             parentfiledata = self._get_filedata(filename)
 
         self._map.reset_state(
             filename,
             wc_tracked,
             p1_tracked,
             # the underlying reference might have changed, we will have to
             # check it.
             has_meaningful_mtime=False,
             parentfiledata=parentfiledata,
         )
         if (
             parentfiledata is not None
             and parentfiledata[2] > 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 = parentfiledata[2]
 
     @requires_parents_change
     def update_file(
         self,
         filename,
         wc_tracked,
         p1_tracked,
-        p2_tracked=False,
-        merged=False,
-        clean_p1=False,
-        clean_p2=False,
+        p2_info=False,
         possibly_dirty=False,
         parentfiledata=None,
     ):
         """update the information about a file in the dirstate
 
         This is to be called when the direstates parent changes to keep track
         of what is the file situation in regards to the working copy and its parent.
 
         This function must be called within a `dirstate.parentchange` context.
 
         note: the API is at an early stage and we might need to adjust it
         depending of what information ends up being relevant and useful to
         other processing.
         """
-        if merged and (clean_p1 or clean_p2):
-            msg = b'`merged` argument incompatible with `clean_p1`/`clean_p2`'
-            raise error.ProgrammingError(msg)
 
         # note: I do not think we need to double check name clash here since we
         # are in a update/merge case that should already have taken care of
         # this. The test agrees
 
         self._dirty = True
 
         need_parent_file_data = (
-            not (possibly_dirty or clean_p2 or merged)
-            and wc_tracked
-            and p1_tracked
+            not possibly_dirty and not p2_info and wc_tracked and p1_tracked
         )
 
         # this mean we are doing call for file we do not really care about the
         # data (eg: added or removed), however this should be a minor overhead
         # compared to the overall update process calling this.
         if need_parent_file_data:
             if parentfiledata is None:
                 parentfiledata = self._get_filedata(filename)
             mtime = parentfiledata[2]
 
             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
 
         self._map.reset_state(
             filename,
             wc_tracked,
             p1_tracked,
-            p2_info=merged or clean_p2,
+            p2_info=p2_info,
             has_meaningful_mtime=not possibly_dirty,
             parentfiledata=parentfiledata,
         )
         if (
             parentfiledata is not None
             and parentfiledata[2] > 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 = parentfiledata[2]
 
     def _check_new_tracked_filename(self, filename):
         scmutil.checkfilename(filename)
         if self._map.hastrackeddir(filename):
             msg = _(b'directory %r already in dirstate')
             msg %= pycompat.bytestr(filename)
             raise error.Abort(msg)
         # shadows
         for d in pathutil.finddirs(filename):
             if self._map.hastrackeddir(d):
                 break
             entry = self._map.get(d)
             if entry is not None and not entry.removed:
                 msg = _(b'file %r in dirstate clashes with %r')
                 msg %= (pycompat.bytestr(d), pycompat.bytestr(filename))
                 raise error.Abort(msg)
 
     def _get_filedata(self, filename):
         """returns"""
         s = os.lstat(self._join(filename))
         mode = s.st_mode
         size = s.st_size
         mtime = s[stat.ST_MTIME]
         return (mode, size, mtime)
 
     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._dirty = True
 
     def rebuild(self, parent, allfiles, changedfiles=None):
         if changedfiles is None:
             # Rebuild entire dirstate
             to_lookup = allfiles
             to_drop = []
             lastnormaltime = self._lastnormaltime
             self.clear()
             self._lastnormaltime = lastnormaltime
         elif len(changedfiles) < 10:
             # Avoid turning allfiles into a set, which can be expensive if it's
             # large.
             to_lookup = []
             to_drop = []
             for f in changedfiles:
                 if f in allfiles:
                     to_lookup.append(f)
                 else:
                     to_drop.append(f)
         else:
             changedfilesset = set(changedfiles)
             to_lookup = changedfilesset & set(allfiles)
             to_drop = changedfilesset - to_lookup
 
         if self._origpl is None:
             self._origpl = self._pl
         self._map.setparents(parent, self._nodeconstants.nullid)
 
         for f in to_lookup:
 
             if self.in_merge:
                 self.set_tracked(f)
             else:
                 self._map.reset_state(
                     f,
                     wc_tracked=True,
                     p1_tracked=True,
                 )
         for f in to_drop:
             self._map.reset_state(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
 
             # record when mtime start to be ambiguous
             now = _getfsnow(self._opener)
 
             # delay writing in-memory changes out
             tr.addfilegenerator(
                 b'dirstate',
                 (self._filename,),
                 lambda f: self._writedirstate(tr, f, now=now),
                 location=b'plain',
             )
             return
 
         st = self._opener(filename, b"w", atomictemp=True, checkambig=True)
         self._writedirstate(tr, 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, tr, st, now=None):
         # 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
 
         if now is 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.need_delay(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(tr, st, now)
         self._lastnormaltime = 0
         self._dirty = False
 
     def _dirignore(self, f):
         if self._ignore(f):
             return True
         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
 
         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
                     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):
                         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)
         if matchtdir:
             for d in work:
                 matchtdir(d[0])
             for d in dirsnotfound:
                 matchtdir(d)
 
         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 _rust_status(self, matcher, list_clean, list_ignored, list_unknown):
         # 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(b"worker", b"numcpus")
         if numcpus is not None:
             encoding.environ.setdefault(b'RAYON_NUM_THREADS', b'%d' % numcpus)
 
         workers_enabled = self._ui.configbool(b"worker", b"enabled", True)
         if not workers_enabled:
             encoding.environ[b"RAYON_NUM_THREADS"] = b"1"
 
         (
             lookup,
             modified,
             added,
             removed,
             deleted,
             clean,
             ignored,
             unknown,
             warnings,
             bad,
             traversed,
             dirty,
         ) = rustmod.status(
             self._map._map,
             matcher,
             self._rootdir,
             self._ignorefiles(),
             self._checkexec,
             self._lastnormaltime,
             bool(list_clean),
             bool(list_ignored),
             bool(list_unknown),
             bool(matcher.traversedir),
         )
 
         self._dirty |= dirty
 
         if matcher.traversedir:
             for dir in traversed:
                 matcher.traversedir(dir)
 
         if self._ui.warn:
             for item in warnings:
                 if isinstance(item, tuple):
                     file_path, syntax = item
                     msg = _(b"%s: ignoring invalid syntax '%s'\n") % (
                         file_path,
                         syntax,
                     )
                     self._ui.warn(msg)
                 else:
                     msg = _(b"skipping unreadable pattern file '%s': %s\n")
                     self._ui.warn(
                         msg
                         % (
                             pathutil.canonpath(
                                 self._rootdir, self._rootdir, item
                             ),
                             b"No such file or directory",
                         )
                     )
 
         for (fn, message) in bad:
             matcher.bad(fn, encoding.strtolocal(message))
 
         status = scmutil.status(
             modified=modified,
             added=added,
             removed=removed,
             deleted=deleted,
             unknown=unknown,
             ignored=ignored,
             clean=clean,
         )
         return (lookup, status)
 
     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
 
         allowed_matchers = (
             matchmod.alwaysmatcher,
             matchmod.exactmatcher,
             matchmod.includematcher,
         )
 
         if rustmod is None:
             use_rust = False
         elif self._checkcase:
             # Case-insensitive filesystems are not handled yet
             use_rust = False
         elif subrepos:
             use_rust = False
         elif sparse.enabled:
             use_rust = False
         elif not isinstance(match, allowed_matchers):
             # Some matchers have yet to be implemented
             use_rust = False
 
         if use_rust:
             try:
                 return self._rust_status(
                     match, listclean, listignored, listunknown
                 )
             except rustmod.FallbackError:
                 pass
 
         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)
             mode = t.mode
             size = t.size
             time = t.mtime
 
             if not st and t.tracked:
                 dadd(fn)
             elif t.merged:
                 madd(fn)
             elif t.added:
                 aadd(fn)
             elif t.removed:
                 radd(fn)
             elif t.tracked:
                 if (
                     size >= 0
                     and (
                         (size != st.st_size and size != st.st_size & _rangemask)
                         or ((mode ^ st.st_mode) & 0o100 and checkexec)
                     )
                     or t.from_p2
                     or fn in copymap
                 ):
                     if stat.S_ISLNK(st.st_mode) and size != st.st_size:
                         # issue6456: Size returned may be longer due to
                         # encryption on EXT-4 fscrypt, undecided.
                         ladd(fn)
                     else:
                         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)
         status = scmutil.status(
             modified, added, removed, deleted, unknown, ignored, clean
         )
         return (lookup, status)
 
     def matches(self, match):
         """
         return files in the dirstate (in whatever state) filtered by match
         """
         dmap = self._map
         if rustmod is not None:
             dmap = self._map._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(
                 tr,
                 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,),
                 lambda f: self._writedirstate(tr, f),
                 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)
 
     def verify(self, m1, m2):
         """check the dirstate content again the parent manifest and yield errors"""
         missing_from_p1 = b"%s in state %s, but not in manifest1\n"
         unexpected_in_p1 = b"%s in state %s, but also in manifest1\n"
         missing_from_ps = b"%s in state %s, but not in either manifest\n"
         missing_from_ds = b"%s in manifest1, but listed as state %s\n"
         for f, entry in self.items():
             state = entry.state
             if state in b"nr" and f not in m1:
                 yield (missing_from_p1, f, state)
             if state in b"a" and f in m1:
                 yield (unexpected_in_p1, f, state)
             if state in b"m" and f not in m1 and f not in m2:
                 yield (missing_from_ps, f, state)
         for f in m1:
             state = self.get_entry(f).state
             if state not in b"nrm":
                 yield (missing_from_ds, f, state)

mercurial/mergestate.py

@@ -1,876 +1,871 b''
 from __future__ import absolute_import
 
 import collections
 import errno
 import shutil
 import struct
 
 from .i18n import _
 from .node import (
     bin,
     hex,
     nullrev,
 )
 from . import (
     error,
     filemerge,
     pycompat,
     util,
 )
 from .utils import hashutil
 
 _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)
 
 
 def _filectxorabsent(hexnode, ctx, f):
     if hexnode == ctx.repo().nodeconstants.nullhex:
         return filemerge.absentfilectx(ctx, f)
     else:
         return ctx[f]
 
 
 # Merge state record types. See ``mergestate`` docs for more.
 
 ####
 # merge records which records metadata about a current merge
 # exists only once in a mergestate
 #####
 RECORD_LOCAL = b'L'
 RECORD_OTHER = b'O'
 # record merge labels
 RECORD_LABELS = b'l'
 
 #####
 # record extra information about files, with one entry containing info about one
 # file. Hence, multiple of them can exists
 #####
 RECORD_FILE_VALUES = b'f'
 
 #####
 # merge records which represents state of individual merges of files/folders
 # These are top level records for each entry containing merge related info.
 # Each record of these has info about one file. Hence multiple of them can
 # exists
 #####
 RECORD_MERGED = b'F'
 RECORD_CHANGEDELETE_CONFLICT = b'C'
 # the path was dir on one side of merge and file on another
 RECORD_PATH_CONFLICT = b'P'
 
 #####
 # possible state which a merge entry can have. These are stored inside top-level
 # merge records mentioned just above.
 #####
 MERGE_RECORD_UNRESOLVED = b'u'
 MERGE_RECORD_RESOLVED = b'r'
 MERGE_RECORD_UNRESOLVED_PATH = b'pu'
 MERGE_RECORD_RESOLVED_PATH = b'pr'
 # represents that the file was automatically merged in favor
 # of other version. This info is used on commit.
 # This is now deprecated and commit related information is now
 # stored in RECORD_FILE_VALUES
 MERGE_RECORD_MERGED_OTHER = b'o'
 
 #####
 # top level record which stores other unknown records. Multiple of these can
 # exists
 #####
 RECORD_OVERRIDE = b't'
 
 #####
 # legacy records which are no longer used but kept to prevent breaking BC
 #####
 # This record was release in 5.4 and usage was removed in 5.5
 LEGACY_RECORD_RESOLVED_OTHER = b'R'
 # This record was release in 3.7 and usage was removed in 5.6
 LEGACY_RECORD_DRIVER_RESOLVED = b'd'
 # This record was release in 3.7 and usage was removed in 5.6
 LEGACY_MERGE_DRIVER_STATE = b'm'
 # This record was release in 3.7 and usage was removed in 5.6
 LEGACY_MERGE_DRIVER_MERGE = 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'
 # the file was absent on local side before merge and we should
 # keep it absent (absent means file not present, it can be a result
 # of file deletion, rename etc.)
 ACTION_KEEP_ABSENT = b'ka'
 # the file is absent on the ancestor and remote side of the merge
 # hence this file is new and we should keep it
 ACTION_KEEP_NEW = b'kn'
 ACTION_EXEC = b'e'
 ACTION_CREATED_MERGE = b'cm'
 
 # actions which are no op
 NO_OP_ACTIONS = (
     ACTION_KEEP,
     ACTION_KEEP_ABSENT,
     ACTION_KEEP_NEW,
 )
 
 
 class _mergestate_base(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
     P: a path conflict (file vs directory)
     f: a (filename, dictionary) tuple of optional values for a given file
     l: the labels for the parts of the merge.
 
     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
     o: file was merged in favor of other parent of merge (DEPRECATED)
 
     The resolve command transitions between 'u' and 'r' for conflicts and
     'pu' and 'pr' for path conflicts.
     """
 
     def __init__(self, repo):
         """Initialize the merge state.
 
         Do not use this directly! Instead call read() or clean()."""
         self._repo = repo
         self._state = {}
         self._stateextras = collections.defaultdict(dict)
         self._local = None
         self._other = None
         self._labels = None
         # contains a mapping of form:
         # {filename : (merge_return_value, action_to_be_performed}
         # these are results of re-running merge process
         # this dict is used to perform actions on dirstate caused by re-running
         # the merge
         self._results = {}
         self._dirty = False
 
     def reset(self):
         pass
 
     def start(self, node, other, labels=None):
         self._local = node
         self._other = other
         self._labels = labels
 
     @util.propertycache
     def local(self):
         if self._local is None:
             msg = b"local accessed but self._local isn't set"
             raise error.ProgrammingError(msg)
         return self._local
 
     @util.propertycache
     def localctx(self):
         return self._repo[self.local]
 
     @util.propertycache
     def other(self):
         if self._other is None:
             msg = b"other accessed but self._other isn't set"
             raise error.ProgrammingError(msg)
         return self._other
 
     @util.propertycache
     def otherctx(self):
         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."
         """
         return bool(self._local) or bool(self._state)
 
     def commit(self):
         """Write current state on disk (if necessary)"""
 
     @staticmethod
     def getlocalkey(path):
         """hash the path of a local file context for storage in the .hg/merge
         directory."""
 
         return hex(hashutil.sha1(path).digest())
 
     def _make_backup(self, fctx, localkey):
         raise NotImplementedError()
 
     def _restore_backup(self, fctx, localkey, flags):
         raise NotImplementedError()
 
     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 = self._repo.nodeconstants.nullhex
         else:
             localkey = mergestate.getlocalkey(fcl.path())
             self._make_backup(fcl, localkey)
         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 addpathconflict(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 addcommitinfo(self, path, data):
         """stores information which is required at commit
         into _stateextras"""
         self._stateextras[path].update(data)
         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 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 allextras(self):
         """return all extras information stored with the mergestate"""
         return self._stateextras
 
     def extras(self, filename):
         """return extras stored with the mergestate for the given filename"""
         return self._stateextras[filename]
 
     def _resolve(self, preresolve, dfile, wctx):
         """rerun merge process for file path `dfile`.
         Returns whether the merge was completed and the return value of merge
         obtained from filemerge._filemerge().
         """
         if self[dfile] in (
             MERGE_RECORD_RESOLVED,
             LEGACY_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 = _filectxorabsent(localkey, wctx, dfile)
         fco = _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.rev() == nullrev 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 != self._repo.nodeconstants.nullhex:
                 self._restore_backup(wctx[dfile], localkey, flags)
             else:
                 wctx[dfile].remove(ignoremissing=True)
             complete, merge_ret, deleted = filemerge.premerge(
                 self._repo,
                 wctx,
                 self._local,
                 lfile,
                 fcd,
                 fco,
                 fca,
                 labels=self._labels,
             )
         else:
             complete, merge_ret, deleted = filemerge.filemerge(
                 self._repo,
                 wctx,
                 self._local,
                 lfile,
                 fcd,
                 fco,
                 fca,
                 labels=self._labels,
             )
         if merge_ret is None:
             # If return value of merge is None, then there are no real conflict
             del self._state[dfile]
             self._dirty = True
         elif not merge_ret:
             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] = merge_ret, action
 
         return complete, merge_ret
 
     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
 
 
 class mergestate(_mergestate_base):
 
     statepathv1 = b'merge/state'
     statepathv2 = b'merge/state2'
 
     @staticmethod
     def clean(repo):
         """Initialize a brand new merge state, removing any existing state on
         disk."""
         ms = mergestate(repo)
         ms.reset()
         return ms
 
     @staticmethod
     def read(repo):
         """Initialize the merge state, reading it from disk."""
         ms = mergestate(repo)
         ms._read()
         return ms
 
     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.
         """
         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 == LEGACY_MERGE_DRIVER_STATE:
                 pass
             elif rtype in (
                 RECORD_MERGED,
                 RECORD_CHANGEDELETE_CONFLICT,
                 RECORD_PATH_CONFLICT,
                 LEGACY_MERGE_DRIVER_MERGE,
                 LEGACY_RECORD_RESOLVED_OTHER,
             ):
                 bits = record.split(b'\0')
                 # merge entry type MERGE_RECORD_MERGED_OTHER is deprecated
                 # and we now store related information in _stateextras, so
                 # lets write to _stateextras directly
                 if bits[1] == MERGE_RECORD_MERGED_OTHER:
                     self._stateextras[bits[0]][b'filenode-source'] = b'other'
                 else:
                     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)
 
         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
 
     def commit(self):
         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)))
         # 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] 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] == self._repo.nodeconstants.nullhex
                 or v[6] == self._repo.nodeconstants.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()
 
     def _make_backup(self, fctx, localkey):
         self._repo.vfs.write(b'merge/' + localkey, fctx.data())
 
     def _restore_backup(self, fctx, localkey, flags):
         with self._repo.vfs(b'merge/' + localkey) as f:
             fctx.write(f.read(), flags)
 
     def reset(self):
         shutil.rmtree(self._repo.vfs.join(b'merge'), True)
 
 
 class memmergestate(_mergestate_base):
     def __init__(self, repo):
         super(memmergestate, self).__init__(repo)
         self._backups = {}
 
     def _make_backup(self, fctx, localkey):
         self._backups[localkey] = fctx.data()
 
     def _restore_backup(self, fctx, localkey, flags):
         fctx.write(self._backups[localkey], flags)
 
 
 def recordupdates(repo, actions, branchmerge, getfiledata):
     """record merge actions to the dirstate"""
     # remove (must come first)
     for f, args, msg in actions.get(ACTION_REMOVE, []):
         if branchmerge:
             repo.dirstate.update_file(f, p1_tracked=True, wc_tracked=False)
         else:
             repo.dirstate.update_file(f, p1_tracked=False, wc_tracked=False)
 
     # forget (must come first)
     for f, args, msg in actions.get(ACTION_FORGET, []):
         repo.dirstate.update_file(f, p1_tracked=False, wc_tracked=False)
 
     # resolve path conflicts
     for f, args, msg in actions.get(ACTION_PATH_CONFLICT_RESOLVE, []):
         (f0, origf0) = args
         repo.dirstate.update_file(f, p1_tracked=False, wc_tracked=True)
         repo.dirstate.copy(origf0, f)
         if f0 == origf0:
             repo.dirstate.update_file(f0, p1_tracked=True, wc_tracked=False)
         else:
             repo.dirstate.update_file(f0, p1_tracked=False, wc_tracked=False)
 
     # re-add
     for f, args, msg in actions.get(ACTION_ADD, []):
         repo.dirstate.update_file(f, p1_tracked=False, wc_tracked=True)
 
     # re-add/mark as modified
     for f, args, msg in actions.get(ACTION_ADD_MODIFIED, []):
         if branchmerge:
             repo.dirstate.update_file(
                 f, p1_tracked=True, wc_tracked=True, possibly_dirty=True
             )
         else:
             repo.dirstate.update_file(f, p1_tracked=False, wc_tracked=True)
 
     # exec change
     for f, args, msg in actions.get(ACTION_EXEC, []):
         repo.dirstate.update_file(
             f, p1_tracked=True, wc_tracked=True, possibly_dirty=True
         )
 
     # keep
     for f, args, msg in actions.get(ACTION_KEEP, []):
         pass
 
     # keep deleted
     for f, args, msg in actions.get(ACTION_KEEP_ABSENT, []):
         pass
 
     # keep new
     for f, args, msg in actions.get(ACTION_KEEP_NEW, []):
         pass
 
     # get
     for f, args, msg in actions.get(ACTION_GET, []):
         if branchmerge:
             # tracked in p1 can be True also but update_file should not care
             old_entry = repo.dirstate.get_entry(f)
             p1_tracked = old_entry.any_tracked and not old_entry.added
             repo.dirstate.update_file(
                 f,
                 p1_tracked=p1_tracked,
-                p2_tracked=True,
                 wc_tracked=True,
-                clean_p2=not p1_tracked,
-                merged=p1_tracked,
+                p2_info=True,
             )
         else:
             parentfiledata = getfiledata[f] if getfiledata else None
             repo.dirstate.update_file(
                 f,
                 p1_tracked=True,
                 wc_tracked=True,
                 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
             p1_tracked = f1 == f
-            p2_tracked = f2 == f
             repo.dirstate.update_file(
                 f,
                 p1_tracked=p1_tracked,
-                p2_tracked=p2_tracked,
                 wc_tracked=True,
-                merged=p1_tracked,
-                clean_p2=not p1_tracked,
+                p2_info=True,
             )
             if f1 != f2:  # copy/rename
                 if move:
                     repo.dirstate.update_file(
                         f1, p1_tracked=True, wc_tracked=False
                     )
                 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.update_file(
                     f, p1_tracked=True, wc_tracked=True, possibly_dirty=True
                 )
             if move:
                 repo.dirstate.update_file(
                     f1, p1_tracked=False, wc_tracked=False
                 )
 
     # directory rename, move local
     for f, args, msg in actions.get(ACTION_DIR_RENAME_MOVE_LOCAL, []):
         f0, flag = args
         if branchmerge:
             repo.dirstate.update_file(f, p1_tracked=False, wc_tracked=True)
             repo.dirstate.update_file(f0, p1_tracked=True, wc_tracked=False)
             repo.dirstate.copy(f0, f)
         else:
             repo.dirstate.update_file(f, p1_tracked=True, wc_tracked=True)
             repo.dirstate.update_file(f0, p1_tracked=False, wc_tracked=False)
 
     # directory rename, get
     for f, args, msg in actions.get(ACTION_LOCAL_DIR_RENAME_GET, []):
         f0, flag = args
         if branchmerge:
             repo.dirstate.update_file(f, p1_tracked=False, wc_tracked=True)
             repo.dirstate.copy(f0, f)
         else:
             repo.dirstate.update_file(f, p1_tracked=True, wc_tracked=True)