dirstate: infer the 'n' state from `possibly_dirty`

This flag is only used with 'n' so lets set the state based on that parameter in
this case.

Differential Revision: https://phab.mercurial-scm.org/D10970

mercurial/dirstate.py

@@ -1,1439 +1,1439 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
 
 dirstatetuple = parsers.dirstatetuple
 
 
 class repocache(filecache):
     """filecache for files in .hg/"""
 
     def join(self, obj, fname):
         return obj._opener.join(fname)
 
 
 class rootcache(filecache):
     """filecache for files in the repository root"""
 
     def join(self, obj, fname):
         return obj._join(fname)
 
 
 def _getfsnow(vfs):
     '''Get "now" timestamp on filesystem'''
     tmpfd, tmpname = vfs.mkstemp()
     try:
         return os.fstat(tmpfd)[stat.ST_MTIME]
     finally:
         os.close(tmpfd)
         vfs.unlink(tmpname)
 
 
 @interfaceutil.implementer(intdirstate.idirstate)
 class dirstate(object):
     def __init__(
         self,
         opener,
         ui,
         root,
         validate,
         sparsematchfn,
         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._updatedfiles = set()
         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.
         """
         entry = self._map.get(key)
         if entry is not None:
             return entry.state
         return b'?'
 
     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 directories(self):
         return self._map.directories()
 
     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
         self._map.setparents(p1, p2)
         copies = {}
         if (
             oldp2 != self._nodeconstants.nullid
             and p2 == self._nodeconstants.nullid
         ):
             candidatefiles = self._map.non_normal_or_other_parent_paths()
 
             for f in candidatefiles:
                 s = self._map.get(f)
                 if s is None:
                     continue
 
                 # Discard "merged" markers when moving away from a merge state
                 if s.merged:
                     source = self._map.copymap.get(f)
                     if source:
                         copies[f] = source
                     self.normallookup(f)
                 # Also fix up otherparent markers
                 elif s.from_p2:
                     source = self._map.copymap.get(f)
                     if source:
                         copies[f] = source
                     self.add(f)
         return copies
 
     def setbranch(self, branch):
         self.__class__._branch.set(self, encoding.fromlocal(branch))
         f = self._opener(b'branch', b'w', atomictemp=True, checkambig=True)
         try:
             f.write(self._branch + b'\n')
             f.close()
 
             # make sure filecache has the correct stat info for _branch after
             # replacing the underlying file
             ce = self._filecache[b'_branch']
             if ce:
                 ce.refresh()
         except:  # re-raises
             f.discard()
             raise
 
     def invalidate(self):
         """Causes the next access to reread the dirstate.
 
         This is different from localrepo.invalidatedirstate() because it always
         rereads the dirstate. Use localrepo.invalidatedirstate() if you want to
         check whether the dirstate has changed before rereading it."""
 
         for a in ("_map", "_branch", "_ignore"):
             if a in self.__dict__:
                 delattr(self, a)
         self._lastnormaltime = 0
         self._dirty = False
         self._updatedfiles.clear()
         self._parentwriters = 0
         self._origpl = None
 
     def copy(self, source, dest):
         """Mark dest as a copy of source. Unmark dest if source is None."""
         if source == dest:
             return
         self._dirty = True
         if source is not None:
             self._map.copymap[dest] = source
             self._updatedfiles.add(source)
             self._updatedfiles.add(dest)
         elif self._map.copymap.pop(dest, None):
             self._updatedfiles.add(dest)
 
     def copied(self, file):
         return self._map.copymap.get(file, None)
 
     def copies(self):
         return self._map.copymap
 
     def _addpath(
         self,
         f,
         state=None,
         mode=0,
         size=None,
         mtime=None,
         added=False,
         merged=False,
         from_p2=False,
         possibly_dirty=False,
     ):
         entry = self._map.get(f)
         if added or entry is not None and entry.removed:
             scmutil.checkfilename(f)
             if self._map.hastrackeddir(f):
                 msg = _(b'directory %r already in dirstate')
                 msg %= pycompat.bytestr(f)
                 raise error.Abort(msg)
             # shadows
             for d in pathutil.finddirs(f):
                 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(f))
                     raise error.Abort(msg)
         self._dirty = True
         self._updatedfiles.add(f)
         self._map.addfile(
             f,
             state=state,
             mode=mode,
             size=size,
             mtime=mtime,
             added=added,
             merged=merged,
             from_p2=from_p2,
             possibly_dirty=possibly_dirty,
         )
 
     def normal(self, f, parentfiledata=None):
         """Mark a file normal and clean.
 
         parentfiledata: (mode, size, mtime) of the clean file
 
         parentfiledata should be computed from memory (for mode,
         size), as or close as possible from the point where we
         determined the file was clean, to limit the risk of the
         file having been changed by an external process between the
         moment where the file was determined to be clean and now."""
         if parentfiledata:
             (mode, size, mtime) = parentfiledata
         else:
             s = os.lstat(self._join(f))
             mode = s.st_mode
             size = s.st_size
             mtime = s[stat.ST_MTIME]
         self._addpath(f, b'n', mode, size, mtime)
         self._map.copymap.pop(f, None)
         if f in self._map.nonnormalset:
             self._map.nonnormalset.remove(f)
         if mtime > self._lastnormaltime:
             # Remember the most recent modification timeslot for status(),
             # to make sure we won't miss future size-preserving file content
             # modifications that happen within the same timeslot.
             self._lastnormaltime = mtime
 
     def normallookup(self, f):
         '''Mark a file normal, but possibly dirty.'''
         if self.in_merge:
             # if there is a merge going on and the file was either
             # "merged" or coming from other parent (-2) before
             # being removed, restore that state.
             entry = self._map.get(f)
             if entry is not None:
                 # XXX this should probably be dealt with a a lower level
                 # (see `merged_removed` and `from_p2_removed`)
                 if entry.merged_removed or entry.from_p2_removed:
                     source = self._map.copymap.get(f)
                     if entry.merged_removed:
                         self.merge(f)
                     elif entry.from_p2_removed:
                         self.otherparent(f)
                     if source is not None:
                         self.copy(source, f)
                     return
                 elif entry.merged or entry.from_p2:
                     return
-        self._addpath(f, b'n', 0, possibly_dirty=True)
+        self._addpath(f, possibly_dirty=True)
         self._map.copymap.pop(f, None)
 
     def otherparent(self, f):
         '''Mark as coming from the other parent, always dirty.'''
         if not self.in_merge:
             msg = _(b"setting %r to other parent only allowed in merges") % f
             raise error.Abort(msg)
         if f in self and self[f] == b'n':
             # merge-like
             self._addpath(f, merged=True)
         else:
             # add-like
             self._addpath(f, b'n', 0, from_p2=True)
         self._map.copymap.pop(f, None)
 
     def add(self, f):
         '''Mark a file added.'''
         self._addpath(f, added=True)
         self._map.copymap.pop(f, None)
 
     def remove(self, f):
         '''Mark a file removed.'''
         self._dirty = True
         self._updatedfiles.add(f)
         self._map.removefile(f, in_merge=self.in_merge)
 
     def merge(self, f):
         '''Mark a file merged.'''
         if not self.in_merge:
             return self.normallookup(f)
         return self.otherparent(f)
 
     def drop(self, f):
         '''Drop a file from the dirstate'''
         oldstate = self[f]
         if self._map.dropfile(f, oldstate):
             self._dirty = True
             self._updatedfiles.add(f)
             self._map.copymap.pop(f, None)
 
     def _discoverpath(self, path, normed, ignoremissing, exists, storemap):
         if exists is None:
             exists = os.path.lexists(os.path.join(self._root, path))
         if not exists:
             # Maybe a path component exists
             if not ignoremissing and b'/' in path:
                 d, f = path.rsplit(b'/', 1)
                 d = self._normalize(d, False, ignoremissing, None)
                 folded = d + b"/" + f
             else:
                 # No path components, preserve original case
                 folded = path
         else:
             # recursively normalize leading directory components
             # against dirstate
             if b'/' in normed:
                 d, f = normed.rsplit(b'/', 1)
                 d = self._normalize(d, False, ignoremissing, True)
                 r = self._root + b"/" + d
                 folded = d + b"/" + util.fspath(f, r)
             else:
                 folded = util.fspath(normed, self._root)
             storemap[normed] = folded
 
         return folded
 
     def _normalizefile(self, path, isknown, ignoremissing=False, exists=None):
         normed = util.normcase(path)
         folded = self._map.filefoldmap.get(normed, None)
         if folded is None:
             if isknown:
                 folded = path
             else:
                 folded = self._discoverpath(
                     path, normed, ignoremissing, exists, self._map.filefoldmap
                 )
         return folded
 
     def _normalize(self, path, isknown, ignoremissing=False, exists=None):
         normed = util.normcase(path)
         folded = self._map.filefoldmap.get(normed, None)
         if folded is None:
             folded = self._map.dirfoldmap.get(normed, None)
         if folded is None:
             if isknown:
                 folded = path
             else:
                 # store discovered result in dirfoldmap so that future
                 # normalizefile calls don't start matching directories
                 folded = self._discoverpath(
                     path, normed, ignoremissing, exists, self._map.dirfoldmap
                 )
         return folded
 
     def normalize(self, path, isknown=False, ignoremissing=False):
         """
         normalize the case of a pathname when on a casefolding filesystem
 
         isknown specifies whether the filename came from walking the
         disk, to avoid extra filesystem access.
 
         If ignoremissing is True, missing path are returned
         unchanged. Otherwise, we try harder to normalize possibly
         existing path components.
 
         The normalized case is determined based on the following precedence:
 
         - version of name already stored in the dirstate
         - version of name stored on disk
         - version provided via command arguments
         """
 
         if self._checkcase:
             return self._normalize(path, isknown, ignoremissing)
         return path
 
     def clear(self):
         self._map.clear()
         self._lastnormaltime = 0
         self._updatedfiles.clear()
         self._dirty = True
 
     def rebuild(self, parent, allfiles, changedfiles=None):
         if changedfiles is None:
             # Rebuild entire dirstate
             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:
             self.normallookup(f)
         for f in to_drop:
             self.drop(f)
 
         self._dirty = True
 
     def identity(self):
         """Return identity of dirstate itself to detect changing in storage
 
         If identity of previous dirstate is equal to this, writing
         changes based on the former dirstate out can keep consistency.
         """
         return self._map.identity
 
     def write(self, tr):
         if not self._dirty:
             return
 
         filename = self._filename
         if tr:
             # 'dirstate.write()' is not only for writing in-memory
             # changes out, but also for dropping ambiguous timestamp.
             # delayed writing re-raise "ambiguous timestamp issue".
             # See also the wiki page below for detail:
             # https://www.mercurial-scm.org/wiki/DirstateTransactionPlan
 
             # emulate dropping timestamp in 'parsers.pack_dirstate'
             now = _getfsnow(self._opener)
             self._map.clearambiguoustimes(self._updatedfiles, now)
 
             # emulate that all 'dirstate.normal' results are written out
             self._lastnormaltime = 0
             self._updatedfiles.clear()
 
             # delay writing in-memory changes out
             tr.addfilegenerator(
                 b'dirstate',
                 (self._filename,),
                 self._writedirstate,
                 location=b'plain',
             )
             return
 
         st = self._opener(filename, b"w", atomictemp=True, checkambig=True)
         self._writedirstate(st)
 
     def addparentchangecallback(self, category, callback):
         """add a callback to be called when the wd parents are changed
 
         Callback will be called with the following arguments:
             dirstate, (oldp1, oldp2), (newp1, newp2)
 
         Category is a unique identifier to allow overwriting an old callback
         with a newer callback.
         """
         self._plchangecallbacks[category] = callback
 
     def _writedirstate(self, st):
         # notify callbacks about parents change
         if self._origpl is not None and self._origpl != self._pl:
             for c, callback in sorted(
                 pycompat.iteritems(self._plchangecallbacks)
             ):
                 callback(self, self._origpl, self._pl)
             self._origpl = None
         # use the modification time of the newly created temporary file as the
         # filesystem's notion of 'now'
         now = util.fstat(st)[stat.ST_MTIME] & _rangemask
 
         # enough 'delaywrite' prevents 'pack_dirstate' from dropping
         # timestamp of each entries in dirstate, because of 'now > mtime'
         delaywrite = self._ui.configint(b'debug', b'dirstate.delaywrite')
         if delaywrite > 0:
             # do we have any files to delay for?
             for f, e in pycompat.iteritems(self._map):
                 if e.state == b'n' and e[3] == now:
                     import time  # to avoid useless import
 
                     # rather than sleep n seconds, sleep until the next
                     # multiple of n seconds
                     clock = time.time()
                     start = int(clock) - (int(clock) % delaywrite)
                     end = start + delaywrite
                     time.sleep(end - clock)
                     now = end  # trust our estimate that the end is near now
                     break
 
         self._map.write(st, now)
         self._lastnormaltime = 0
         self._dirty = False
 
     def _dirignore(self, f):
         if self._ignore(f):
             return True
         for p in 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._rustmap,
             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)
             state = t.state
             mode = t[1]
             size = t[2]
             time = t[3]
 
             if not st and state in b"nma":
                 dadd(fn)
             elif state == b'n':
                 if (
                     size >= 0
                     and (
                         (size != st.st_size and size != st.st_size & _rangemask)
                         or ((mode ^ st.st_mode) & 0o100 and checkexec)
                     )
                     or 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)
             elif t.merged:
                 madd(fn)
             elif t.added:
                 aadd(fn)
             elif t.removed:
                 radd(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._rustmap
 
         if match.always():
             return dmap.keys()
         files = match.files()
         if match.isexact():
             # fast path -- filter the other way around, since typically files is
             # much smaller than dmap
             return [f for f in files if f in dmap]
         if match.prefix() and all(fn in dmap for fn in files):
             # fast path -- all the values are known to be files, so just return
             # that
             return list(files)
         return [f for f in dmap if match(f)]
 
     def _actualfilename(self, tr):
         if tr:
             return self._pendingfilename
         else:
             return self._filename
 
     def savebackup(self, tr, backupname):
         '''Save current dirstate into backup file'''
         filename = self._actualfilename(tr)
         assert backupname != filename
 
         # use '_writedirstate' instead of 'write' to write changes certainly,
         # because the latter omits writing out if transaction is running.
         # output file will be used to create backup of dirstate at this point.
         if self._dirty or not self._opener.exists(filename):
             self._writedirstate(
                 self._opener(filename, b"w", atomictemp=True, checkambig=True)
             )
 
         if tr:
             # ensure that subsequent tr.writepending returns True for
             # changes written out above, even if dirstate is never
             # changed after this
             tr.addfilegenerator(
                 b'dirstate',
                 (self._filename,),
                 self._writedirstate,
                 location=b'plain',
             )
 
             # ensure that pending file written above is unlinked at
             # failure, even if tr.writepending isn't invoked until the
             # end of this transaction
             tr.registertmp(filename, location=b'plain')
 
         self._opener.tryunlink(backupname)
         # hardlink backup is okay because _writedirstate is always called
         # with an "atomictemp=True" file.
         util.copyfile(
             self._opener.join(filename),
             self._opener.join(backupname),
             hardlink=True,
         )
 
     def restorebackup(self, tr, backupname):
         '''Restore dirstate by backup file'''
         # this "invalidate()" prevents "wlock.release()" from writing
         # changes of dirstate out after restoring from backup file
         self.invalidate()
         filename = self._actualfilename(tr)
         o = self._opener
         if util.samefile(o.join(backupname), o.join(filename)):
             o.unlink(backupname)
         else:
             o.rename(backupname, filename, checkambig=True)
 
     def clearbackup(self, tr, backupname):
         '''Clear backup file'''
         self._opener.unlink(backupname)

mercurial/dirstatemap.py

@@ -1,683 +1,684 b''
 # dirstatemap.py
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import errno
 
 from .i18n import _
 
 from . import (
     error,
     pathutil,
     policy,
     pycompat,
     txnutil,
     util,
 )
 
 parsers = policy.importmod('parsers')
 rustmod = policy.importrust('dirstate')
 
 propertycache = util.propertycache
 
 dirstatetuple = parsers.dirstatetuple
 
 
 # a special value used internally for `size` if the file come from the other parent
 FROM_P2 = -2
 
 # a special value used internally for `size` if the file is modified/merged/added
 NONNORMAL = -1
 
 # a special value used internally for `time` if the time is ambigeous
 AMBIGUOUS_TIME = -1
 
 rangemask = 0x7FFFFFFF
 
 
 class dirstatemap(object):
     """Map encapsulating the dirstate's contents.
 
     The dirstate contains the following state:
 
     - `identity` is the identity of the dirstate file, which can be used to
       detect when changes have occurred to the dirstate file.
 
     - `parents` is a pair containing the parents of the working copy. The
       parents are updated by calling `setparents`.
 
     - the state map maps filenames to tuples of (state, mode, size, mtime),
       where state is a single character representing 'normal', 'added',
       'removed', or 'merged'. It is read by treating the dirstate as a
       dict.  File state is updated by calling the `addfile`, `removefile` and
       `dropfile` methods.
 
     - `copymap` maps destination filenames to their source filename.
 
     The dirstate also provides the following views onto the state:
 
     - `nonnormalset` is a set of the filenames that have state other
       than 'normal', or are normal but have an mtime of -1 ('normallookup').
 
     - `otherparentset` is a set of the filenames that are marked as coming
       from the second parent when the dirstate is currently being merged.
 
     - `filefoldmap` is a dict mapping normalized filenames to the denormalized
       form that they appear as in the dirstate.
 
     - `dirfoldmap` is a dict mapping normalized directory names to the
       denormalized form that they appear as in the dirstate.
     """
 
     def __init__(self, ui, opener, root, nodeconstants, use_dirstate_v2):
         self._ui = ui
         self._opener = opener
         self._root = root
         self._filename = b'dirstate'
         self._nodelen = 20
         self._nodeconstants = nodeconstants
         assert (
             not use_dirstate_v2
         ), "should have detected unsupported requirement"
 
         self._parents = None
         self._dirtyparents = False
 
         # for consistent view between _pl() and _read() invocations
         self._pendingmode = None
 
     @propertycache
     def _map(self):
         self._map = {}
         self.read()
         return self._map
 
     @propertycache
     def copymap(self):
         self.copymap = {}
         self._map
         return self.copymap
 
     def directories(self):
         # Rust / dirstate-v2 only
         return []
 
     def clear(self):
         self._map.clear()
         self.copymap.clear()
         self.setparents(self._nodeconstants.nullid, self._nodeconstants.nullid)
         util.clearcachedproperty(self, b"_dirs")
         util.clearcachedproperty(self, b"_alldirs")
         util.clearcachedproperty(self, b"filefoldmap")
         util.clearcachedproperty(self, b"dirfoldmap")
         util.clearcachedproperty(self, b"nonnormalset")
         util.clearcachedproperty(self, b"otherparentset")
 
     def items(self):
         return pycompat.iteritems(self._map)
 
     # forward for python2,3 compat
     iteritems = items
 
     def __len__(self):
         return len(self._map)
 
     def __iter__(self):
         return iter(self._map)
 
     def get(self, key, default=None):
         return self._map.get(key, default)
 
     def __contains__(self, key):
         return key in self._map
 
     def __getitem__(self, key):
         return self._map[key]
 
     def keys(self):
         return self._map.keys()
 
     def preload(self):
         """Loads the underlying data, if it's not already loaded"""
         self._map
 
     def addfile(
         self,
         f,
         state=None,
         mode=0,
         size=None,
         mtime=None,
         added=False,
         merged=False,
         from_p2=False,
         possibly_dirty=False,
     ):
         """Add a tracked file to the dirstate."""
         if added:
             assert not merged
             assert not possibly_dirty
             assert not from_p2
             state = b'a'
             size = NONNORMAL
             mtime = AMBIGUOUS_TIME
         elif merged:
             assert not possibly_dirty
             assert not from_p2
             state = b'm'
             size = FROM_P2
             mtime = AMBIGUOUS_TIME
         elif from_p2:
             assert not possibly_dirty
             size = FROM_P2
             mtime = AMBIGUOUS_TIME
         elif possibly_dirty:
+            state = b'n'
             size = NONNORMAL
             mtime = AMBIGUOUS_TIME
         else:
             assert state != b'a'
             assert size != FROM_P2
             assert size != NONNORMAL
             size = size & rangemask
             mtime = mtime & rangemask
         assert state is not None
         assert size is not None
         assert mtime is not None
         old_entry = self.get(f)
         if (
             old_entry is None or old_entry.removed
         ) and "_dirs" in self.__dict__:
             self._dirs.addpath(f)
         if old_entry is None and "_alldirs" in self.__dict__:
             self._alldirs.addpath(f)
         self._map[f] = dirstatetuple(state, mode, size, mtime)
         if state != b'n' or mtime == AMBIGUOUS_TIME:
             self.nonnormalset.add(f)
         if size == FROM_P2:
             self.otherparentset.add(f)
 
     def removefile(self, f, in_merge=False):
         """
         Mark a file as removed in the dirstate.
 
         The `size` parameter is used to store sentinel values that indicate
         the file's previous state.  In the future, we should refactor this
         to be more explicit about what that state is.
         """
         entry = self.get(f)
         size = 0
         if in_merge:
             # XXX we should not be able to have 'm' state and 'FROM_P2' if not
             # during a merge. So I (marmoute) am not sure we need the
             # conditionnal at all. Adding double checking this with assert
             # would be nice.
             if entry is not None:
                 # backup the previous state
                 if entry.merged:  # merge
                     size = NONNORMAL
                 elif entry[0] == b'n' and entry.from_p2:
                     size = FROM_P2
                     self.otherparentset.add(f)
         if size == 0:
             self.copymap.pop(f, None)
 
         if entry is not None and entry[0] != b'r' and "_dirs" in self.__dict__:
             self._dirs.delpath(f)
         if entry is None and "_alldirs" in self.__dict__:
             self._alldirs.addpath(f)
         if "filefoldmap" in self.__dict__:
             normed = util.normcase(f)
             self.filefoldmap.pop(normed, None)
         self._map[f] = dirstatetuple(b'r', 0, size, 0)
         self.nonnormalset.add(f)
 
     def dropfile(self, f, oldstate):
         """
         Remove a file from the dirstate.  Returns True if the file was
         previously recorded.
         """
         exists = self._map.pop(f, None) is not None
         if exists:
             if oldstate != b"r" and "_dirs" in self.__dict__:
                 self._dirs.delpath(f)
             if "_alldirs" in self.__dict__:
                 self._alldirs.delpath(f)
         if "filefoldmap" in self.__dict__:
             normed = util.normcase(f)
             self.filefoldmap.pop(normed, None)
         self.nonnormalset.discard(f)
         return exists
 
     def clearambiguoustimes(self, files, now):
         for f in files:
             e = self.get(f)
             if e is not None and e[0] == b'n' and e[3] == now:
                 self._map[f] = dirstatetuple(e[0], e[1], e[2], AMBIGUOUS_TIME)
                 self.nonnormalset.add(f)
 
     def nonnormalentries(self):
         '''Compute the nonnormal dirstate entries from the dmap'''
         try:
             return parsers.nonnormalotherparententries(self._map)
         except AttributeError:
             nonnorm = set()
             otherparent = set()
             for fname, e in pycompat.iteritems(self._map):
                 if e[0] != b'n' or e[3] == AMBIGUOUS_TIME:
                     nonnorm.add(fname)
                 if e[0] == b'n' and e[2] == FROM_P2:
                     otherparent.add(fname)
             return nonnorm, otherparent
 
     @propertycache
     def filefoldmap(self):
         """Returns a dictionary mapping normalized case paths to their
         non-normalized versions.
         """
         try:
             makefilefoldmap = parsers.make_file_foldmap
         except AttributeError:
             pass
         else:
             return makefilefoldmap(
                 self._map, util.normcasespec, util.normcasefallback
             )
 
         f = {}
         normcase = util.normcase
         for name, s in pycompat.iteritems(self._map):
             if s[0] != b'r':
                 f[normcase(name)] = name
         f[b'.'] = b'.'  # prevents useless util.fspath() invocation
         return f
 
     def hastrackeddir(self, d):
         """
         Returns True if the dirstate contains a tracked (not removed) file
         in this directory.
         """
         return d in self._dirs
 
     def hasdir(self, d):
         """
         Returns True if the dirstate contains a file (tracked or removed)
         in this directory.
         """
         return d in self._alldirs
 
     @propertycache
     def _dirs(self):
         return pathutil.dirs(self._map, b'r')
 
     @propertycache
     def _alldirs(self):
         return pathutil.dirs(self._map)
 
     def _opendirstatefile(self):
         fp, mode = txnutil.trypending(self._root, self._opener, self._filename)
         if self._pendingmode is not None and self._pendingmode != mode:
             fp.close()
             raise error.Abort(
                 _(b'working directory state may be changed parallelly')
             )
         self._pendingmode = mode
         return fp
 
     def parents(self):
         if not self._parents:
             try:
                 fp = self._opendirstatefile()
                 st = fp.read(2 * self._nodelen)
                 fp.close()
             except IOError as err:
                 if err.errno != errno.ENOENT:
                     raise
                 # File doesn't exist, so the current state is empty
                 st = b''
 
             l = len(st)
             if l == self._nodelen * 2:
                 self._parents = (
                     st[: self._nodelen],
                     st[self._nodelen : 2 * self._nodelen],
                 )
             elif l == 0:
                 self._parents = (
                     self._nodeconstants.nullid,
                     self._nodeconstants.nullid,
                 )
             else:
                 raise error.Abort(
                     _(b'working directory state appears damaged!')
                 )
 
         return self._parents
 
     def setparents(self, p1, p2):
         self._parents = (p1, p2)
         self._dirtyparents = True
 
     def read(self):
         # ignore HG_PENDING because identity is used only for writing
         self.identity = util.filestat.frompath(
             self._opener.join(self._filename)
         )
 
         try:
             fp = self._opendirstatefile()
             try:
                 st = fp.read()
             finally:
                 fp.close()
         except IOError as err:
             if err.errno != errno.ENOENT:
                 raise
             return
         if not st:
             return
 
         if util.safehasattr(parsers, b'dict_new_presized'):
             # Make an estimate of the number of files in the dirstate based on
             # its size. This trades wasting some memory for avoiding costly
             # resizes. Each entry have a prefix of 17 bytes followed by one or
             # two path names. Studies on various large-scale real-world repositories
             # found 54 bytes a reasonable upper limit for the average path names.
             # Copy entries are ignored for the sake of this estimate.
             self._map = parsers.dict_new_presized(len(st) // 71)
 
         # Python's garbage collector triggers a GC each time a certain number
         # of container objects (the number being defined by
         # gc.get_threshold()) are allocated. parse_dirstate creates a tuple
         # for each file in the dirstate. The C version then immediately marks
         # them as not to be tracked by the collector. However, this has no
         # effect on when GCs are triggered, only on what objects the GC looks
         # into. This means that O(number of files) GCs are unavoidable.
         # Depending on when in the process's lifetime the dirstate is parsed,
         # this can get very expensive. As a workaround, disable GC while
         # parsing the dirstate.
         #
         # (we cannot decorate the function directly since it is in a C module)
         parse_dirstate = util.nogc(parsers.parse_dirstate)
         p = parse_dirstate(self._map, self.copymap, st)
         if not self._dirtyparents:
             self.setparents(*p)
 
         # Avoid excess attribute lookups by fast pathing certain checks
         self.__contains__ = self._map.__contains__
         self.__getitem__ = self._map.__getitem__
         self.get = self._map.get
 
     def write(self, st, now):
         st.write(
             parsers.pack_dirstate(self._map, self.copymap, self.parents(), now)
         )
         st.close()
         self._dirtyparents = False
         self.nonnormalset, self.otherparentset = self.nonnormalentries()
 
     @propertycache
     def nonnormalset(self):
         nonnorm, otherparents = self.nonnormalentries()
         self.otherparentset = otherparents
         return nonnorm
 
     @propertycache
     def otherparentset(self):
         nonnorm, otherparents = self.nonnormalentries()
         self.nonnormalset = nonnorm
         return otherparents
 
     def non_normal_or_other_parent_paths(self):
         return self.nonnormalset.union(self.otherparentset)
 
     @propertycache
     def identity(self):
         self._map
         return self.identity
 
     @propertycache
     def dirfoldmap(self):
         f = {}
         normcase = util.normcase
         for name in self._dirs:
             f[normcase(name)] = name
         return f
 
 
 if rustmod is not None:
 
     class dirstatemap(object):
         def __init__(self, ui, opener, root, nodeconstants, use_dirstate_v2):
             self._use_dirstate_v2 = use_dirstate_v2
             self._nodeconstants = nodeconstants
             self._ui = ui
             self._opener = opener
             self._root = root
             self._filename = b'dirstate'
             self._nodelen = 20  # Also update Rust code when changing this!
             self._parents = None
             self._dirtyparents = False
 
             # for consistent view between _pl() and _read() invocations
             self._pendingmode = None
 
             self._use_dirstate_tree = self._ui.configbool(
                 b"experimental",
                 b"dirstate-tree.in-memory",
                 False,
             )
 
         def addfile(
             self,
             f,
             state=None,
             mode=0,
             size=None,
             mtime=None,
             added=False,
             merged=False,
             from_p2=False,
             possibly_dirty=False,
         ):
             return self._rustmap.addfile(
                 f,
                 state,
                 mode,
                 size,
                 mtime,
                 added,
                 merged,
                 from_p2,
                 possibly_dirty,
             )
 
         def removefile(self, *args, **kwargs):
             return self._rustmap.removefile(*args, **kwargs)
 
         def dropfile(self, *args, **kwargs):
             return self._rustmap.dropfile(*args, **kwargs)
 
         def clearambiguoustimes(self, *args, **kwargs):
             return self._rustmap.clearambiguoustimes(*args, **kwargs)
 
         def nonnormalentries(self):
             return self._rustmap.nonnormalentries()
 
         def get(self, *args, **kwargs):
             return self._rustmap.get(*args, **kwargs)
 
         @property
         def copymap(self):
             return self._rustmap.copymap()
 
         def directories(self):
             return self._rustmap.directories()
 
         def preload(self):
             self._rustmap
 
         def clear(self):
             self._rustmap.clear()
             self.setparents(
                 self._nodeconstants.nullid, self._nodeconstants.nullid
             )
             util.clearcachedproperty(self, b"_dirs")
             util.clearcachedproperty(self, b"_alldirs")
             util.clearcachedproperty(self, b"dirfoldmap")
 
         def items(self):
             return self._rustmap.items()
 
         def keys(self):
             return iter(self._rustmap)
 
         def __contains__(self, key):
             return key in self._rustmap
 
         def __getitem__(self, item):
             return self._rustmap[item]
 
         def __len__(self):
             return len(self._rustmap)
 
         def __iter__(self):
             return iter(self._rustmap)
 
         # forward for python2,3 compat
         iteritems = items
 
         def _opendirstatefile(self):
             fp, mode = txnutil.trypending(
                 self._root, self._opener, self._filename
             )
             if self._pendingmode is not None and self._pendingmode != mode:
                 fp.close()
                 raise error.Abort(
                     _(b'working directory state may be changed parallelly')
                 )
             self._pendingmode = mode
             return fp
 
         def setparents(self, p1, p2):
             self._parents = (p1, p2)
             self._dirtyparents = True
 
         def parents(self):
             if not self._parents:
                 if self._use_dirstate_v2:
                     offset = len(rustmod.V2_FORMAT_MARKER)
                 else:
                     offset = 0
                 read_len = offset + self._nodelen * 2
                 try:
                     fp = self._opendirstatefile()
                     st = fp.read(read_len)
                     fp.close()
                 except IOError as err:
                     if err.errno != errno.ENOENT:
                         raise
                     # File doesn't exist, so the current state is empty
                     st = b''
 
                 l = len(st)
                 if l == read_len:
                     st = st[offset:]
                     self._parents = (
                         st[: self._nodelen],
                         st[self._nodelen : 2 * self._nodelen],
                     )
                 elif l == 0:
                     self._parents = (
                         self._nodeconstants.nullid,
                         self._nodeconstants.nullid,
                     )
                 else:
                     raise error.Abort(
                         _(b'working directory state appears damaged!')
                     )
 
             return self._parents
 
         @propertycache
         def _rustmap(self):
             """
             Fills the Dirstatemap when called.
             """
             # ignore HG_PENDING because identity is used only for writing
             self.identity = util.filestat.frompath(
                 self._opener.join(self._filename)
             )
 
             try:
                 fp = self._opendirstatefile()
                 try:
                     st = fp.read()
                 finally:
                     fp.close()
             except IOError as err:
                 if err.errno != errno.ENOENT:
                     raise
                 st = b''
 
             self._rustmap, parents = rustmod.DirstateMap.new(
                 self._use_dirstate_tree, self._use_dirstate_v2, st
             )
 
             if parents and not self._dirtyparents:
                 self.setparents(*parents)
 
             self.__contains__ = self._rustmap.__contains__
             self.__getitem__ = self._rustmap.__getitem__
             self.get = self._rustmap.get
             return self._rustmap
 
         def write(self, st, now):
             parents = self.parents()
             packed = self._rustmap.write(
                 self._use_dirstate_v2, parents[0], parents[1], now
             )
             st.write(packed)
             st.close()
             self._dirtyparents = False
 
         @propertycache
         def filefoldmap(self):
             """Returns a dictionary mapping normalized case paths to their
             non-normalized versions.
             """
             return self._rustmap.filefoldmapasdict()
 
         def hastrackeddir(self, d):
             return self._rustmap.hastrackeddir(d)
 
         def hasdir(self, d):
             return self._rustmap.hasdir(d)
 
         @propertycache
         def identity(self):
             self._rustmap
             return self.identity
 
         @property
         def nonnormalset(self):
             nonnorm = self._rustmap.non_normal_entries()
             return nonnorm
 
         @propertycache
         def otherparentset(self):
             otherparents = self._rustmap.other_parent_entries()
             return otherparents
 
         def non_normal_or_other_parent_paths(self):
             return self._rustmap.non_normal_or_other_parent_paths()
 
         @propertycache
         def dirfoldmap(self):
             f = {}
             normcase = util.normcase
             for name, _pseudo_entry in self.directories():
                 f[normcase(name)] = name
             return f

rust/hg-core/src/dirstate/dirstate_map.rs

@@ -1,475 +1,476 b''
 // dirstate_map.rs
 //
 // Copyright 2019 Raphaël Gomès <[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.
 
 use crate::dirstate::parsers::Timestamp;
 use crate::{
     dirstate::EntryState,
     dirstate::MTIME_UNSET,
     dirstate::SIZE_FROM_OTHER_PARENT,
     dirstate::SIZE_NON_NORMAL,
     dirstate::V1_RANGEMASK,
     pack_dirstate, parse_dirstate,
     utils::hg_path::{HgPath, HgPathBuf},
     CopyMap, DirsMultiset, DirstateEntry, DirstateError, DirstateParents,
     StateMap,
 };
 use micro_timer::timed;
 use std::collections::HashSet;
 use std::iter::FromIterator;
 use std::ops::Deref;
 
 #[derive(Default)]
 pub struct DirstateMap {
     state_map: StateMap,
     pub copy_map: CopyMap,
     pub dirs: Option<DirsMultiset>,
     pub all_dirs: Option<DirsMultiset>,
     non_normal_set: Option<HashSet<HgPathBuf>>,
     other_parent_set: Option<HashSet<HgPathBuf>>,
 }
 
 /// Should only really be used in python interface code, for clarity
 impl Deref for DirstateMap {
     type Target = StateMap;
 
     fn deref(&self) -> &Self::Target {
         &self.state_map
     }
 }
 
 impl FromIterator<(HgPathBuf, DirstateEntry)> for DirstateMap {
     fn from_iter<I: IntoIterator<Item = (HgPathBuf, DirstateEntry)>>(
         iter: I,
     ) -> Self {
         Self {
             state_map: iter.into_iter().collect(),
             ..Self::default()
         }
     }
 }
 
 impl DirstateMap {
     pub fn new() -> Self {
         Self::default()
     }
 
     pub fn clear(&mut self) {
         self.state_map = StateMap::default();
         self.copy_map.clear();
         self.non_normal_set = None;
         self.other_parent_set = None;
     }
 
     /// Add a tracked file to the dirstate
     pub fn add_file(
         &mut self,
         filename: &HgPath,
         entry: DirstateEntry,
         // XXX once the dust settle this should probably become an enum
         added: bool,
         merged: bool,
         from_p2: bool,
         possibly_dirty: bool,
     ) -> Result<(), DirstateError> {
         let mut entry = entry;
         if added {
             assert!(!merged);
             assert!(!possibly_dirty);
             assert!(!from_p2);
             entry.state = EntryState::Added;
             entry.size = SIZE_NON_NORMAL;
             entry.mtime = MTIME_UNSET;
         } else if merged {
             assert!(!possibly_dirty);
             assert!(!from_p2);
             entry.state = EntryState::Merged;
             entry.size = SIZE_FROM_OTHER_PARENT;
             entry.mtime = MTIME_UNSET;
         } else if from_p2 {
             assert!(!possibly_dirty);
             entry.size = SIZE_FROM_OTHER_PARENT;
             entry.mtime = MTIME_UNSET;
         } else if possibly_dirty {
+            entry.state = EntryState::Normal;
             entry.size = SIZE_NON_NORMAL;
             entry.mtime = MTIME_UNSET;
         } else {
             entry.size = entry.size & V1_RANGEMASK;
             entry.mtime = entry.mtime & V1_RANGEMASK;
         }
         let old_state = match self.get(filename) {
             Some(e) => e.state,
             None => EntryState::Unknown,
         };
         if old_state == EntryState::Unknown || old_state == EntryState::Removed
         {
             if let Some(ref mut dirs) = self.dirs {
                 dirs.add_path(filename)?;
             }
         }
         if old_state == EntryState::Unknown {
             if let Some(ref mut all_dirs) = self.all_dirs {
                 all_dirs.add_path(filename)?;
             }
         }
         self.state_map.insert(filename.to_owned(), entry.to_owned());
 
         if entry.is_non_normal() {
             self.get_non_normal_other_parent_entries()
                 .0
                 .insert(filename.to_owned());
         }
 
         if entry.is_from_other_parent() {
             self.get_non_normal_other_parent_entries()
                 .1
                 .insert(filename.to_owned());
         }
         Ok(())
     }
 
     /// Mark a file as removed in the dirstate.
     ///
     /// The `size` parameter is used to store sentinel values that indicate
     /// the file's previous state.  In the future, we should refactor this
     /// to be more explicit about what that state is.
     pub fn remove_file(
         &mut self,
         filename: &HgPath,
         in_merge: bool,
     ) -> Result<(), DirstateError> {
         let old_entry_opt = self.get(filename);
         let old_state = match old_entry_opt {
             Some(e) => e.state,
             None => EntryState::Unknown,
         };
         let mut size = 0;
         if in_merge {
             // XXX we should not be able to have 'm' state and 'FROM_P2' if not
             // during a merge. So I (marmoute) am not sure we need the
             // conditionnal at all. Adding double checking this with assert
             // would be nice.
             if let Some(old_entry) = old_entry_opt {
                 // backup the previous state
                 if old_entry.state == EntryState::Merged {
                     size = SIZE_NON_NORMAL;
                 } else if old_entry.state == EntryState::Normal
                     && old_entry.size == SIZE_FROM_OTHER_PARENT
                 {
                     // other parent
                     size = SIZE_FROM_OTHER_PARENT;
                     self.get_non_normal_other_parent_entries()
                         .1
                         .insert(filename.to_owned());
                 }
             }
         }
         if old_state != EntryState::Unknown && old_state != EntryState::Removed
         {
             if let Some(ref mut dirs) = self.dirs {
                 dirs.delete_path(filename)?;
             }
         }
         if old_state == EntryState::Unknown {
             if let Some(ref mut all_dirs) = self.all_dirs {
                 all_dirs.add_path(filename)?;
             }
         }
         if size == 0 {
             self.copy_map.remove(filename);
         }
 
         self.state_map.insert(
             filename.to_owned(),
             DirstateEntry {
                 state: EntryState::Removed,
                 mode: 0,
                 size,
                 mtime: 0,
             },
         );
         self.get_non_normal_other_parent_entries()
             .0
             .insert(filename.to_owned());
         Ok(())
     }
 
     /// Remove a file from the dirstate.
     /// Returns `true` if the file was previously recorded.
     pub fn drop_file(
         &mut self,
         filename: &HgPath,
         old_state: EntryState,
     ) -> Result<bool, DirstateError> {
         let exists = self.state_map.remove(filename).is_some();
 
         if exists {
             if old_state != EntryState::Removed {
                 if let Some(ref mut dirs) = self.dirs {
                     dirs.delete_path(filename)?;
                 }
             }
             if let Some(ref mut all_dirs) = self.all_dirs {
                 all_dirs.delete_path(filename)?;
             }
         }
         self.get_non_normal_other_parent_entries()
             .0
             .remove(filename);
 
         Ok(exists)
     }
 
     pub fn clear_ambiguous_times(
         &mut self,
         filenames: Vec<HgPathBuf>,
         now: i32,
     ) {
         for filename in filenames {
             if let Some(entry) = self.state_map.get_mut(&filename) {
                 if entry.clear_ambiguous_mtime(now) {
                     self.get_non_normal_other_parent_entries()
                         .0
                         .insert(filename.to_owned());
                 }
             }
         }
     }
 
     pub fn non_normal_entries_remove(&mut self, key: impl AsRef<HgPath>) {
         self.get_non_normal_other_parent_entries()
             .0
             .remove(key.as_ref());
     }
 
     pub fn non_normal_entries_union(
         &mut self,
         other: HashSet<HgPathBuf>,
     ) -> Vec<HgPathBuf> {
         self.get_non_normal_other_parent_entries()
             .0
             .union(&other)
             .map(ToOwned::to_owned)
             .collect()
     }
 
     pub fn get_non_normal_other_parent_entries(
         &mut self,
     ) -> (&mut HashSet<HgPathBuf>, &mut HashSet<HgPathBuf>) {
         self.set_non_normal_other_parent_entries(false);
         (
             self.non_normal_set.as_mut().unwrap(),
             self.other_parent_set.as_mut().unwrap(),
         )
     }
 
     /// Useful to get immutable references to those sets in contexts where
     /// you only have an immutable reference to the `DirstateMap`, like when
     /// sharing references with Python.
     ///
     /// TODO, get rid of this along with the other "setter/getter" stuff when
     /// a nice typestate plan is defined.
     ///
     /// # Panics
     ///
     /// Will panic if either set is `None`.
     pub fn get_non_normal_other_parent_entries_panic(
         &self,
     ) -> (&HashSet<HgPathBuf>, &HashSet<HgPathBuf>) {
         (
             self.non_normal_set.as_ref().unwrap(),
             self.other_parent_set.as_ref().unwrap(),
         )
     }
 
     pub fn set_non_normal_other_parent_entries(&mut self, force: bool) {
         if !force
             && self.non_normal_set.is_some()
             && self.other_parent_set.is_some()
         {
             return;
         }
         let mut non_normal = HashSet::new();
         let mut other_parent = HashSet::new();
 
         for (filename, entry) in self.state_map.iter() {
             if entry.is_non_normal() {
                 non_normal.insert(filename.to_owned());
             }
             if entry.is_from_other_parent() {
                 other_parent.insert(filename.to_owned());
             }
         }
         self.non_normal_set = Some(non_normal);
         self.other_parent_set = Some(other_parent);
     }
 
     /// Both of these setters and their uses appear to be the simplest way to
     /// emulate a Python lazy property, but it is ugly and unidiomatic.
     /// TODO One day, rewriting this struct using the typestate might be a
     /// good idea.
     pub fn set_all_dirs(&mut self) -> Result<(), DirstateError> {
         if self.all_dirs.is_none() {
             self.all_dirs = Some(DirsMultiset::from_dirstate(
                 self.state_map.iter().map(|(k, v)| Ok((k, *v))),
                 None,
             )?);
         }
         Ok(())
     }
 
     pub fn set_dirs(&mut self) -> Result<(), DirstateError> {
         if self.dirs.is_none() {
             self.dirs = Some(DirsMultiset::from_dirstate(
                 self.state_map.iter().map(|(k, v)| Ok((k, *v))),
                 Some(EntryState::Removed),
             )?);
         }
         Ok(())
     }
 
     pub fn has_tracked_dir(
         &mut self,
         directory: &HgPath,
     ) -> Result<bool, DirstateError> {
         self.set_dirs()?;
         Ok(self.dirs.as_ref().unwrap().contains(directory))
     }
 
     pub fn has_dir(
         &mut self,
         directory: &HgPath,
     ) -> Result<bool, DirstateError> {
         self.set_all_dirs()?;
         Ok(self.all_dirs.as_ref().unwrap().contains(directory))
     }
 
     #[timed]
     pub fn read(
         &mut self,
         file_contents: &[u8],
     ) -> Result<Option<DirstateParents>, DirstateError> {
         if file_contents.is_empty() {
             return Ok(None);
         }
 
         let (parents, entries, copies) = parse_dirstate(file_contents)?;
         self.state_map.extend(
             entries
                 .into_iter()
                 .map(|(path, entry)| (path.to_owned(), entry)),
         );
         self.copy_map.extend(
             copies
                 .into_iter()
                 .map(|(path, copy)| (path.to_owned(), copy.to_owned())),
         );
         Ok(Some(parents.clone()))
     }
 
     pub fn pack(
         &mut self,
         parents: DirstateParents,
         now: Timestamp,
     ) -> Result<Vec<u8>, DirstateError> {
         let packed =
             pack_dirstate(&mut self.state_map, &self.copy_map, parents, now)?;
 
         self.set_non_normal_other_parent_entries(true);
         Ok(packed)
     }
 }
 
 #[cfg(test)]
 mod tests {
     use super::*;
 
     #[test]
     fn test_dirs_multiset() {
         let mut map = DirstateMap::new();
         assert!(map.dirs.is_none());
         assert!(map.all_dirs.is_none());
 
         assert_eq!(map.has_dir(HgPath::new(b"nope")).unwrap(), false);
         assert!(map.all_dirs.is_some());
         assert!(map.dirs.is_none());
 
         assert_eq!(map.has_tracked_dir(HgPath::new(b"nope")).unwrap(), false);
         assert!(map.dirs.is_some());
     }
 
     #[test]
     fn test_add_file() {
         let mut map = DirstateMap::new();
 
         assert_eq!(0, map.len());
 
         map.add_file(
             HgPath::new(b"meh"),
             DirstateEntry {
                 state: EntryState::Normal,
                 mode: 1337,
                 mtime: 1337,
                 size: 1337,
             },
             false,
             false,
             false,
             false,
         )
         .unwrap();
 
         assert_eq!(1, map.len());
         assert_eq!(0, map.get_non_normal_other_parent_entries().0.len());
         assert_eq!(0, map.get_non_normal_other_parent_entries().1.len());
     }
 
     #[test]
     fn test_non_normal_other_parent_entries() {
         let mut map: DirstateMap = [
             (b"f1", (EntryState::Removed, 1337, 1337, 1337)),
             (b"f2", (EntryState::Normal, 1337, 1337, -1)),
             (b"f3", (EntryState::Normal, 1337, 1337, 1337)),
             (b"f4", (EntryState::Normal, 1337, -2, 1337)),
             (b"f5", (EntryState::Added, 1337, 1337, 1337)),
             (b"f6", (EntryState::Added, 1337, 1337, -1)),
             (b"f7", (EntryState::Merged, 1337, 1337, -1)),
             (b"f8", (EntryState::Merged, 1337, 1337, 1337)),
             (b"f9", (EntryState::Merged, 1337, -2, 1337)),
             (b"fa", (EntryState::Added, 1337, -2, 1337)),
             (b"fb", (EntryState::Removed, 1337, -2, 1337)),
         ]
         .iter()
         .map(|(fname, (state, mode, size, mtime))| {
             (
                 HgPathBuf::from_bytes(fname.as_ref()),
                 DirstateEntry {
                     state: *state,
                     mode: *mode,
                     size: *size,
                     mtime: *mtime,
                 },
             )
         })
         .collect();
 
         let mut non_normal = [
             b"f1", b"f2", b"f5", b"f6", b"f7", b"f8", b"f9", b"fa", b"fb",
         ]
         .iter()
         .map(|x| HgPathBuf::from_bytes(x.as_ref()))
         .collect();
 
         let mut other_parent = HashSet::new();
         other_parent.insert(HgPathBuf::from_bytes(b"f4"));
         let entries = map.get_non_normal_other_parent_entries();
 
         assert_eq!(
             (&mut non_normal, &mut other_parent),
             (entries.0, entries.1)
         );
     }
 }

rust/hg-core/src/dirstate_tree/dirstate_map.rs

@@ -1,1203 +1,1204 b''
 use bytes_cast::BytesCast;
 use micro_timer::timed;
 use std::borrow::Cow;
 use std::convert::TryInto;
 use std::path::PathBuf;
 
 use super::on_disk;
 use super::on_disk::DirstateV2ParseError;
 use super::path_with_basename::WithBasename;
 use crate::dirstate::parsers::pack_entry;
 use crate::dirstate::parsers::packed_entry_size;
 use crate::dirstate::parsers::parse_dirstate_entries;
 use crate::dirstate::parsers::Timestamp;
 use crate::dirstate::MTIME_UNSET;
 use crate::dirstate::SIZE_FROM_OTHER_PARENT;
 use crate::dirstate::SIZE_NON_NORMAL;
 use crate::dirstate::V1_RANGEMASK;
 use crate::matchers::Matcher;
 use crate::utils::hg_path::{HgPath, HgPathBuf};
 use crate::CopyMapIter;
 use crate::DirstateEntry;
 use crate::DirstateError;
 use crate::DirstateParents;
 use crate::DirstateStatus;
 use crate::EntryState;
 use crate::FastHashMap;
 use crate::PatternFileWarning;
 use crate::StateMapIter;
 use crate::StatusError;
 use crate::StatusOptions;
 
 pub struct DirstateMap<'on_disk> {
     /// Contents of the `.hg/dirstate` file
     pub(super) on_disk: &'on_disk [u8],
 
     pub(super) root: ChildNodes<'on_disk>,
 
     /// Number of nodes anywhere in the tree that have `.entry.is_some()`.
     pub(super) nodes_with_entry_count: u32,
 
     /// Number of nodes anywhere in the tree that have
     /// `.copy_source.is_some()`.
     pub(super) nodes_with_copy_source_count: u32,
 
     /// See on_disk::Header
     pub(super) ignore_patterns_hash: on_disk::IgnorePatternsHash,
 }
 
 /// Using a plain `HgPathBuf` of the full path from the repository root as a
 /// map key would also work: all paths in a given map have the same parent
 /// path, so comparing full paths gives the same result as comparing base
 /// names. However `HashMap` would waste time always re-hashing the same
 /// string prefix.
 pub(super) type NodeKey<'on_disk> = WithBasename<Cow<'on_disk, HgPath>>;
 
 /// Similar to `&'tree Cow<'on_disk, HgPath>`, but can also be returned
 /// for on-disk nodes that don’t actually have a `Cow` to borrow.
 pub(super) enum BorrowedPath<'tree, 'on_disk> {
     InMemory(&'tree HgPathBuf),
     OnDisk(&'on_disk HgPath),
 }
 
 pub(super) enum ChildNodes<'on_disk> {
     InMemory(FastHashMap<NodeKey<'on_disk>, Node<'on_disk>>),
     OnDisk(&'on_disk [on_disk::Node]),
 }
 
 pub(super) enum ChildNodesRef<'tree, 'on_disk> {
     InMemory(&'tree FastHashMap<NodeKey<'on_disk>, Node<'on_disk>>),
     OnDisk(&'on_disk [on_disk::Node]),
 }
 
 pub(super) enum NodeRef<'tree, 'on_disk> {
     InMemory(&'tree NodeKey<'on_disk>, &'tree Node<'on_disk>),
     OnDisk(&'on_disk on_disk::Node),
 }
 
 impl<'tree, 'on_disk> BorrowedPath<'tree, 'on_disk> {
     pub fn detach_from_tree(&self) -> Cow<'on_disk, HgPath> {
         match *self {
             BorrowedPath::InMemory(in_memory) => Cow::Owned(in_memory.clone()),
             BorrowedPath::OnDisk(on_disk) => Cow::Borrowed(on_disk),
         }
     }
 }
 
 impl<'tree, 'on_disk> std::ops::Deref for BorrowedPath<'tree, 'on_disk> {
     type Target = HgPath;
 
     fn deref(&self) -> &HgPath {
         match *self {
             BorrowedPath::InMemory(in_memory) => in_memory,
             BorrowedPath::OnDisk(on_disk) => on_disk,
         }
     }
 }
 
 impl Default for ChildNodes<'_> {
     fn default() -> Self {
         ChildNodes::InMemory(Default::default())
     }
 }
 
 impl<'on_disk> ChildNodes<'on_disk> {
     pub(super) fn as_ref<'tree>(
         &'tree self,
     ) -> ChildNodesRef<'tree, 'on_disk> {
         match self {
             ChildNodes::InMemory(nodes) => ChildNodesRef::InMemory(nodes),
             ChildNodes::OnDisk(nodes) => ChildNodesRef::OnDisk(nodes),
         }
     }
 
     pub(super) fn is_empty(&self) -> bool {
         match self {
             ChildNodes::InMemory(nodes) => nodes.is_empty(),
             ChildNodes::OnDisk(nodes) => nodes.is_empty(),
         }
     }
 
     pub(super) fn make_mut(
         &mut self,
         on_disk: &'on_disk [u8],
     ) -> Result<
         &mut FastHashMap<NodeKey<'on_disk>, Node<'on_disk>>,
         DirstateV2ParseError,
     > {
         match self {
             ChildNodes::InMemory(nodes) => Ok(nodes),
             ChildNodes::OnDisk(nodes) => {
                 let nodes = nodes
                     .iter()
                     .map(|node| {
                         Ok((
                             node.path(on_disk)?,
                             node.to_in_memory_node(on_disk)?,
                         ))
                     })
                     .collect::<Result<_, _>>()?;
                 *self = ChildNodes::InMemory(nodes);
                 match self {
                     ChildNodes::InMemory(nodes) => Ok(nodes),
                     ChildNodes::OnDisk(_) => unreachable!(),
                 }
             }
         }
     }
 }
 
 impl<'tree, 'on_disk> ChildNodesRef<'tree, 'on_disk> {
     pub(super) fn get(
         &self,
         base_name: &HgPath,
         on_disk: &'on_disk [u8],
     ) -> Result<Option<NodeRef<'tree, 'on_disk>>, DirstateV2ParseError> {
         match self {
             ChildNodesRef::InMemory(nodes) => Ok(nodes
                 .get_key_value(base_name)
                 .map(|(k, v)| NodeRef::InMemory(k, v))),
             ChildNodesRef::OnDisk(nodes) => {
                 let mut parse_result = Ok(());
                 let search_result = nodes.binary_search_by(|node| {
                     match node.base_name(on_disk) {
                         Ok(node_base_name) => node_base_name.cmp(base_name),
                         Err(e) => {
                             parse_result = Err(e);
                             // Dummy comparison result, `search_result` won’t
                             // be used since `parse_result` is an error
                             std::cmp::Ordering::Equal
                         }
                     }
                 });
                 parse_result.map(|()| {
                     search_result.ok().map(|i| NodeRef::OnDisk(&nodes[i]))
                 })
             }
         }
     }
 
     /// Iterate in undefined order
     pub(super) fn iter(
         &self,
     ) -> impl Iterator<Item = NodeRef<'tree, 'on_disk>> {
         match self {
             ChildNodesRef::InMemory(nodes) => itertools::Either::Left(
                 nodes.iter().map(|(k, v)| NodeRef::InMemory(k, v)),
             ),
             ChildNodesRef::OnDisk(nodes) => {
                 itertools::Either::Right(nodes.iter().map(NodeRef::OnDisk))
             }
         }
     }
 
     /// Iterate in parallel in undefined order
     pub(super) fn par_iter(
         &self,
     ) -> impl rayon::iter::ParallelIterator<Item = NodeRef<'tree, 'on_disk>>
     {
         use rayon::prelude::*;
         match self {
             ChildNodesRef::InMemory(nodes) => rayon::iter::Either::Left(
                 nodes.par_iter().map(|(k, v)| NodeRef::InMemory(k, v)),
             ),
             ChildNodesRef::OnDisk(nodes) => rayon::iter::Either::Right(
                 nodes.par_iter().map(NodeRef::OnDisk),
             ),
         }
     }
 
     pub(super) fn sorted(&self) -> Vec<NodeRef<'tree, 'on_disk>> {
         match self {
             ChildNodesRef::InMemory(nodes) => {
                 let mut vec: Vec<_> = nodes
                     .iter()
                     .map(|(k, v)| NodeRef::InMemory(k, v))
                     .collect();
                 fn sort_key<'a>(node: &'a NodeRef) -> &'a HgPath {
                     match node {
                         NodeRef::InMemory(path, _node) => path.base_name(),
                         NodeRef::OnDisk(_) => unreachable!(),
                     }
                 }
                 // `sort_unstable_by_key` doesn’t allow keys borrowing from the
                 // value: https://github.com/rust-lang/rust/issues/34162
                 vec.sort_unstable_by(|a, b| sort_key(a).cmp(sort_key(b)));
                 vec
             }
             ChildNodesRef::OnDisk(nodes) => {
                 // Nodes on disk are already sorted
                 nodes.iter().map(NodeRef::OnDisk).collect()
             }
         }
     }
 }
 
 impl<'tree, 'on_disk> NodeRef<'tree, 'on_disk> {
     pub(super) fn full_path(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<&'tree HgPath, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(path, _node) => Ok(path.full_path()),
             NodeRef::OnDisk(node) => node.full_path(on_disk),
         }
     }
 
     /// Returns a `BorrowedPath`, which can be turned into a `Cow<'on_disk,
     /// HgPath>` detached from `'tree`
     pub(super) fn full_path_borrowed(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<BorrowedPath<'tree, 'on_disk>, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(path, _node) => match path.full_path() {
                 Cow::Borrowed(on_disk) => Ok(BorrowedPath::OnDisk(on_disk)),
                 Cow::Owned(in_memory) => Ok(BorrowedPath::InMemory(in_memory)),
             },
             NodeRef::OnDisk(node) => {
                 Ok(BorrowedPath::OnDisk(node.full_path(on_disk)?))
             }
         }
     }
 
     pub(super) fn base_name(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<&'tree HgPath, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(path, _node) => Ok(path.base_name()),
             NodeRef::OnDisk(node) => node.base_name(on_disk),
         }
     }
 
     pub(super) fn children(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<ChildNodesRef<'tree, 'on_disk>, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(_path, node) => Ok(node.children.as_ref()),
             NodeRef::OnDisk(node) => {
                 Ok(ChildNodesRef::OnDisk(node.children(on_disk)?))
             }
         }
     }
 
     pub(super) fn has_copy_source(&self) -> bool {
         match self {
             NodeRef::InMemory(_path, node) => node.copy_source.is_some(),
             NodeRef::OnDisk(node) => node.has_copy_source(),
         }
     }
 
     pub(super) fn copy_source(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<Option<&'tree HgPath>, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(_path, node) => {
                 Ok(node.copy_source.as_ref().map(|s| &**s))
             }
             NodeRef::OnDisk(node) => node.copy_source(on_disk),
         }
     }
 
     pub(super) fn entry(
         &self,
     ) -> Result<Option<DirstateEntry>, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(_path, node) => {
                 Ok(node.data.as_entry().copied())
             }
             NodeRef::OnDisk(node) => node.entry(),
         }
     }
 
     pub(super) fn state(
         &self,
     ) -> Result<Option<EntryState>, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(_path, node) => {
                 Ok(node.data.as_entry().map(|entry| entry.state))
             }
             NodeRef::OnDisk(node) => node.state(),
         }
     }
 
     pub(super) fn cached_directory_mtime(
         &self,
     ) -> Option<&'tree on_disk::Timestamp> {
         match self {
             NodeRef::InMemory(_path, node) => match &node.data {
                 NodeData::CachedDirectory { mtime } => Some(mtime),
                 _ => None,
             },
             NodeRef::OnDisk(node) => node.cached_directory_mtime(),
         }
     }
 
     pub(super) fn descendants_with_entry_count(&self) -> u32 {
         match self {
             NodeRef::InMemory(_path, node) => {
                 node.descendants_with_entry_count
             }
             NodeRef::OnDisk(node) => node.descendants_with_entry_count.get(),
         }
     }
 
     pub(super) fn tracked_descendants_count(&self) -> u32 {
         match self {
             NodeRef::InMemory(_path, node) => node.tracked_descendants_count,
             NodeRef::OnDisk(node) => node.tracked_descendants_count.get(),
         }
     }
 }
 
 /// Represents a file or a directory
 #[derive(Default)]
 pub(super) struct Node<'on_disk> {
     pub(super) data: NodeData,
 
     pub(super) copy_source: Option<Cow<'on_disk, HgPath>>,
 
     pub(super) children: ChildNodes<'on_disk>,
 
     /// How many (non-inclusive) descendants of this node have an entry.
     pub(super) descendants_with_entry_count: u32,
 
     /// How many (non-inclusive) descendants of this node have an entry whose
     /// state is "tracked".
     pub(super) tracked_descendants_count: u32,
 }
 
 pub(super) enum NodeData {
     Entry(DirstateEntry),
     CachedDirectory { mtime: on_disk::Timestamp },
     None,
 }
 
 impl Default for NodeData {
     fn default() -> Self {
         NodeData::None
     }
 }
 
 impl NodeData {
     fn has_entry(&self) -> bool {
         match self {
             NodeData::Entry(_) => true,
             _ => false,
         }
     }
 
     fn as_entry(&self) -> Option<&DirstateEntry> {
         match self {
             NodeData::Entry(entry) => Some(entry),
             _ => None,
         }
     }
 }
 
 impl<'on_disk> DirstateMap<'on_disk> {
     pub(super) fn empty(on_disk: &'on_disk [u8]) -> Self {
         Self {
             on_disk,
             root: ChildNodes::default(),
             nodes_with_entry_count: 0,
             nodes_with_copy_source_count: 0,
             ignore_patterns_hash: [0; on_disk::IGNORE_PATTERNS_HASH_LEN],
         }
     }
 
     #[timed]
     pub fn new_v2(
         on_disk: &'on_disk [u8],
     ) -> Result<(Self, Option<DirstateParents>), DirstateError> {
         Ok(on_disk::read(on_disk)?)
     }
 
     #[timed]
     pub fn new_v1(
         on_disk: &'on_disk [u8],
     ) -> Result<(Self, Option<DirstateParents>), DirstateError> {
         let mut map = Self::empty(on_disk);
         if map.on_disk.is_empty() {
             return Ok((map, None));
         }
 
         let parents = parse_dirstate_entries(
             map.on_disk,
             |path, entry, copy_source| {
                 let tracked = entry.state.is_tracked();
                 let node = Self::get_or_insert_node(
                     map.on_disk,
                     &mut map.root,
                     path,
                     WithBasename::to_cow_borrowed,
                     |ancestor| {
                         if tracked {
                             ancestor.tracked_descendants_count += 1
                         }
                         ancestor.descendants_with_entry_count += 1
                     },
                 )?;
                 assert!(
                     !node.data.has_entry(),
                     "duplicate dirstate entry in read"
                 );
                 assert!(
                     node.copy_source.is_none(),
                     "duplicate dirstate entry in read"
                 );
                 node.data = NodeData::Entry(*entry);
                 node.copy_source = copy_source.map(Cow::Borrowed);
                 map.nodes_with_entry_count += 1;
                 if copy_source.is_some() {
                     map.nodes_with_copy_source_count += 1
                 }
                 Ok(())
             },
         )?;
         let parents = Some(parents.clone());
 
         Ok((map, parents))
     }
 
     fn get_node<'tree>(
         &'tree self,
         path: &HgPath,
     ) -> Result<Option<NodeRef<'tree, 'on_disk>>, DirstateV2ParseError> {
         let mut children = self.root.as_ref();
         let mut components = path.components();
         let mut component =
             components.next().expect("expected at least one components");
         loop {
             if let Some(child) = children.get(component, self.on_disk)? {
                 if let Some(next_component) = components.next() {
                     component = next_component;
                     children = child.children(self.on_disk)?;
                 } else {
                     return Ok(Some(child));
                 }
             } else {
                 return Ok(None);
             }
         }
     }
 
     /// Returns a mutable reference to the node at `path` if it exists
     ///
     /// This takes `root` instead of `&mut self` so that callers can mutate
     /// other fields while the returned borrow is still valid
     fn get_node_mut<'tree>(
         on_disk: &'on_disk [u8],
         root: &'tree mut ChildNodes<'on_disk>,
         path: &HgPath,
     ) -> Result<Option<&'tree mut Node<'on_disk>>, DirstateV2ParseError> {
         let mut children = root;
         let mut components = path.components();
         let mut component =
             components.next().expect("expected at least one components");
         loop {
             if let Some(child) = children.make_mut(on_disk)?.get_mut(component)
             {
                 if let Some(next_component) = components.next() {
                     component = next_component;
                     children = &mut child.children;
                 } else {
                     return Ok(Some(child));
                 }
             } else {
                 return Ok(None);
             }
         }
     }
 
     pub(super) fn get_or_insert<'tree, 'path>(
         &'tree mut self,
         path: &HgPath,
     ) -> Result<&'tree mut Node<'on_disk>, DirstateV2ParseError> {
         Self::get_or_insert_node(
             self.on_disk,
             &mut self.root,
             path,
             WithBasename::to_cow_owned,
             |_| {},
         )
     }
 
     pub(super) fn get_or_insert_node<'tree, 'path>(
         on_disk: &'on_disk [u8],
         root: &'tree mut ChildNodes<'on_disk>,
         path: &'path HgPath,
         to_cow: impl Fn(
             WithBasename<&'path HgPath>,
         ) -> WithBasename<Cow<'on_disk, HgPath>>,
         mut each_ancestor: impl FnMut(&mut Node),
     ) -> Result<&'tree mut Node<'on_disk>, DirstateV2ParseError> {
         let mut child_nodes = root;
         let mut inclusive_ancestor_paths =
             WithBasename::inclusive_ancestors_of(path);
         let mut ancestor_path = inclusive_ancestor_paths
             .next()
             .expect("expected at least one inclusive ancestor");
         loop {
             // TODO: can we avoid allocating an owned key in cases where the
             // map already contains that key, without introducing double
             // lookup?
             let child_node = child_nodes
                 .make_mut(on_disk)?
                 .entry(to_cow(ancestor_path))
                 .or_default();
             if let Some(next) = inclusive_ancestor_paths.next() {
                 each_ancestor(child_node);
                 ancestor_path = next;
                 child_nodes = &mut child_node.children;
             } else {
                 return Ok(child_node);
             }
         }
     }
 
     fn add_or_remove_file(
         &mut self,
         path: &HgPath,
         old_state: EntryState,
         new_entry: DirstateEntry,
     ) -> Result<(), DirstateV2ParseError> {
         let had_entry = old_state != EntryState::Unknown;
         let tracked_count_increment =
             match (old_state.is_tracked(), new_entry.state.is_tracked()) {
                 (false, true) => 1,
                 (true, false) => -1,
                 _ => 0,
             };
 
         let node = Self::get_or_insert_node(
             self.on_disk,
             &mut self.root,
             path,
             WithBasename::to_cow_owned,
             |ancestor| {
                 if !had_entry {
                     ancestor.descendants_with_entry_count += 1;
                 }
 
                 // We can’t use `+= increment` because the counter is unsigned,
                 // and we want debug builds to detect accidental underflow
                 // through zero
                 match tracked_count_increment {
                     1 => ancestor.tracked_descendants_count += 1,
                     -1 => ancestor.tracked_descendants_count -= 1,
                     _ => {}
                 }
             },
         )?;
         if !had_entry {
             self.nodes_with_entry_count += 1
         }
         node.data = NodeData::Entry(new_entry);
         Ok(())
     }
 
     fn iter_nodes<'tree>(
         &'tree self,
     ) -> impl Iterator<
         Item = Result<NodeRef<'tree, 'on_disk>, DirstateV2ParseError>,
     > + 'tree {
         // Depth first tree traversal.
         //
         // If we could afford internal iteration and recursion,
         // this would look like:
         //
         // ```
         // fn traverse_children(
         //     children: &ChildNodes,
         //     each: &mut impl FnMut(&Node),
         // ) {
         //     for child in children.values() {
         //         traverse_children(&child.children, each);
         //         each(child);
         //     }
         // }
         // ```
         //
         // However we want an external iterator and therefore can’t use the
         // call stack. Use an explicit stack instead:
         let mut stack = Vec::new();
         let mut iter = self.root.as_ref().iter();
         std::iter::from_fn(move || {
             while let Some(child_node) = iter.next() {
                 let children = match child_node.children(self.on_disk) {
                     Ok(children) => children,
                     Err(error) => return Some(Err(error)),
                 };
                 // Pseudo-recursion
                 let new_iter = children.iter();
                 let old_iter = std::mem::replace(&mut iter, new_iter);
                 stack.push((child_node, old_iter));
             }
             // Found the end of a `children.iter()` iterator.
             if let Some((child_node, next_iter)) = stack.pop() {
                 // "Return" from pseudo-recursion by restoring state from the
                 // explicit stack
                 iter = next_iter;
 
                 Some(Ok(child_node))
             } else {
                 // Reached the bottom of the stack, we’re done
                 None
             }
         })
     }
 
     fn clear_known_ambiguous_mtimes(
         &mut self,
         paths: &[impl AsRef<HgPath>],
     ) -> Result<(), DirstateV2ParseError> {
         for path in paths {
             if let Some(node) = Self::get_node_mut(
                 self.on_disk,
                 &mut self.root,
                 path.as_ref(),
             )? {
                 if let NodeData::Entry(entry) = &mut node.data {
                     entry.clear_mtime();
                 }
             }
         }
         Ok(())
     }
 
     /// Return a faillilble iterator of full paths of nodes that have an
     /// `entry` for which the given `predicate` returns true.
     ///
     /// Fallibility means that each iterator item is a `Result`, which may
     /// indicate a parse error of the on-disk dirstate-v2 format. Such errors
     /// should only happen if Mercurial is buggy or a repository is corrupted.
     fn filter_full_paths<'tree>(
         &'tree self,
         predicate: impl Fn(&DirstateEntry) -> bool + 'tree,
     ) -> impl Iterator<Item = Result<&HgPath, DirstateV2ParseError>> + 'tree
     {
         filter_map_results(self.iter_nodes(), move |node| {
             if let Some(entry) = node.entry()? {
                 if predicate(&entry) {
                     return Ok(Some(node.full_path(self.on_disk)?));
                 }
             }
             Ok(None)
         })
     }
 }
 
 /// Like `Iterator::filter_map`, but over a fallible iterator of `Result`s.
 ///
 /// The callback is only called for incoming `Ok` values. Errors are passed
 /// through as-is. In order to let it use the `?` operator the callback is
 /// expected to return a `Result` of `Option`, instead of an `Option` of
 /// `Result`.
 fn filter_map_results<'a, I, F, A, B, E>(
     iter: I,
     f: F,
 ) -> impl Iterator<Item = Result<B, E>> + 'a
 where
     I: Iterator<Item = Result<A, E>> + 'a,
     F: Fn(A) -> Result<Option<B>, E> + 'a,
 {
     iter.filter_map(move |result| match result {
         Ok(node) => f(node).transpose(),
         Err(e) => Some(Err(e)),
     })
 }
 
 impl<'on_disk> super::dispatch::DirstateMapMethods for DirstateMap<'on_disk> {
     fn clear(&mut self) {
         self.root = Default::default();
         self.nodes_with_entry_count = 0;
         self.nodes_with_copy_source_count = 0;
     }
 
     fn add_file(
         &mut self,
         filename: &HgPath,
         entry: DirstateEntry,
         added: bool,
         merged: bool,
         from_p2: bool,
         possibly_dirty: bool,
     ) -> Result<(), DirstateError> {
         let mut entry = entry;
         if added {
             assert!(!possibly_dirty);
             assert!(!from_p2);
             entry.state = EntryState::Added;
             entry.size = SIZE_NON_NORMAL;
             entry.mtime = MTIME_UNSET;
         } else if merged {
             assert!(!possibly_dirty);
             assert!(!from_p2);
             entry.state = EntryState::Merged;
             entry.size = SIZE_FROM_OTHER_PARENT;
             entry.mtime = MTIME_UNSET;
         } else if from_p2 {
             assert!(!possibly_dirty);
             entry.size = SIZE_FROM_OTHER_PARENT;
             entry.mtime = MTIME_UNSET;
         } else if possibly_dirty {
+            entry.state = EntryState::Normal;
             entry.size = SIZE_NON_NORMAL;
             entry.mtime = MTIME_UNSET;
         } else {
             entry.size = entry.size & V1_RANGEMASK;
             entry.mtime = entry.mtime & V1_RANGEMASK;
         }
 
         let old_state = match self.get(filename)? {
             Some(e) => e.state,
             None => EntryState::Unknown,
         };
 
         Ok(self.add_or_remove_file(filename, old_state, entry)?)
     }
 
     fn remove_file(
         &mut self,
         filename: &HgPath,
         in_merge: bool,
     ) -> Result<(), DirstateError> {
         let old_entry_opt = self.get(filename)?;
         let old_state = match old_entry_opt {
             Some(e) => e.state,
             None => EntryState::Unknown,
         };
         let mut size = 0;
         if in_merge {
             // XXX we should not be able to have 'm' state and 'FROM_P2' if not
             // during a merge. So I (marmoute) am not sure we need the
             // conditionnal at all. Adding double checking this with assert
             // would be nice.
             if let Some(old_entry) = old_entry_opt {
                 // backup the previous state
                 if old_entry.state == EntryState::Merged {
                     size = SIZE_NON_NORMAL;
                 } else if old_entry.state == EntryState::Normal
                     && old_entry.size == SIZE_FROM_OTHER_PARENT
                 {
                     // other parent
                     size = SIZE_FROM_OTHER_PARENT;
                 }
             }
         }
         if size == 0 {
             self.copy_map_remove(filename)?;
         }
         let entry = DirstateEntry {
             state: EntryState::Removed,
             mode: 0,
             size,
             mtime: 0,
         };
         Ok(self.add_or_remove_file(filename, old_state, entry)?)
     }
 
     fn drop_file(
         &mut self,
         filename: &HgPath,
         old_state: EntryState,
     ) -> Result<bool, DirstateError> {
         struct Dropped {
             was_tracked: bool,
             had_entry: bool,
             had_copy_source: bool,
         }
 
         /// If this returns `Ok(Some((dropped, removed)))`, then
         ///
         /// * `dropped` is about the leaf node that was at `filename`
         /// * `removed` is whether this particular level of recursion just
         ///   removed a node in `nodes`.
         fn recur<'on_disk>(
             on_disk: &'on_disk [u8],
             nodes: &mut ChildNodes<'on_disk>,
             path: &HgPath,
         ) -> Result<Option<(Dropped, bool)>, DirstateV2ParseError> {
             let (first_path_component, rest_of_path) =
                 path.split_first_component();
             let node = if let Some(node) =
                 nodes.make_mut(on_disk)?.get_mut(first_path_component)
             {
                 node
             } else {
                 return Ok(None);
             };
             let dropped;
             if let Some(rest) = rest_of_path {
                 if let Some((d, removed)) =
                     recur(on_disk, &mut node.children, rest)?
                 {
                     dropped = d;
                     if dropped.had_entry {
                         node.descendants_with_entry_count -= 1;
                     }
                     if dropped.was_tracked {
                         node.tracked_descendants_count -= 1;
                     }
 
                     // Directory caches must be invalidated when removing a
                     // child node
                     if removed {
                         if let NodeData::CachedDirectory { .. } = &node.data {
                             node.data = NodeData::None
                         }
                     }
                 } else {
                     return Ok(None);
                 }
             } else {
                 let had_entry = node.data.has_entry();
                 if had_entry {
                     node.data = NodeData::None
                 }
                 dropped = Dropped {
                     was_tracked: node
                         .data
                         .as_entry()
                         .map_or(false, |entry| entry.state.is_tracked()),
                     had_entry,
                     had_copy_source: node.copy_source.take().is_some(),
                 };
             }
             // After recursion, for both leaf (rest_of_path is None) nodes and
             // parent nodes, remove a node if it just became empty.
             let remove = !node.data.has_entry()
                 && node.copy_source.is_none()
                 && node.children.is_empty();
             if remove {
                 nodes.make_mut(on_disk)?.remove(first_path_component);
             }
             Ok(Some((dropped, remove)))
         }
 
         if let Some((dropped, _removed)) =
             recur(self.on_disk, &mut self.root, filename)?
         {
             if dropped.had_entry {
                 self.nodes_with_entry_count -= 1
             }
             if dropped.had_copy_source {
                 self.nodes_with_copy_source_count -= 1
             }
             Ok(dropped.had_entry)
         } else {
             debug_assert!(!old_state.is_tracked());
             Ok(false)
         }
     }
 
     fn clear_ambiguous_times(
         &mut self,
         filenames: Vec<HgPathBuf>,
         now: i32,
     ) -> Result<(), DirstateV2ParseError> {
         for filename in filenames {
             if let Some(node) =
                 Self::get_node_mut(self.on_disk, &mut self.root, &filename)?
             {
                 if let NodeData::Entry(entry) = &mut node.data {
                     entry.clear_ambiguous_mtime(now);
                 }
             }
         }
         Ok(())
     }
 
     fn non_normal_entries_contains(
         &mut self,
         key: &HgPath,
     ) -> Result<bool, DirstateV2ParseError> {
         Ok(if let Some(node) = self.get_node(key)? {
             node.entry()?.map_or(false, |entry| entry.is_non_normal())
         } else {
             false
         })
     }
 
     fn non_normal_entries_remove(&mut self, _key: &HgPath) {
         // Do nothing, this `DirstateMap` does not have a separate "non normal
         // entries" set that need to be kept up to date
     }
 
     fn non_normal_or_other_parent_paths(
         &mut self,
     ) -> Box<dyn Iterator<Item = Result<&HgPath, DirstateV2ParseError>> + '_>
     {
         Box::new(self.filter_full_paths(|entry| {
             entry.is_non_normal() || entry.is_from_other_parent()
         }))
     }
 
     fn set_non_normal_other_parent_entries(&mut self, _force: bool) {
         // Do nothing, this `DirstateMap` does not have a separate "non normal
         // entries" and "from other parent" sets that need to be recomputed
     }
 
     fn iter_non_normal_paths(
         &mut self,
     ) -> Box<
         dyn Iterator<Item = Result<&HgPath, DirstateV2ParseError>> + Send + '_,
     > {
         self.iter_non_normal_paths_panic()
     }
 
     fn iter_non_normal_paths_panic(
         &self,
     ) -> Box<
         dyn Iterator<Item = Result<&HgPath, DirstateV2ParseError>> + Send + '_,
     > {
         Box::new(self.filter_full_paths(|entry| entry.is_non_normal()))
     }
 
     fn iter_other_parent_paths(
         &mut self,
     ) -> Box<
         dyn Iterator<Item = Result<&HgPath, DirstateV2ParseError>> + Send + '_,
     > {
         Box::new(self.filter_full_paths(|entry| entry.is_from_other_parent()))
     }
 
     fn has_tracked_dir(
         &mut self,
         directory: &HgPath,
     ) -> Result<bool, DirstateError> {
         if let Some(node) = self.get_node(directory)? {
             // A node without a `DirstateEntry` was created to hold child
             // nodes, and is therefore a directory.
             let state = node.state()?;
             Ok(state.is_none() && node.tracked_descendants_count() > 0)
         } else {
             Ok(false)
         }
     }
 
     fn has_dir(&mut self, directory: &HgPath) -> Result<bool, DirstateError> {
         if let Some(node) = self.get_node(directory)? {
             // A node without a `DirstateEntry` was created to hold child
             // nodes, and is therefore a directory.
             let state = node.state()?;
             Ok(state.is_none() && node.descendants_with_entry_count() > 0)
         } else {
             Ok(false)
         }
     }
 
     #[timed]
     fn pack_v1(
         &mut self,
         parents: DirstateParents,
         now: Timestamp,
     ) -> Result<Vec<u8>, DirstateError> {
         let now: i32 = now.0.try_into().expect("time overflow");
         let mut ambiguous_mtimes = Vec::new();
         // Optizimation (to be measured?): pre-compute size to avoid `Vec`
         // reallocations
         let mut size = parents.as_bytes().len();
         for node in self.iter_nodes() {
             let node = node?;
             if let Some(entry) = node.entry()? {
                 size += packed_entry_size(
                     node.full_path(self.on_disk)?,
                     node.copy_source(self.on_disk)?,
                 );
                 if entry.mtime_is_ambiguous(now) {
                     ambiguous_mtimes.push(
                         node.full_path_borrowed(self.on_disk)?
                             .detach_from_tree(),
                     )
                 }
             }
         }
         self.clear_known_ambiguous_mtimes(&ambiguous_mtimes)?;
 
         let mut packed = Vec::with_capacity(size);
         packed.extend(parents.as_bytes());
 
         for node in self.iter_nodes() {
             let node = node?;
             if let Some(entry) = node.entry()? {
                 pack_entry(
                     node.full_path(self.on_disk)?,
                     &entry,
                     node.copy_source(self.on_disk)?,
                     &mut packed,
                 );
             }
         }
         Ok(packed)
     }
 
     #[timed]
     fn pack_v2(
         &mut self,
         parents: DirstateParents,
         now: Timestamp,
     ) -> Result<Vec<u8>, DirstateError> {
         // TODO: how do we want to handle this in 2038?
         let now: i32 = now.0.try_into().expect("time overflow");
         let mut paths = Vec::new();
         for node in self.iter_nodes() {
             let node = node?;
             if let Some(entry) = node.entry()? {
                 if entry.mtime_is_ambiguous(now) {
                     paths.push(
                         node.full_path_borrowed(self.on_disk)?
                             .detach_from_tree(),
                     )
                 }
             }
         }
         // Borrow of `self` ends here since we collect cloned paths
 
         self.clear_known_ambiguous_mtimes(&paths)?;
 
         on_disk::write(self, parents)
     }
 
     fn status<'a>(
         &'a mut self,
         matcher: &'a (dyn Matcher + Sync),
         root_dir: PathBuf,
         ignore_files: Vec<PathBuf>,
         options: StatusOptions,
     ) -> Result<(DirstateStatus<'a>, Vec<PatternFileWarning>), StatusError>
     {
         super::status::status(self, matcher, root_dir, ignore_files, options)
     }
 
     fn copy_map_len(&self) -> usize {
         self.nodes_with_copy_source_count as usize
     }
 
     fn copy_map_iter(&self) -> CopyMapIter<'_> {
         Box::new(filter_map_results(self.iter_nodes(), move |node| {
             Ok(if let Some(source) = node.copy_source(self.on_disk)? {
                 Some((node.full_path(self.on_disk)?, source))
             } else {
                 None
             })
         }))
     }
 
     fn copy_map_contains_key(
         &self,
         key: &HgPath,
     ) -> Result<bool, DirstateV2ParseError> {
         Ok(if let Some(node) = self.get_node(key)? {
             node.has_copy_source()
         } else {
             false
         })
     }
 
     fn copy_map_get(
         &self,
         key: &HgPath,
     ) -> Result<Option<&HgPath>, DirstateV2ParseError> {
         if let Some(node) = self.get_node(key)? {
             if let Some(source) = node.copy_source(self.on_disk)? {
                 return Ok(Some(source));
             }
         }
         Ok(None)
     }
 
     fn copy_map_remove(
         &mut self,
         key: &HgPath,
     ) -> Result<Option<HgPathBuf>, DirstateV2ParseError> {
         let count = &mut self.nodes_with_copy_source_count;
         Ok(
             Self::get_node_mut(self.on_disk, &mut self.root, key)?.and_then(
                 |node| {
                     if node.copy_source.is_some() {
                         *count -= 1
                     }
                     node.copy_source.take().map(Cow::into_owned)
                 },
             ),
         )
     }
 
     fn copy_map_insert(
         &mut self,
         key: HgPathBuf,
         value: HgPathBuf,
     ) -> Result<Option<HgPathBuf>, DirstateV2ParseError> {
         let node = Self::get_or_insert_node(
             self.on_disk,
             &mut self.root,
             &key,
             WithBasename::to_cow_owned,
             |_ancestor| {},
         )?;
         if node.copy_source.is_none() {
             self.nodes_with_copy_source_count += 1
         }
         Ok(node.copy_source.replace(value.into()).map(Cow::into_owned))
     }
 
     fn len(&self) -> usize {
         self.nodes_with_entry_count as usize
     }
 
     fn contains_key(
         &self,
         key: &HgPath,
     ) -> Result<bool, DirstateV2ParseError> {
         Ok(self.get(key)?.is_some())
     }
 
     fn get(
         &self,
         key: &HgPath,
     ) -> Result<Option<DirstateEntry>, DirstateV2ParseError> {
         Ok(if let Some(node) = self.get_node(key)? {
             node.entry()?
         } else {
             None
         })
     }
 
     fn iter(&self) -> StateMapIter<'_> {
         Box::new(filter_map_results(self.iter_nodes(), move |node| {
             Ok(if let Some(entry) = node.entry()? {
                 Some((node.full_path(self.on_disk)?, entry))
             } else {
                 None
             })
         }))
     }
 
     fn iter_directories(
         &self,
     ) -> Box<
         dyn Iterator<
                 Item = Result<
                     (&HgPath, Option<Timestamp>),
                     DirstateV2ParseError,
                 >,
             > + Send
             + '_,
     > {
         Box::new(filter_map_results(self.iter_nodes(), move |node| {
             Ok(if node.state()?.is_none() {
                 Some((
                     node.full_path(self.on_disk)?,
                     node.cached_directory_mtime()
                         .map(|mtime| Timestamp(mtime.seconds())),
                 ))
             } else {
                 None
             })
         }))
     }
 }