dirstate: use keyword arguments to clarify status()'s callers

The arguments are especially non-obvious because the order is
different from dirstate.walk().

Differential Revision: https://phab.mercurial-scm.org/D847

hgext/largefiles/lfutil.py

@@ -1,673 +1,674 b''
 # Copyright 2009-2010 Gregory P. Ward
 # Copyright 2009-2010 Intelerad Medical Systems Incorporated
 # Copyright 2010-2011 Fog Creek Software
 # Copyright 2010-2011 Unity Technologies
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 '''largefiles utility code: must not import other modules in this package.'''
 from __future__ import absolute_import
 
 import copy
 import hashlib
 import os
 import platform
 import stat
 
 from mercurial.i18n import _
 
 from mercurial import (
     dirstate,
     encoding,
     error,
     httpconnection,
     match as matchmod,
     node,
     pycompat,
     scmutil,
     sparse,
     util,
     vfs as vfsmod,
 )
 
 shortname = '.hglf'
 shortnameslash = shortname + '/'
 longname = 'largefiles'
 
 # -- Private worker functions ------------------------------------------
 
 def getminsize(ui, assumelfiles, opt, default=10):
     lfsize = opt
     if not lfsize and assumelfiles:
         lfsize = ui.config(longname, 'minsize', default=default)
     if lfsize:
         try:
             lfsize = float(lfsize)
         except ValueError:
             raise error.Abort(_('largefiles: size must be number (not %s)\n')
                              % lfsize)
     if lfsize is None:
         raise error.Abort(_('minimum size for largefiles must be specified'))
     return lfsize
 
 def link(src, dest):
     """Try to create hardlink - if that fails, efficiently make a copy."""
     util.makedirs(os.path.dirname(dest))
     try:
         util.oslink(src, dest)
     except OSError:
         # if hardlinks fail, fallback on atomic copy
         with open(src, 'rb') as srcf, util.atomictempfile(dest) as dstf:
             for chunk in util.filechunkiter(srcf):
                 dstf.write(chunk)
         os.chmod(dest, os.stat(src).st_mode)
 
 def usercachepath(ui, hash):
     '''Return the correct location in the "global" largefiles cache for a file
     with the given hash.
     This cache is used for sharing of largefiles across repositories - both
     to preserve download bandwidth and storage space.'''
     return os.path.join(_usercachedir(ui), hash)
 
 def _usercachedir(ui):
     '''Return the location of the "global" largefiles cache.'''
     path = ui.configpath(longname, 'usercache', None)
     if path:
         return path
     if pycompat.osname == 'nt':
         appdata = encoding.environ.get('LOCALAPPDATA',\
                         encoding.environ.get('APPDATA'))
         if appdata:
             return os.path.join(appdata, longname)
     elif platform.system() == 'Darwin':
         home = encoding.environ.get('HOME')
         if home:
             return os.path.join(home, 'Library', 'Caches', longname)
     elif pycompat.osname == 'posix':
         path = encoding.environ.get('XDG_CACHE_HOME')
         if path:
             return os.path.join(path, longname)
         home = encoding.environ.get('HOME')
         if home:
             return os.path.join(home, '.cache', longname)
     else:
         raise error.Abort(_('unknown operating system: %s\n')
                           % pycompat.osname)
     raise error.Abort(_('unknown %s usercache location') % longname)
 
 def inusercache(ui, hash):
     path = usercachepath(ui, hash)
     return os.path.exists(path)
 
 def findfile(repo, hash):
     '''Return store path of the largefile with the specified hash.
     As a side effect, the file might be linked from user cache.
     Return None if the file can't be found locally.'''
     path, exists = findstorepath(repo, hash)
     if exists:
         repo.ui.note(_('found %s in store\n') % hash)
         return path
     elif inusercache(repo.ui, hash):
         repo.ui.note(_('found %s in system cache\n') % hash)
         path = storepath(repo, hash)
         link(usercachepath(repo.ui, hash), path)
         return path
     return None
 
 class largefilesdirstate(dirstate.dirstate):
     def __getitem__(self, key):
         return super(largefilesdirstate, self).__getitem__(unixpath(key))
     def normal(self, f):
         return super(largefilesdirstate, self).normal(unixpath(f))
     def remove(self, f):
         return super(largefilesdirstate, self).remove(unixpath(f))
     def add(self, f):
         return super(largefilesdirstate, self).add(unixpath(f))
     def drop(self, f):
         return super(largefilesdirstate, self).drop(unixpath(f))
     def forget(self, f):
         return super(largefilesdirstate, self).forget(unixpath(f))
     def normallookup(self, f):
         return super(largefilesdirstate, self).normallookup(unixpath(f))
     def _ignore(self, f):
         return False
     def write(self, tr=False):
         # (1) disable PENDING mode always
         #     (lfdirstate isn't yet managed as a part of the transaction)
         # (2) avoid develwarn 'use dirstate.write with ....'
         super(largefilesdirstate, self).write(None)
 
 def openlfdirstate(ui, repo, create=True):
     '''
     Return a dirstate object that tracks largefiles: i.e. its root is
     the repo root, but it is saved in .hg/largefiles/dirstate.
     '''
     vfs = repo.vfs
     lfstoredir = longname
     opener = vfsmod.vfs(vfs.join(lfstoredir))
     lfdirstate = largefilesdirstate(opener, ui, repo.root,
                                     repo.dirstate._validate,
                                     lambda: sparse.matcher(repo))
 
     # If the largefiles dirstate does not exist, populate and create
     # it. This ensures that we create it on the first meaningful
     # largefiles operation in a new clone.
     if create and not vfs.exists(vfs.join(lfstoredir, 'dirstate')):
         matcher = getstandinmatcher(repo)
         standins = repo.dirstate.walk(matcher, subrepos=[], unknown=False,
                                       ignored=False)
 
         if len(standins) > 0:
             vfs.makedirs(lfstoredir)
 
         for standin in standins:
             lfile = splitstandin(standin)
             lfdirstate.normallookup(lfile)
     return lfdirstate
 
 def lfdirstatestatus(lfdirstate, repo):
     pctx = repo['.']
     match = matchmod.always(repo.root, repo.getcwd())
-    unsure, s = lfdirstate.status(match, [], False, False, False)
+    unsure, s = lfdirstate.status(match, subrepos=[], ignored=False,
+                                  clean=False, unknown=False)
     modified, clean = s.modified, s.clean
     for lfile in unsure:
         try:
             fctx = pctx[standin(lfile)]
         except LookupError:
             fctx = None
         if not fctx or readasstandin(fctx) != hashfile(repo.wjoin(lfile)):
             modified.append(lfile)
         else:
             clean.append(lfile)
             lfdirstate.normal(lfile)
     return s
 
 def listlfiles(repo, rev=None, matcher=None):
     '''return a list of largefiles in the working copy or the
     specified changeset'''
 
     if matcher is None:
         matcher = getstandinmatcher(repo)
 
     # ignore unknown files in working directory
     return [splitstandin(f)
             for f in repo[rev].walk(matcher)
             if rev is not None or repo.dirstate[f] != '?']
 
 def instore(repo, hash, forcelocal=False):
     '''Return true if a largefile with the given hash exists in the store'''
     return os.path.exists(storepath(repo, hash, forcelocal))
 
 def storepath(repo, hash, forcelocal=False):
     '''Return the correct location in the repository largefiles store for a
     file with the given hash.'''
     if not forcelocal and repo.shared():
         return repo.vfs.reljoin(repo.sharedpath, longname, hash)
     return repo.vfs.join(longname, hash)
 
 def findstorepath(repo, hash):
     '''Search through the local store path(s) to find the file for the given
     hash.  If the file is not found, its path in the primary store is returned.
     The return value is a tuple of (path, exists(path)).
     '''
     # For shared repos, the primary store is in the share source.  But for
     # backward compatibility, force a lookup in the local store if it wasn't
     # found in the share source.
     path = storepath(repo, hash, False)
 
     if instore(repo, hash):
         return (path, True)
     elif repo.shared() and instore(repo, hash, True):
         return storepath(repo, hash, True), True
 
     return (path, False)
 
 def copyfromcache(repo, hash, filename):
     '''Copy the specified largefile from the repo or system cache to
     filename in the repository. Return true on success or false if the
     file was not found in either cache (which should not happened:
     this is meant to be called only after ensuring that the needed
     largefile exists in the cache).'''
     wvfs = repo.wvfs
     path = findfile(repo, hash)
     if path is None:
         return False
     wvfs.makedirs(wvfs.dirname(wvfs.join(filename)))
     # The write may fail before the file is fully written, but we
     # don't use atomic writes in the working copy.
     with open(path, 'rb') as srcfd, wvfs(filename, 'wb') as destfd:
         gothash = copyandhash(
             util.filechunkiter(srcfd), destfd)
     if gothash != hash:
         repo.ui.warn(_('%s: data corruption in %s with hash %s\n')
                      % (filename, path, gothash))
         wvfs.unlink(filename)
         return False
     return True
 
 def copytostore(repo, ctx, file, fstandin):
     wvfs = repo.wvfs
     hash = readasstandin(ctx[fstandin])
     if instore(repo, hash):
         return
     if wvfs.exists(file):
         copytostoreabsolute(repo, wvfs.join(file), hash)
     else:
         repo.ui.warn(_("%s: largefile %s not available from local store\n") %
                      (file, hash))
 
 def copyalltostore(repo, node):
     '''Copy all largefiles in a given revision to the store'''
 
     ctx = repo[node]
     for filename in ctx.files():
         realfile = splitstandin(filename)
         if realfile is not None and filename in ctx.manifest():
             copytostore(repo, ctx, realfile, filename)
 
 def copytostoreabsolute(repo, file, hash):
     if inusercache(repo.ui, hash):
         link(usercachepath(repo.ui, hash), storepath(repo, hash))
     else:
         util.makedirs(os.path.dirname(storepath(repo, hash)))
         with open(file, 'rb') as srcf:
             with util.atomictempfile(storepath(repo, hash),
                                      createmode=repo.store.createmode) as dstf:
                 for chunk in util.filechunkiter(srcf):
                     dstf.write(chunk)
         linktousercache(repo, hash)
 
 def linktousercache(repo, hash):
     '''Link / copy the largefile with the specified hash from the store
     to the cache.'''
     path = usercachepath(repo.ui, hash)
     link(storepath(repo, hash), path)
 
 def getstandinmatcher(repo, rmatcher=None):
     '''Return a match object that applies rmatcher to the standin directory'''
     wvfs = repo.wvfs
     standindir = shortname
 
     # no warnings about missing files or directories
     badfn = lambda f, msg: None
 
     if rmatcher and not rmatcher.always():
         pats = [wvfs.join(standindir, pat) for pat in rmatcher.files()]
         if not pats:
             pats = [wvfs.join(standindir)]
         match = scmutil.match(repo[None], pats, badfn=badfn)
     else:
         # no patterns: relative to repo root
         match = scmutil.match(repo[None], [wvfs.join(standindir)], badfn=badfn)
     return match
 
 def composestandinmatcher(repo, rmatcher):
     '''Return a matcher that accepts standins corresponding to the
     files accepted by rmatcher. Pass the list of files in the matcher
     as the paths specified by the user.'''
     smatcher = getstandinmatcher(repo, rmatcher)
     isstandin = smatcher.matchfn
     def composedmatchfn(f):
         return isstandin(f) and rmatcher.matchfn(splitstandin(f))
     smatcher.matchfn = composedmatchfn
 
     return smatcher
 
 def standin(filename):
     '''Return the repo-relative path to the standin for the specified big
     file.'''
     # Notes:
     # 1) Some callers want an absolute path, but for instance addlargefiles
     #    needs it repo-relative so it can be passed to repo[None].add().  So
     #    leave it up to the caller to use repo.wjoin() to get an absolute path.
     # 2) Join with '/' because that's what dirstate always uses, even on
     #    Windows. Change existing separator to '/' first in case we are
     #    passed filenames from an external source (like the command line).
     return shortnameslash + util.pconvert(filename)
 
 def isstandin(filename):
     '''Return true if filename is a big file standin. filename must be
     in Mercurial's internal form (slash-separated).'''
     return filename.startswith(shortnameslash)
 
 def splitstandin(filename):
     # Split on / because that's what dirstate always uses, even on Windows.
     # Change local separator to / first just in case we are passed filenames
     # from an external source (like the command line).
     bits = util.pconvert(filename).split('/', 1)
     if len(bits) == 2 and bits[0] == shortname:
         return bits[1]
     else:
         return None
 
 def updatestandin(repo, lfile, standin):
     """Re-calculate hash value of lfile and write it into standin
 
     This assumes that "lfutil.standin(lfile) == standin", for efficiency.
     """
     file = repo.wjoin(lfile)
     if repo.wvfs.exists(lfile):
         hash = hashfile(file)
         executable = getexecutable(file)
         writestandin(repo, standin, hash, executable)
     else:
         raise error.Abort(_('%s: file not found!') % lfile)
 
 def readasstandin(fctx):
     '''read hex hash from given filectx of standin file
 
     This encapsulates how "standin" data is stored into storage layer.'''
     return fctx.data().strip()
 
 def writestandin(repo, standin, hash, executable):
     '''write hash to <repo.root>/<standin>'''
     repo.wwrite(standin, hash + '\n', executable and 'x' or '')
 
 def copyandhash(instream, outfile):
     '''Read bytes from instream (iterable) and write them to outfile,
     computing the SHA-1 hash of the data along the way. Return the hash.'''
     hasher = hashlib.sha1('')
     for data in instream:
         hasher.update(data)
         outfile.write(data)
     return hasher.hexdigest()
 
 def hashfile(file):
     if not os.path.exists(file):
         return ''
     with open(file, 'rb') as fd:
         return hexsha1(fd)
 
 def getexecutable(filename):
     mode = os.stat(filename).st_mode
     return ((mode & stat.S_IXUSR) and
             (mode & stat.S_IXGRP) and
             (mode & stat.S_IXOTH))
 
 def urljoin(first, second, *arg):
     def join(left, right):
         if not left.endswith('/'):
             left += '/'
         if right.startswith('/'):
             right = right[1:]
         return left + right
 
     url = join(first, second)
     for a in arg:
         url = join(url, a)
     return url
 
 def hexsha1(fileobj):
     """hexsha1 returns the hex-encoded sha1 sum of the data in the file-like
     object data"""
     h = hashlib.sha1()
     for chunk in util.filechunkiter(fileobj):
         h.update(chunk)
     return h.hexdigest()
 
 def httpsendfile(ui, filename):
     return httpconnection.httpsendfile(ui, filename, 'rb')
 
 def unixpath(path):
     '''Return a version of path normalized for use with the lfdirstate.'''
     return util.pconvert(os.path.normpath(path))
 
 def islfilesrepo(repo):
     '''Return true if the repo is a largefile repo.'''
     if ('largefiles' in repo.requirements and
             any(shortnameslash in f[0] for f in repo.store.datafiles())):
         return True
 
     return any(openlfdirstate(repo.ui, repo, False))
 
 class storeprotonotcapable(Exception):
     def __init__(self, storetypes):
         self.storetypes = storetypes
 
 def getstandinsstate(repo):
     standins = []
     matcher = getstandinmatcher(repo)
     wctx = repo[None]
     for standin in repo.dirstate.walk(matcher, subrepos=[], unknown=False,
                                       ignored=False):
         lfile = splitstandin(standin)
         try:
             hash = readasstandin(wctx[standin])
         except IOError:
             hash = None
         standins.append((lfile, hash))
     return standins
 
 def synclfdirstate(repo, lfdirstate, lfile, normallookup):
     lfstandin = standin(lfile)
     if lfstandin in repo.dirstate:
         stat = repo.dirstate._map[lfstandin]
         state, mtime = stat[0], stat[3]
     else:
         state, mtime = '?', -1
     if state == 'n':
         if (normallookup or mtime < 0 or
             not repo.wvfs.exists(lfile)):
             # state 'n' doesn't ensure 'clean' in this case
             lfdirstate.normallookup(lfile)
         else:
             lfdirstate.normal(lfile)
     elif state == 'm':
         lfdirstate.normallookup(lfile)
     elif state == 'r':
         lfdirstate.remove(lfile)
     elif state == 'a':
         lfdirstate.add(lfile)
     elif state == '?':
         lfdirstate.drop(lfile)
 
 def markcommitted(orig, ctx, node):
     repo = ctx.repo()
 
     orig(node)
 
     # ATTENTION: "ctx.files()" may differ from "repo[node].files()"
     # because files coming from the 2nd parent are omitted in the latter.
     #
     # The former should be used to get targets of "synclfdirstate",
     # because such files:
     # - are marked as "a" by "patch.patch()" (e.g. via transplant), and
     # - have to be marked as "n" after commit, but
     # - aren't listed in "repo[node].files()"
 
     lfdirstate = openlfdirstate(repo.ui, repo)
     for f in ctx.files():
         lfile = splitstandin(f)
         if lfile is not None:
             synclfdirstate(repo, lfdirstate, lfile, False)
     lfdirstate.write()
 
     # As part of committing, copy all of the largefiles into the cache.
     #
     # Using "node" instead of "ctx" implies additional "repo[node]"
     # lookup while copyalltostore(), but can omit redundant check for
     # files comming from the 2nd parent, which should exist in store
     # at merging.
     copyalltostore(repo, node)
 
 def getlfilestoupdate(oldstandins, newstandins):
     changedstandins = set(oldstandins).symmetric_difference(set(newstandins))
     filelist = []
     for f in changedstandins:
         if f[0] not in filelist:
             filelist.append(f[0])
     return filelist
 
 def getlfilestoupload(repo, missing, addfunc):
     for i, n in enumerate(missing):
         repo.ui.progress(_('finding outgoing largefiles'), i,
             unit=_('revisions'), total=len(missing))
         parents = [p for p in repo[n].parents() if p != node.nullid]
 
         oldlfstatus = repo.lfstatus
         repo.lfstatus = False
         try:
             ctx = repo[n]
         finally:
             repo.lfstatus = oldlfstatus
 
         files = set(ctx.files())
         if len(parents) == 2:
             mc = ctx.manifest()
             mp1 = ctx.parents()[0].manifest()
             mp2 = ctx.parents()[1].manifest()
             for f in mp1:
                 if f not in mc:
                     files.add(f)
             for f in mp2:
                 if f not in mc:
                     files.add(f)
             for f in mc:
                 if mc[f] != mp1.get(f, None) or mc[f] != mp2.get(f, None):
                     files.add(f)
         for fn in files:
             if isstandin(fn) and fn in ctx:
                 addfunc(fn, readasstandin(ctx[fn]))
     repo.ui.progress(_('finding outgoing largefiles'), None)
 
 def updatestandinsbymatch(repo, match):
     '''Update standins in the working directory according to specified match
 
     This returns (possibly modified) ``match`` object to be used for
     subsequent commit process.
     '''
 
     ui = repo.ui
 
     # Case 1: user calls commit with no specific files or
     # include/exclude patterns: refresh and commit all files that
     # are "dirty".
     if match is None or match.always():
         # Spend a bit of time here to get a list of files we know
         # are modified so we can compare only against those.
         # It can cost a lot of time (several seconds)
         # otherwise to update all standins if the largefiles are
         # large.
         lfdirstate = openlfdirstate(ui, repo)
         dirtymatch = matchmod.always(repo.root, repo.getcwd())
-        unsure, s = lfdirstate.status(dirtymatch, [], False, False,
-                                      False)
+        unsure, s = lfdirstate.status(dirtymatch, subrepos=[], ignored=False,
+                                      clean=False, unknown=False)
         modifiedfiles = unsure + s.modified + s.added + s.removed
         lfiles = listlfiles(repo)
         # this only loops through largefiles that exist (not
         # removed/renamed)
         for lfile in lfiles:
             if lfile in modifiedfiles:
                 fstandin = standin(lfile)
                 if repo.wvfs.exists(fstandin):
                     # this handles the case where a rebase is being
                     # performed and the working copy is not updated
                     # yet.
                     if repo.wvfs.exists(lfile):
                         updatestandin(repo, lfile, fstandin)
 
         return match
 
     lfiles = listlfiles(repo)
     match._files = repo._subdirlfs(match.files(), lfiles)
 
     # Case 2: user calls commit with specified patterns: refresh
     # any matching big files.
     smatcher = composestandinmatcher(repo, match)
     standins = repo.dirstate.walk(smatcher, subrepos=[], unknown=False,
                                   ignored=False)
 
     # No matching big files: get out of the way and pass control to
     # the usual commit() method.
     if not standins:
         return match
 
     # Refresh all matching big files.  It's possible that the
     # commit will end up failing, in which case the big files will
     # stay refreshed.  No harm done: the user modified them and
     # asked to commit them, so sooner or later we're going to
     # refresh the standins.  Might as well leave them refreshed.
     lfdirstate = openlfdirstate(ui, repo)
     for fstandin in standins:
         lfile = splitstandin(fstandin)
         if lfdirstate[lfile] != 'r':
             updatestandin(repo, lfile, fstandin)
 
     # Cook up a new matcher that only matches regular files or
     # standins corresponding to the big files requested by the
     # user.  Have to modify _files to prevent commit() from
     # complaining "not tracked" for big files.
     match = copy.copy(match)
     origmatchfn = match.matchfn
 
     # Check both the list of largefiles and the list of
     # standins because if a largefile was removed, it
     # won't be in the list of largefiles at this point
     match._files += sorted(standins)
 
     actualfiles = []
     for f in match._files:
         fstandin = standin(f)
 
         # For largefiles, only one of the normal and standin should be
         # committed (except if one of them is a remove).  In the case of a
         # standin removal, drop the normal file if it is unknown to dirstate.
         # Thus, skip plain largefile names but keep the standin.
         if f in lfiles or fstandin in standins:
             if repo.dirstate[fstandin] != 'r':
                 if repo.dirstate[f] != 'r':
                     continue
             elif repo.dirstate[f] == '?':
                 continue
 
         actualfiles.append(f)
     match._files = actualfiles
 
     def matchfn(f):
         if origmatchfn(f):
             return f not in lfiles
         else:
             return f in standins
 
     match.matchfn = matchfn
 
     return match
 
 class automatedcommithook(object):
     '''Stateful hook to update standins at the 1st commit of resuming
 
     For efficiency, updating standins in the working directory should
     be avoided while automated committing (like rebase, transplant and
     so on), because they should be updated before committing.
 
     But the 1st commit of resuming automated committing (e.g. ``rebase
     --continue``) should update them, because largefiles may be
     modified manually.
     '''
     def __init__(self, resuming):
         self.resuming = resuming
 
     def __call__(self, repo, match):
         if self.resuming:
             self.resuming = False # avoids updating at subsequent commits
             return updatestandinsbymatch(repo, match)
         else:
             return match
 
 def getstatuswriter(ui, repo, forcibly=None):
     '''Return the function to write largefiles specific status out
 
     If ``forcibly`` is ``None``, this returns the last element of
     ``repo._lfstatuswriters`` as "default" writer function.
 
     Otherwise, this returns the function to always write out (or
     ignore if ``not forcibly``) status.
     '''
     if forcibly is None and util.safehasattr(repo, '_largefilesenabled'):
         return repo._lfstatuswriters[-1]
     else:
         if forcibly:
             return ui.status # forcibly WRITE OUT
         else:
             return lambda *msg, **opts: None # forcibly IGNORE

hgext/largefiles/overrides.py

@@ -1,1470 +1,1472 b''
 # Copyright 2009-2010 Gregory P. Ward
 # Copyright 2009-2010 Intelerad Medical Systems Incorporated
 # Copyright 2010-2011 Fog Creek Software
 # Copyright 2010-2011 Unity Technologies
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 '''Overridden Mercurial commands and functions for the largefiles extension'''
 from __future__ import absolute_import
 
 import copy
 import os
 
 from mercurial.i18n import _
 
 from mercurial import (
     archival,
     cmdutil,
     error,
     hg,
     match as matchmod,
     pathutil,
     registrar,
     scmutil,
     smartset,
     util,
 )
 
 from . import (
     lfcommands,
     lfutil,
     storefactory,
 )
 
 # -- Utility functions: commonly/repeatedly needed functionality ---------------
 
 def composelargefilematcher(match, manifest):
     '''create a matcher that matches only the largefiles in the original
     matcher'''
     m = copy.copy(match)
     lfile = lambda f: lfutil.standin(f) in manifest
     m._files = filter(lfile, m._files)
     m._fileset = set(m._files)
     m.always = lambda: False
     origmatchfn = m.matchfn
     m.matchfn = lambda f: lfile(f) and origmatchfn(f)
     return m
 
 def composenormalfilematcher(match, manifest, exclude=None):
     excluded = set()
     if exclude is not None:
         excluded.update(exclude)
 
     m = copy.copy(match)
     notlfile = lambda f: not (lfutil.isstandin(f) or lfutil.standin(f) in
             manifest or f in excluded)
     m._files = filter(notlfile, m._files)
     m._fileset = set(m._files)
     m.always = lambda: False
     origmatchfn = m.matchfn
     m.matchfn = lambda f: notlfile(f) and origmatchfn(f)
     return m
 
 def installnormalfilesmatchfn(manifest):
     '''installmatchfn with a matchfn that ignores all largefiles'''
     def overridematch(ctx, pats=(), opts=None, globbed=False,
             default='relpath', badfn=None):
         if opts is None:
             opts = {}
         match = oldmatch(ctx, pats, opts, globbed, default, badfn=badfn)
         return composenormalfilematcher(match, manifest)
     oldmatch = installmatchfn(overridematch)
 
 def installmatchfn(f):
     '''monkey patch the scmutil module with a custom match function.
     Warning: it is monkey patching the _module_ on runtime! Not thread safe!'''
     oldmatch = scmutil.match
     setattr(f, 'oldmatch', oldmatch)
     scmutil.match = f
     return oldmatch
 
 def restorematchfn():
     '''restores scmutil.match to what it was before installmatchfn
     was called.  no-op if scmutil.match is its original function.
 
     Note that n calls to installmatchfn will require n calls to
     restore the original matchfn.'''
     scmutil.match = getattr(scmutil.match, 'oldmatch')
 
 def installmatchandpatsfn(f):
     oldmatchandpats = scmutil.matchandpats
     setattr(f, 'oldmatchandpats', oldmatchandpats)
     scmutil.matchandpats = f
     return oldmatchandpats
 
 def restorematchandpatsfn():
     '''restores scmutil.matchandpats to what it was before
     installmatchandpatsfn was called. No-op if scmutil.matchandpats
     is its original function.
 
     Note that n calls to installmatchandpatsfn will require n calls
     to restore the original matchfn.'''
     scmutil.matchandpats = getattr(scmutil.matchandpats, 'oldmatchandpats',
             scmutil.matchandpats)
 
 def addlargefiles(ui, repo, isaddremove, matcher, **opts):
     large = opts.get(r'large')
     lfsize = lfutil.getminsize(
         ui, lfutil.islfilesrepo(repo), opts.get(r'lfsize'))
 
     lfmatcher = None
     if lfutil.islfilesrepo(repo):
         lfpats = ui.configlist(lfutil.longname, 'patterns', default=[])
         if lfpats:
             lfmatcher = matchmod.match(repo.root, '', list(lfpats))
 
     lfnames = []
     m = matcher
 
     wctx = repo[None]
     for f in wctx.walk(matchmod.badmatch(m, lambda x, y: None)):
         exact = m.exact(f)
         lfile = lfutil.standin(f) in wctx
         nfile = f in wctx
         exists = lfile or nfile
 
         # addremove in core gets fancy with the name, add doesn't
         if isaddremove:
             name = m.uipath(f)
         else:
             name = m.rel(f)
 
         # Don't warn the user when they attempt to add a normal tracked file.
         # The normal add code will do that for us.
         if exact and exists:
             if lfile:
                 ui.warn(_('%s already a largefile\n') % name)
             continue
 
         if (exact or not exists) and not lfutil.isstandin(f):
             # In case the file was removed previously, but not committed
             # (issue3507)
             if not repo.wvfs.exists(f):
                 continue
 
             abovemin = (lfsize and
                         repo.wvfs.lstat(f).st_size >= lfsize * 1024 * 1024)
             if large or abovemin or (lfmatcher and lfmatcher(f)):
                 lfnames.append(f)
                 if ui.verbose or not exact:
                     ui.status(_('adding %s as a largefile\n') % name)
 
     bad = []
 
     # Need to lock, otherwise there could be a race condition between
     # when standins are created and added to the repo.
     with repo.wlock():
         if not opts.get('dry_run'):
             standins = []
             lfdirstate = lfutil.openlfdirstate(ui, repo)
             for f in lfnames:
                 standinname = lfutil.standin(f)
                 lfutil.writestandin(repo, standinname, hash='',
                     executable=lfutil.getexecutable(repo.wjoin(f)))
                 standins.append(standinname)
                 if lfdirstate[f] == 'r':
                     lfdirstate.normallookup(f)
                 else:
                     lfdirstate.add(f)
             lfdirstate.write()
             bad += [lfutil.splitstandin(f)
                     for f in repo[None].add(standins)
                     if f in m.files()]
 
         added = [f for f in lfnames if f not in bad]
     return added, bad
 
 def removelargefiles(ui, repo, isaddremove, matcher, **opts):
     after = opts.get('after')
     m = composelargefilematcher(matcher, repo[None].manifest())
     try:
         repo.lfstatus = True
         s = repo.status(match=m, clean=not isaddremove)
     finally:
         repo.lfstatus = False
     manifest = repo[None].manifest()
     modified, added, deleted, clean = [[f for f in list
                                         if lfutil.standin(f) in manifest]
                                        for list in (s.modified, s.added,
                                                     s.deleted, s.clean)]
 
     def warn(files, msg):
         for f in files:
             ui.warn(msg % m.rel(f))
         return int(len(files) > 0)
 
     result = 0
 
     if after:
         remove = deleted
         result = warn(modified + added + clean,
                       _('not removing %s: file still exists\n'))
     else:
         remove = deleted + clean
         result = warn(modified, _('not removing %s: file is modified (use -f'
                                   ' to force removal)\n'))
         result = warn(added, _('not removing %s: file has been marked for add'
                                ' (use forget to undo)\n')) or result
 
     # Need to lock because standin files are deleted then removed from the
     # repository and we could race in-between.
     with repo.wlock():
         lfdirstate = lfutil.openlfdirstate(ui, repo)
         for f in sorted(remove):
             if ui.verbose or not m.exact(f):
                 # addremove in core gets fancy with the name, remove doesn't
                 if isaddremove:
                     name = m.uipath(f)
                 else:
                     name = m.rel(f)
                 ui.status(_('removing %s\n') % name)
 
             if not opts.get('dry_run'):
                 if not after:
                     repo.wvfs.unlinkpath(f, ignoremissing=True)
 
         if opts.get('dry_run'):
             return result
 
         remove = [lfutil.standin(f) for f in remove]
         # If this is being called by addremove, let the original addremove
         # function handle this.
         if not isaddremove:
             for f in remove:
                 repo.wvfs.unlinkpath(f, ignoremissing=True)
         repo[None].forget(remove)
 
         for f in remove:
             lfutil.synclfdirstate(repo, lfdirstate, lfutil.splitstandin(f),
                                   False)
 
         lfdirstate.write()
 
     return result
 
 # For overriding mercurial.hgweb.webcommands so that largefiles will
 # appear at their right place in the manifests.
 def decodepath(orig, path):
     return lfutil.splitstandin(path) or path
 
 # -- Wrappers: modify existing commands --------------------------------
 
 def overrideadd(orig, ui, repo, *pats, **opts):
     if opts.get('normal') and opts.get('large'):
         raise error.Abort(_('--normal cannot be used with --large'))
     return orig(ui, repo, *pats, **opts)
 
 def cmdutiladd(orig, ui, repo, matcher, prefix, explicitonly, **opts):
     # The --normal flag short circuits this override
     if opts.get(r'normal'):
         return orig(ui, repo, matcher, prefix, explicitonly, **opts)
 
     ladded, lbad = addlargefiles(ui, repo, False, matcher, **opts)
     normalmatcher = composenormalfilematcher(matcher, repo[None].manifest(),
                                              ladded)
     bad = orig(ui, repo, normalmatcher, prefix, explicitonly, **opts)
 
     bad.extend(f for f in lbad)
     return bad
 
 def cmdutilremove(orig, ui, repo, matcher, prefix, after, force, subrepos):
     normalmatcher = composenormalfilematcher(matcher, repo[None].manifest())
     result = orig(ui, repo, normalmatcher, prefix, after, force, subrepos)
     return removelargefiles(ui, repo, False, matcher, after=after,
                             force=force) or result
 
 def overridestatusfn(orig, repo, rev2, **opts):
     try:
         repo._repo.lfstatus = True
         return orig(repo, rev2, **opts)
     finally:
         repo._repo.lfstatus = False
 
 def overridestatus(orig, ui, repo, *pats, **opts):
     try:
         repo.lfstatus = True
         return orig(ui, repo, *pats, **opts)
     finally:
         repo.lfstatus = False
 
 def overridedirty(orig, repo, ignoreupdate=False, missing=False):
     try:
         repo._repo.lfstatus = True
         return orig(repo, ignoreupdate=ignoreupdate, missing=missing)
     finally:
         repo._repo.lfstatus = False
 
 def overridelog(orig, ui, repo, *pats, **opts):
     def overridematchandpats(ctx, pats=(), opts=None, globbed=False,
             default='relpath', badfn=None):
         """Matcher that merges root directory with .hglf, suitable for log.
         It is still possible to match .hglf directly.
         For any listed files run log on the standin too.
         matchfn tries both the given filename and with .hglf stripped.
         """
         if opts is None:
             opts = {}
         matchandpats = oldmatchandpats(ctx, pats, opts, globbed, default,
                                        badfn=badfn)
         m, p = copy.copy(matchandpats)
 
         if m.always():
             # We want to match everything anyway, so there's no benefit trying
             # to add standins.
             return matchandpats
 
         pats = set(p)
 
         def fixpats(pat, tostandin=lfutil.standin):
             if pat.startswith('set:'):
                 return pat
 
             kindpat = matchmod._patsplit(pat, None)
 
             if kindpat[0] is not None:
                 return kindpat[0] + ':' + tostandin(kindpat[1])
             return tostandin(kindpat[1])
 
         if m._cwd:
             hglf = lfutil.shortname
             back = util.pconvert(m.rel(hglf)[:-len(hglf)])
 
             def tostandin(f):
                 # The file may already be a standin, so truncate the back
                 # prefix and test before mangling it.  This avoids turning
                 # 'glob:../.hglf/foo*' into 'glob:../.hglf/../.hglf/foo*'.
                 if f.startswith(back) and lfutil.splitstandin(f[len(back):]):
                     return f
 
                 # An absolute path is from outside the repo, so truncate the
                 # path to the root before building the standin.  Otherwise cwd
                 # is somewhere in the repo, relative to root, and needs to be
                 # prepended before building the standin.
                 if os.path.isabs(m._cwd):
                     f = f[len(back):]
                 else:
                     f = m._cwd + '/' + f
                 return back + lfutil.standin(f)
         else:
             def tostandin(f):
                 if lfutil.isstandin(f):
                     return f
                 return lfutil.standin(f)
         pats.update(fixpats(f, tostandin) for f in p)
 
         for i in range(0, len(m._files)):
             # Don't add '.hglf' to m.files, since that is already covered by '.'
             if m._files[i] == '.':
                 continue
             standin = lfutil.standin(m._files[i])
             # If the "standin" is a directory, append instead of replace to
             # support naming a directory on the command line with only
             # largefiles.  The original directory is kept to support normal
             # files.
             if standin in ctx:
                 m._files[i] = standin
             elif m._files[i] not in ctx and repo.wvfs.isdir(standin):
                 m._files.append(standin)
 
         m._fileset = set(m._files)
         m.always = lambda: False
         origmatchfn = m.matchfn
         def lfmatchfn(f):
             lf = lfutil.splitstandin(f)
             if lf is not None and origmatchfn(lf):
                 return True
             r = origmatchfn(f)
             return r
         m.matchfn = lfmatchfn
 
         ui.debug('updated patterns: %s\n' % ', '.join(sorted(pats)))
         return m, pats
 
     # For hg log --patch, the match object is used in two different senses:
     # (1) to determine what revisions should be printed out, and
     # (2) to determine what files to print out diffs for.
     # The magic matchandpats override should be used for case (1) but not for
     # case (2).
     def overridemakelogfilematcher(repo, pats, opts, badfn=None):
         wctx = repo[None]
         match, pats = oldmatchandpats(wctx, pats, opts, badfn=badfn)
         return lambda rev: match
 
     oldmatchandpats = installmatchandpatsfn(overridematchandpats)
     oldmakelogfilematcher = cmdutil._makenofollowlogfilematcher
     setattr(cmdutil, '_makenofollowlogfilematcher', overridemakelogfilematcher)
 
     try:
         return orig(ui, repo, *pats, **opts)
     finally:
         restorematchandpatsfn()
         setattr(cmdutil, '_makenofollowlogfilematcher', oldmakelogfilematcher)
 
 def overrideverify(orig, ui, repo, *pats, **opts):
     large = opts.pop('large', False)
     all = opts.pop('lfa', False)
     contents = opts.pop('lfc', False)
 
     result = orig(ui, repo, *pats, **opts)
     if large or all or contents:
         result = result or lfcommands.verifylfiles(ui, repo, all, contents)
     return result
 
 def overridedebugstate(orig, ui, repo, *pats, **opts):
     large = opts.pop('large', False)
     if large:
         class fakerepo(object):
             dirstate = lfutil.openlfdirstate(ui, repo)
         orig(ui, fakerepo, *pats, **opts)
     else:
         orig(ui, repo, *pats, **opts)
 
 # Before starting the manifest merge, merge.updates will call
 # _checkunknownfile to check if there are any files in the merged-in
 # changeset that collide with unknown files in the working copy.
 #
 # The largefiles are seen as unknown, so this prevents us from merging
 # in a file 'foo' if we already have a largefile with the same name.
 #
 # The overridden function filters the unknown files by removing any
 # largefiles. This makes the merge proceed and we can then handle this
 # case further in the overridden calculateupdates function below.
 def overridecheckunknownfile(origfn, repo, wctx, mctx, f, f2=None):
     if lfutil.standin(repo.dirstate.normalize(f)) in wctx:
         return False
     return origfn(repo, wctx, mctx, f, f2)
 
 # The manifest merge handles conflicts on the manifest level. We want
 # to handle changes in largefile-ness of files at this level too.
 #
 # The strategy is to run the original calculateupdates and then process
 # the action list it outputs. There are two cases we need to deal with:
 #
 # 1. Normal file in p1, largefile in p2. Here the largefile is
 #    detected via its standin file, which will enter the working copy
 #    with a "get" action. It is not "merge" since the standin is all
 #    Mercurial is concerned with at this level -- the link to the
 #    existing normal file is not relevant here.
 #
 # 2. Largefile in p1, normal file in p2. Here we get a "merge" action
 #    since the largefile will be present in the working copy and
 #    different from the normal file in p2. Mercurial therefore
 #    triggers a merge action.
 #
 # In both cases, we prompt the user and emit new actions to either
 # remove the standin (if the normal file was kept) or to remove the
 # normal file and get the standin (if the largefile was kept). The
 # default prompt answer is to use the largefile version since it was
 # presumably changed on purpose.
 #
 # Finally, the merge.applyupdates function will then take care of
 # writing the files into the working copy and lfcommands.updatelfiles
 # will update the largefiles.
 def overridecalculateupdates(origfn, repo, p1, p2, pas, branchmerge, force,
                              acceptremote, *args, **kwargs):
     overwrite = force and not branchmerge
     actions, diverge, renamedelete = origfn(
         repo, p1, p2, pas, branchmerge, force, acceptremote, *args, **kwargs)
 
     if overwrite:
         return actions, diverge, renamedelete
 
     # Convert to dictionary with filename as key and action as value.
     lfiles = set()
     for f in actions:
         splitstandin = lfutil.splitstandin(f)
         if splitstandin in p1:
             lfiles.add(splitstandin)
         elif lfutil.standin(f) in p1:
             lfiles.add(f)
 
     for lfile in sorted(lfiles):
         standin = lfutil.standin(lfile)
         (lm, largs, lmsg) = actions.get(lfile, (None, None, None))
         (sm, sargs, smsg) = actions.get(standin, (None, None, None))
         if sm in ('g', 'dc') and lm != 'r':
             if sm == 'dc':
                 f1, f2, fa, move, anc = sargs
                 sargs = (p2[f2].flags(), False)
             # Case 1: normal file in the working copy, largefile in
             # the second parent
             usermsg = _('remote turned local normal file %s into a largefile\n'
                         'use (l)argefile or keep (n)ormal file?'
                         '$$ &Largefile $$ &Normal file') % lfile
             if repo.ui.promptchoice(usermsg, 0) == 0: # pick remote largefile
                 actions[lfile] = ('r', None, 'replaced by standin')
                 actions[standin] = ('g', sargs, 'replaces standin')
             else: # keep local normal file
                 actions[lfile] = ('k', None, 'replaces standin')
                 if branchmerge:
                     actions[standin] = ('k', None, 'replaced by non-standin')
                 else:
                     actions[standin] = ('r', None, 'replaced by non-standin')
         elif lm in ('g', 'dc') and sm != 'r':
             if lm == 'dc':
                 f1, f2, fa, move, anc = largs
                 largs = (p2[f2].flags(), False)
             # Case 2: largefile in the working copy, normal file in
             # the second parent
             usermsg = _('remote turned local largefile %s into a normal file\n'
                     'keep (l)argefile or use (n)ormal file?'
                     '$$ &Largefile $$ &Normal file') % lfile
             if repo.ui.promptchoice(usermsg, 0) == 0: # keep local largefile
                 if branchmerge:
                     # largefile can be restored from standin safely
                     actions[lfile] = ('k', None, 'replaced by standin')
                     actions[standin] = ('k', None, 'replaces standin')
                 else:
                     # "lfile" should be marked as "removed" without
                     # removal of itself
                     actions[lfile] = ('lfmr', None,
                                       'forget non-standin largefile')
 
                     # linear-merge should treat this largefile as 're-added'
                     actions[standin] = ('a', None, 'keep standin')
             else: # pick remote normal file
                 actions[lfile] = ('g', largs, 'replaces standin')
                 actions[standin] = ('r', None, 'replaced by non-standin')
 
     return actions, diverge, renamedelete
 
 def mergerecordupdates(orig, repo, actions, branchmerge):
     if 'lfmr' in actions:
         lfdirstate = lfutil.openlfdirstate(repo.ui, repo)
         for lfile, args, msg in actions['lfmr']:
             # this should be executed before 'orig', to execute 'remove'
             # before all other actions
             repo.dirstate.remove(lfile)
             # make sure lfile doesn't get synclfdirstate'd as normal
             lfdirstate.add(lfile)
         lfdirstate.write()
 
     return orig(repo, actions, branchmerge)
 
 # Override filemerge to prompt the user about how they wish to merge
 # largefiles. This will handle identical edits without prompting the user.
 def overridefilemerge(origfn, premerge, repo, wctx, mynode, orig, fcd, fco, fca,
                       labels=None):
     if not lfutil.isstandin(orig) or fcd.isabsent() or fco.isabsent():
         return origfn(premerge, repo, wctx, mynode, orig, fcd, fco, fca,
                       labels=labels)
 
     ahash = lfutil.readasstandin(fca).lower()
     dhash = lfutil.readasstandin(fcd).lower()
     ohash = lfutil.readasstandin(fco).lower()
     if (ohash != ahash and
         ohash != dhash and
         (dhash == ahash or
          repo.ui.promptchoice(
              _('largefile %s has a merge conflict\nancestor was %s\n'
                'keep (l)ocal %s or\ntake (o)ther %s?'
                '$$ &Local $$ &Other') %
                (lfutil.splitstandin(orig), ahash, dhash, ohash),
              0) == 1)):
         repo.wwrite(fcd.path(), fco.data(), fco.flags())
     return True, 0, False
 
 def copiespathcopies(orig, ctx1, ctx2, match=None):
     copies = orig(ctx1, ctx2, match=match)
     updated = {}
 
     for k, v in copies.iteritems():
         updated[lfutil.splitstandin(k) or k] = lfutil.splitstandin(v) or v
 
     return updated
 
 # Copy first changes the matchers to match standins instead of
 # largefiles.  Then it overrides util.copyfile in that function it
 # checks if the destination largefile already exists. It also keeps a
 # list of copied files so that the largefiles can be copied and the
 # dirstate updated.
 def overridecopy(orig, ui, repo, pats, opts, rename=False):
     # doesn't remove largefile on rename
     if len(pats) < 2:
         # this isn't legal, let the original function deal with it
         return orig(ui, repo, pats, opts, rename)
 
     # This could copy both lfiles and normal files in one command,
     # but we don't want to do that. First replace their matcher to
     # only match normal files and run it, then replace it to just
     # match largefiles and run it again.
     nonormalfiles = False
     nolfiles = False
     installnormalfilesmatchfn(repo[None].manifest())
     try:
         result = orig(ui, repo, pats, opts, rename)
     except error.Abort as e:
         if str(e) != _('no files to copy'):
             raise e
         else:
             nonormalfiles = True
         result = 0
     finally:
         restorematchfn()
 
     # The first rename can cause our current working directory to be removed.
     # In that case there is nothing left to copy/rename so just quit.
     try:
         repo.getcwd()
     except OSError:
         return result
 
     def makestandin(relpath):
         path = pathutil.canonpath(repo.root, repo.getcwd(), relpath)
         return repo.wvfs.join(lfutil.standin(path))
 
     fullpats = scmutil.expandpats(pats)
     dest = fullpats[-1]
 
     if os.path.isdir(dest):
         if not os.path.isdir(makestandin(dest)):
             os.makedirs(makestandin(dest))
 
     try:
         # When we call orig below it creates the standins but we don't add
         # them to the dir state until later so lock during that time.
         wlock = repo.wlock()
 
         manifest = repo[None].manifest()
         def overridematch(ctx, pats=(), opts=None, globbed=False,
                 default='relpath', badfn=None):
             if opts is None:
                 opts = {}
             newpats = []
             # The patterns were previously mangled to add the standin
             # directory; we need to remove that now
             for pat in pats:
                 if matchmod.patkind(pat) is None and lfutil.shortname in pat:
                     newpats.append(pat.replace(lfutil.shortname, ''))
                 else:
                     newpats.append(pat)
             match = oldmatch(ctx, newpats, opts, globbed, default, badfn=badfn)
             m = copy.copy(match)
             lfile = lambda f: lfutil.standin(f) in manifest
             m._files = [lfutil.standin(f) for f in m._files if lfile(f)]
             m._fileset = set(m._files)
             origmatchfn = m.matchfn
             def matchfn(f):
                 lfile = lfutil.splitstandin(f)
                 return (lfile is not None and
                         (f in manifest) and
                         origmatchfn(lfile) or
                         None)
             m.matchfn = matchfn
             return m
         oldmatch = installmatchfn(overridematch)
         listpats = []
         for pat in pats:
             if matchmod.patkind(pat) is not None:
                 listpats.append(pat)
             else:
                 listpats.append(makestandin(pat))
 
         try:
             origcopyfile = util.copyfile
             copiedfiles = []
             def overridecopyfile(src, dest):
                 if (lfutil.shortname in src and
                     dest.startswith(repo.wjoin(lfutil.shortname))):
                     destlfile = dest.replace(lfutil.shortname, '')
                     if not opts['force'] and os.path.exists(destlfile):
                         raise IOError('',
                             _('destination largefile already exists'))
                 copiedfiles.append((src, dest))
                 origcopyfile(src, dest)
 
             util.copyfile = overridecopyfile
             result += orig(ui, repo, listpats, opts, rename)
         finally:
             util.copyfile = origcopyfile
 
         lfdirstate = lfutil.openlfdirstate(ui, repo)
         for (src, dest) in copiedfiles:
             if (lfutil.shortname in src and
                 dest.startswith(repo.wjoin(lfutil.shortname))):
                 srclfile = src.replace(repo.wjoin(lfutil.standin('')), '')
                 destlfile = dest.replace(repo.wjoin(lfutil.standin('')), '')
                 destlfiledir = repo.wvfs.dirname(repo.wjoin(destlfile)) or '.'
                 if not os.path.isdir(destlfiledir):
                     os.makedirs(destlfiledir)
                 if rename:
                     os.rename(repo.wjoin(srclfile), repo.wjoin(destlfile))
 
                     # The file is gone, but this deletes any empty parent
                     # directories as a side-effect.
                     repo.wvfs.unlinkpath(srclfile, ignoremissing=True)
                     lfdirstate.remove(srclfile)
                 else:
                     util.copyfile(repo.wjoin(srclfile),
                                   repo.wjoin(destlfile))
 
                 lfdirstate.add(destlfile)
         lfdirstate.write()
     except error.Abort as e:
         if str(e) != _('no files to copy'):
             raise e
         else:
             nolfiles = True
     finally:
         restorematchfn()
         wlock.release()
 
     if nolfiles and nonormalfiles:
         raise error.Abort(_('no files to copy'))
 
     return result
 
 # When the user calls revert, we have to be careful to not revert any
 # changes to other largefiles accidentally. This means we have to keep
 # track of the largefiles that are being reverted so we only pull down
 # the necessary largefiles.
 #
 # Standins are only updated (to match the hash of largefiles) before
 # commits. Update the standins then run the original revert, changing
 # the matcher to hit standins instead of largefiles. Based on the
 # resulting standins update the largefiles.
 def overriderevert(orig, ui, repo, ctx, parents, *pats, **opts):
     # Because we put the standins in a bad state (by updating them)
     # and then return them to a correct state we need to lock to
     # prevent others from changing them in their incorrect state.
     with repo.wlock():
         lfdirstate = lfutil.openlfdirstate(ui, repo)
         s = lfutil.lfdirstatestatus(lfdirstate, repo)
         lfdirstate.write()
         for lfile in s.modified:
             lfutil.updatestandin(repo, lfile, lfutil.standin(lfile))
         for lfile in s.deleted:
             fstandin = lfutil.standin(lfile)
             if (repo.wvfs.exists(fstandin)):
                 repo.wvfs.unlink(fstandin)
 
         oldstandins = lfutil.getstandinsstate(repo)
 
         def overridematch(mctx, pats=(), opts=None, globbed=False,
                 default='relpath', badfn=None):
             if opts is None:
                 opts = {}
             match = oldmatch(mctx, pats, opts, globbed, default, badfn=badfn)
             m = copy.copy(match)
 
             # revert supports recursing into subrepos, and though largefiles
             # currently doesn't work correctly in that case, this match is
             # called, so the lfdirstate above may not be the correct one for
             # this invocation of match.
             lfdirstate = lfutil.openlfdirstate(mctx.repo().ui, mctx.repo(),
                                                False)
 
             wctx = repo[None]
             matchfiles = []
             for f in m._files:
                 standin = lfutil.standin(f)
                 if standin in ctx or standin in mctx:
                     matchfiles.append(standin)
                 elif standin in wctx or lfdirstate[f] == 'r':
                     continue
                 else:
                     matchfiles.append(f)
             m._files = matchfiles
             m._fileset = set(m._files)
             origmatchfn = m.matchfn
             def matchfn(f):
                 lfile = lfutil.splitstandin(f)
                 if lfile is not None:
                     return (origmatchfn(lfile) and
                             (f in ctx or f in mctx))
                 return origmatchfn(f)
             m.matchfn = matchfn
             return m
         oldmatch = installmatchfn(overridematch)
         try:
             orig(ui, repo, ctx, parents, *pats, **opts)
         finally:
             restorematchfn()
 
         newstandins = lfutil.getstandinsstate(repo)
         filelist = lfutil.getlfilestoupdate(oldstandins, newstandins)
         # lfdirstate should be 'normallookup'-ed for updated files,
         # because reverting doesn't touch dirstate for 'normal' files
         # when target revision is explicitly specified: in such case,
         # 'n' and valid timestamp in dirstate doesn't ensure 'clean'
         # of target (standin) file.
         lfcommands.updatelfiles(ui, repo, filelist, printmessage=False,
                                 normallookup=True)
 
 # after pulling changesets, we need to take some extra care to get
 # largefiles updated remotely
 def overridepull(orig, ui, repo, source=None, **opts):
     revsprepull = len(repo)
     if not source:
         source = 'default'
     repo.lfpullsource = source
     result = orig(ui, repo, source, **opts)
     revspostpull = len(repo)
     lfrevs = opts.get('lfrev', [])
     if opts.get('all_largefiles'):
         lfrevs.append('pulled()')
     if lfrevs and revspostpull > revsprepull:
         numcached = 0
         repo.firstpulled = revsprepull # for pulled() revset expression
         try:
             for rev in scmutil.revrange(repo, lfrevs):
                 ui.note(_('pulling largefiles for revision %s\n') % rev)
                 (cached, missing) = lfcommands.cachelfiles(ui, repo, rev)
                 numcached += len(cached)
         finally:
             del repo.firstpulled
         ui.status(_("%d largefiles cached\n") % numcached)
     return result
 
 def overridepush(orig, ui, repo, *args, **kwargs):
     """Override push command and store --lfrev parameters in opargs"""
     lfrevs = kwargs.pop('lfrev', None)
     if lfrevs:
         opargs = kwargs.setdefault('opargs', {})
         opargs['lfrevs'] = scmutil.revrange(repo, lfrevs)
     return orig(ui, repo, *args, **kwargs)
 
 def exchangepushoperation(orig, *args, **kwargs):
     """Override pushoperation constructor and store lfrevs parameter"""
     lfrevs = kwargs.pop('lfrevs', None)
     pushop = orig(*args, **kwargs)
     pushop.lfrevs = lfrevs
     return pushop
 
 revsetpredicate = registrar.revsetpredicate()
 
 @revsetpredicate('pulled()')
 def pulledrevsetsymbol(repo, subset, x):
     """Changesets that just has been pulled.
 
     Only available with largefiles from pull --lfrev expressions.
 
     .. container:: verbose
 
       Some examples:
 
       - pull largefiles for all new changesets::
 
           hg pull -lfrev "pulled()"
 
       - pull largefiles for all new branch heads::
 
           hg pull -lfrev "head(pulled()) and not closed()"
 
     """
 
     try:
         firstpulled = repo.firstpulled
     except AttributeError:
         raise error.Abort(_("pulled() only available in --lfrev"))
     return smartset.baseset([r for r in subset if r >= firstpulled])
 
 def overrideclone(orig, ui, source, dest=None, **opts):
     d = dest
     if d is None:
         d = hg.defaultdest(source)
     if opts.get('all_largefiles') and not hg.islocal(d):
             raise error.Abort(_(
             '--all-largefiles is incompatible with non-local destination %s') %
             d)
 
     return orig(ui, source, dest, **opts)
 
 def hgclone(orig, ui, opts, *args, **kwargs):
     result = orig(ui, opts, *args, **kwargs)
 
     if result is not None:
         sourcerepo, destrepo = result
         repo = destrepo.local()
 
         # When cloning to a remote repo (like through SSH), no repo is available
         # from the peer.   Therefore the largefiles can't be downloaded and the
         # hgrc can't be updated.
         if not repo:
             return result
 
         # If largefiles is required for this repo, permanently enable it locally
         if 'largefiles' in repo.requirements:
             with repo.vfs('hgrc', 'a', text=True) as fp:
                 fp.write('\n[extensions]\nlargefiles=\n')
 
         # Caching is implicitly limited to 'rev' option, since the dest repo was
         # truncated at that point.  The user may expect a download count with
         # this option, so attempt whether or not this is a largefile repo.
         if opts.get('all_largefiles'):
             success, missing = lfcommands.downloadlfiles(ui, repo, None)
 
             if missing != 0:
                 return None
 
     return result
 
 def hgpostshare(orig, sourcerepo, destrepo, bookmarks=True, defaultpath=None):
     orig(sourcerepo, destrepo, bookmarks, defaultpath)
 
     # If largefiles is required for this repo, permanently enable it locally
     if 'largefiles' in destrepo.requirements:
         with destrepo.vfs('hgrc', 'a+', text=True) as fp:
             fp.write('\n[extensions]\nlargefiles=\n')
 
 def overriderebase(orig, ui, repo, **opts):
     if not util.safehasattr(repo, '_largefilesenabled'):
         return orig(ui, repo, **opts)
 
     resuming = opts.get('continue')
     repo._lfcommithooks.append(lfutil.automatedcommithook(resuming))
     repo._lfstatuswriters.append(lambda *msg, **opts: None)
     try:
         return orig(ui, repo, **opts)
     finally:
         repo._lfstatuswriters.pop()
         repo._lfcommithooks.pop()
 
 def overridearchivecmd(orig, ui, repo, dest, **opts):
     repo.unfiltered().lfstatus = True
 
     try:
         return orig(ui, repo.unfiltered(), dest, **opts)
     finally:
         repo.unfiltered().lfstatus = False
 
 def hgwebarchive(orig, web, req, tmpl):
     web.repo.lfstatus = True
 
     try:
         return orig(web, req, tmpl)
     finally:
         web.repo.lfstatus = False
 
 def overridearchive(orig, repo, dest, node, kind, decode=True, matchfn=None,
             prefix='', mtime=None, subrepos=None):
     # For some reason setting repo.lfstatus in hgwebarchive only changes the
     # unfiltered repo's attr, so check that as well.
     if not repo.lfstatus and not repo.unfiltered().lfstatus:
         return orig(repo, dest, node, kind, decode, matchfn, prefix, mtime,
                     subrepos)
 
     # No need to lock because we are only reading history and
     # largefile caches, neither of which are modified.
     if node is not None:
         lfcommands.cachelfiles(repo.ui, repo, node)
 
     if kind not in archival.archivers:
         raise error.Abort(_("unknown archive type '%s'") % kind)
 
     ctx = repo[node]
 
     if kind == 'files':
         if prefix:
             raise error.Abort(
                 _('cannot give prefix when archiving to files'))
     else:
         prefix = archival.tidyprefix(dest, kind, prefix)
 
     def write(name, mode, islink, getdata):
         if matchfn and not matchfn(name):
             return
         data = getdata()
         if decode:
             data = repo.wwritedata(name, data)
         archiver.addfile(prefix + name, mode, islink, data)
 
     archiver = archival.archivers[kind](dest, mtime or ctx.date()[0])
 
     if repo.ui.configbool("ui", "archivemeta"):
         write('.hg_archival.txt', 0o644, False,
               lambda: archival.buildmetadata(ctx))
 
     for f in ctx:
         ff = ctx.flags(f)
         getdata = ctx[f].data
         lfile = lfutil.splitstandin(f)
         if lfile is not None:
             if node is not None:
                 path = lfutil.findfile(repo, getdata().strip())
 
                 if path is None:
                     raise error.Abort(
                        _('largefile %s not found in repo store or system cache')
                        % lfile)
             else:
                 path = lfile
 
             f = lfile
 
             getdata = lambda: util.readfile(path)
         write(f, 'x' in ff and 0o755 or 0o644, 'l' in ff, getdata)
 
     if subrepos:
         for subpath in sorted(ctx.substate):
             sub = ctx.workingsub(subpath)
             submatch = matchmod.subdirmatcher(subpath, matchfn)
             sub._repo.lfstatus = True
             sub.archive(archiver, prefix, submatch)
 
     archiver.done()
 
 def hgsubrepoarchive(orig, repo, archiver, prefix, match=None, decode=True):
     lfenabled = util.safehasattr(repo._repo, '_largefilesenabled')
     if not lfenabled or not repo._repo.lfstatus:
         return orig(repo, archiver, prefix, match, decode)
 
     repo._get(repo._state + ('hg',))
     rev = repo._state[1]
     ctx = repo._repo[rev]
 
     if ctx.node() is not None:
         lfcommands.cachelfiles(repo.ui, repo._repo, ctx.node())
 
     def write(name, mode, islink, getdata):
         # At this point, the standin has been replaced with the largefile name,
         # so the normal matcher works here without the lfutil variants.
         if match and not match(f):
             return
         data = getdata()
         if decode:
             data = repo._repo.wwritedata(name, data)
 
         archiver.addfile(prefix + repo._path + '/' + name, mode, islink, data)
 
     for f in ctx:
         ff = ctx.flags(f)
         getdata = ctx[f].data
         lfile = lfutil.splitstandin(f)
         if lfile is not None:
             if ctx.node() is not None:
                 path = lfutil.findfile(repo._repo, getdata().strip())
 
                 if path is None:
                     raise error.Abort(
                        _('largefile %s not found in repo store or system cache')
                        % lfile)
             else:
                 path = lfile
 
             f = lfile
 
             getdata = lambda: util.readfile(os.path.join(prefix, path))
 
         write(f, 'x' in ff and 0o755 or 0o644, 'l' in ff, getdata)
 
     for subpath in sorted(ctx.substate):
         sub = ctx.workingsub(subpath)
         submatch = matchmod.subdirmatcher(subpath, match)
         sub._repo.lfstatus = True
         sub.archive(archiver, prefix + repo._path + '/', submatch, decode)
 
 # If a largefile is modified, the change is not reflected in its
 # standin until a commit. cmdutil.bailifchanged() raises an exception
 # if the repo has uncommitted changes. Wrap it to also check if
 # largefiles were changed. This is used by bisect, backout and fetch.
 def overridebailifchanged(orig, repo, *args, **kwargs):
     orig(repo, *args, **kwargs)
     repo.lfstatus = True
     s = repo.status()
     repo.lfstatus = False
     if s.modified or s.added or s.removed or s.deleted:
         raise error.Abort(_('uncommitted changes'))
 
 def postcommitstatus(orig, repo, *args, **kwargs):
     repo.lfstatus = True
     try:
         return orig(repo, *args, **kwargs)
     finally:
         repo.lfstatus = False
 
 def cmdutilforget(orig, ui, repo, match, prefix, explicitonly):
     normalmatcher = composenormalfilematcher(match, repo[None].manifest())
     bad, forgot = orig(ui, repo, normalmatcher, prefix, explicitonly)
     m = composelargefilematcher(match, repo[None].manifest())
 
     try:
         repo.lfstatus = True
         s = repo.status(match=m, clean=True)
     finally:
         repo.lfstatus = False
     manifest = repo[None].manifest()
     forget = sorted(s.modified + s.added + s.deleted + s.clean)
     forget = [f for f in forget if lfutil.standin(f) in manifest]
 
     for f in forget:
         fstandin = lfutil.standin(f)
         if fstandin not in repo.dirstate and not repo.wvfs.isdir(fstandin):
             ui.warn(_('not removing %s: file is already untracked\n')
                     % m.rel(f))
             bad.append(f)
 
     for f in forget:
         if ui.verbose or not m.exact(f):
             ui.status(_('removing %s\n') % m.rel(f))
 
     # Need to lock because standin files are deleted then removed from the
     # repository and we could race in-between.
     with repo.wlock():
         lfdirstate = lfutil.openlfdirstate(ui, repo)
         for f in forget:
             if lfdirstate[f] == 'a':
                 lfdirstate.drop(f)
             else:
                 lfdirstate.remove(f)
         lfdirstate.write()
         standins = [lfutil.standin(f) for f in forget]
         for f in standins:
             repo.wvfs.unlinkpath(f, ignoremissing=True)
         rejected = repo[None].forget(standins)
 
     bad.extend(f for f in rejected if f in m.files())
     forgot.extend(f for f in forget if f not in rejected)
     return bad, forgot
 
 def _getoutgoings(repo, other, missing, addfunc):
     """get pairs of filename and largefile hash in outgoing revisions
     in 'missing'.
 
     largefiles already existing on 'other' repository are ignored.
 
     'addfunc' is invoked with each unique pairs of filename and
     largefile hash value.
     """
     knowns = set()
     lfhashes = set()
     def dedup(fn, lfhash):
         k = (fn, lfhash)
         if k not in knowns:
             knowns.add(k)
             lfhashes.add(lfhash)
     lfutil.getlfilestoupload(repo, missing, dedup)
     if lfhashes:
         lfexists = storefactory.openstore(repo, other).exists(lfhashes)
         for fn, lfhash in knowns:
             if not lfexists[lfhash]: # lfhash doesn't exist on "other"
                 addfunc(fn, lfhash)
 
 def outgoinghook(ui, repo, other, opts, missing):
     if opts.pop('large', None):
         lfhashes = set()
         if ui.debugflag:
             toupload = {}
             def addfunc(fn, lfhash):
                 if fn not in toupload:
                     toupload[fn] = []
                 toupload[fn].append(lfhash)
                 lfhashes.add(lfhash)
             def showhashes(fn):
                 for lfhash in sorted(toupload[fn]):
                     ui.debug('    %s\n' % (lfhash))
         else:
             toupload = set()
             def addfunc(fn, lfhash):
                 toupload.add(fn)
                 lfhashes.add(lfhash)
             def showhashes(fn):
                 pass
         _getoutgoings(repo, other, missing, addfunc)
 
         if not toupload:
             ui.status(_('largefiles: no files to upload\n'))
         else:
             ui.status(_('largefiles to upload (%d entities):\n')
                       % (len(lfhashes)))
             for file in sorted(toupload):
                 ui.status(lfutil.splitstandin(file) + '\n')
                 showhashes(file)
             ui.status('\n')
 
 def summaryremotehook(ui, repo, opts, changes):
     largeopt = opts.get('large', False)
     if changes is None:
         if largeopt:
             return (False, True) # only outgoing check is needed
         else:
             return (False, False)
     elif largeopt:
         url, branch, peer, outgoing = changes[1]
         if peer is None:
             # i18n: column positioning for "hg summary"
             ui.status(_('largefiles: (no remote repo)\n'))
             return
 
         toupload = set()
         lfhashes = set()
         def addfunc(fn, lfhash):
             toupload.add(fn)
             lfhashes.add(lfhash)
         _getoutgoings(repo, peer, outgoing.missing, addfunc)
 
         if not toupload:
             # i18n: column positioning for "hg summary"
             ui.status(_('largefiles: (no files to upload)\n'))
         else:
             # i18n: column positioning for "hg summary"
             ui.status(_('largefiles: %d entities for %d files to upload\n')
                       % (len(lfhashes), len(toupload)))
 
 def overridesummary(orig, ui, repo, *pats, **opts):
     try:
         repo.lfstatus = True
         orig(ui, repo, *pats, **opts)
     finally:
         repo.lfstatus = False
 
 def scmutiladdremove(orig, repo, matcher, prefix, opts=None, dry_run=None,
                      similarity=None):
     if opts is None:
         opts = {}
     if not lfutil.islfilesrepo(repo):
         return orig(repo, matcher, prefix, opts, dry_run, similarity)
     # Get the list of missing largefiles so we can remove them
     lfdirstate = lfutil.openlfdirstate(repo.ui, repo)
-    unsure, s = lfdirstate.status(matchmod.always(repo.root, repo.getcwd()), [],
-                                  False, False, False)
+    unsure, s = lfdirstate.status(matchmod.always(repo.root, repo.getcwd()),
+                                  subrepos=[], ignored=False, clean=False,
+                                  unknown=False)
 
     # Call into the normal remove code, but the removing of the standin, we want
     # to have handled by original addremove.  Monkey patching here makes sure
     # we don't remove the standin in the largefiles code, preventing a very
     # confused state later.
     if s.deleted:
         m = copy.copy(matcher)
 
         # The m._files and m._map attributes are not changed to the deleted list
         # because that affects the m.exact() test, which in turn governs whether
         # or not the file name is printed, and how.  Simply limit the original
         # matches to those in the deleted status list.
         matchfn = m.matchfn
         m.matchfn = lambda f: f in s.deleted and matchfn(f)
 
         removelargefiles(repo.ui, repo, True, m, **opts)
     # Call into the normal add code, and any files that *should* be added as
     # largefiles will be
     added, bad = addlargefiles(repo.ui, repo, True, matcher, **opts)
     # Now that we've handled largefiles, hand off to the original addremove
     # function to take care of the rest.  Make sure it doesn't do anything with
     # largefiles by passing a matcher that will ignore them.
     matcher = composenormalfilematcher(matcher, repo[None].manifest(), added)
     return orig(repo, matcher, prefix, opts, dry_run, similarity)
 
 # Calling purge with --all will cause the largefiles to be deleted.
 # Override repo.status to prevent this from happening.
 def overridepurge(orig, ui, repo, *dirs, **opts):
     # XXX Monkey patching a repoview will not work. The assigned attribute will
     # be set on the unfiltered repo, but we will only lookup attributes in the
     # unfiltered repo if the lookup in the repoview object itself fails. As the
     # monkey patched method exists on the repoview class the lookup will not
     # fail. As a result, the original version will shadow the monkey patched
     # one, defeating the monkey patch.
     #
     # As a work around we use an unfiltered repo here. We should do something
     # cleaner instead.
     repo = repo.unfiltered()
     oldstatus = repo.status
     def overridestatus(node1='.', node2=None, match=None, ignored=False,
                         clean=False, unknown=False, listsubrepos=False):
         r = oldstatus(node1, node2, match, ignored, clean, unknown,
                       listsubrepos)
         lfdirstate = lfutil.openlfdirstate(ui, repo)
         unknown = [f for f in r.unknown if lfdirstate[f] == '?']
         ignored = [f for f in r.ignored if lfdirstate[f] == '?']
         return scmutil.status(r.modified, r.added, r.removed, r.deleted,
                               unknown, ignored, r.clean)
     repo.status = overridestatus
     orig(ui, repo, *dirs, **opts)
     repo.status = oldstatus
 def overriderollback(orig, ui, repo, **opts):
     with repo.wlock():
         before = repo.dirstate.parents()
         orphans = set(f for f in repo.dirstate
                       if lfutil.isstandin(f) and repo.dirstate[f] != 'r')
         result = orig(ui, repo, **opts)
         after = repo.dirstate.parents()
         if before == after:
             return result # no need to restore standins
 
         pctx = repo['.']
         for f in repo.dirstate:
             if lfutil.isstandin(f):
                 orphans.discard(f)
                 if repo.dirstate[f] == 'r':
                     repo.wvfs.unlinkpath(f, ignoremissing=True)
                 elif f in pctx:
                     fctx = pctx[f]
                     repo.wwrite(f, fctx.data(), fctx.flags())
                 else:
                     # content of standin is not so important in 'a',
                     # 'm' or 'n' (coming from the 2nd parent) cases
                     lfutil.writestandin(repo, f, '', False)
         for standin in orphans:
             repo.wvfs.unlinkpath(standin, ignoremissing=True)
 
         lfdirstate = lfutil.openlfdirstate(ui, repo)
         orphans = set(lfdirstate)
         lfiles = lfutil.listlfiles(repo)
         for file in lfiles:
             lfutil.synclfdirstate(repo, lfdirstate, file, True)
             orphans.discard(file)
         for lfile in orphans:
             lfdirstate.drop(lfile)
         lfdirstate.write()
     return result
 
 def overridetransplant(orig, ui, repo, *revs, **opts):
     resuming = opts.get('continue')
     repo._lfcommithooks.append(lfutil.automatedcommithook(resuming))
     repo._lfstatuswriters.append(lambda *msg, **opts: None)
     try:
         result = orig(ui, repo, *revs, **opts)
     finally:
         repo._lfstatuswriters.pop()
         repo._lfcommithooks.pop()
     return result
 
 def overridecat(orig, ui, repo, file1, *pats, **opts):
     ctx = scmutil.revsingle(repo, opts.get('rev'))
     err = 1
     notbad = set()
     m = scmutil.match(ctx, (file1,) + pats, opts)
     origmatchfn = m.matchfn
     def lfmatchfn(f):
         if origmatchfn(f):
             return True
         lf = lfutil.splitstandin(f)
         if lf is None:
             return False
         notbad.add(lf)
         return origmatchfn(lf)
     m.matchfn = lfmatchfn
     origbadfn = m.bad
     def lfbadfn(f, msg):
         if not f in notbad:
             origbadfn(f, msg)
     m.bad = lfbadfn
 
     origvisitdirfn = m.visitdir
     def lfvisitdirfn(dir):
         if dir == lfutil.shortname:
             return True
         ret = origvisitdirfn(dir)
         if ret:
             return ret
         lf = lfutil.splitstandin(dir)
         if lf is None:
             return False
         return origvisitdirfn(lf)
     m.visitdir = lfvisitdirfn
 
     for f in ctx.walk(m):
         with cmdutil.makefileobj(repo, opts.get('output'), ctx.node(),
                                  pathname=f) as fp:
             lf = lfutil.splitstandin(f)
             if lf is None or origmatchfn(f):
                 # duplicating unreachable code from commands.cat
                 data = ctx[f].data()
                 if opts.get('decode'):
                     data = repo.wwritedata(f, data)
                 fp.write(data)
             else:
                 hash = lfutil.readasstandin(ctx[f])
                 if not lfutil.inusercache(repo.ui, hash):
                     store = storefactory.openstore(repo)
                     success, missing = store.get([(lf, hash)])
                     if len(success) != 1:
                         raise error.Abort(
                             _('largefile %s is not in cache and could not be '
                               'downloaded')  % lf)
                 path = lfutil.usercachepath(repo.ui, hash)
                 with open(path, "rb") as fpin:
                     for chunk in util.filechunkiter(fpin):
                         fp.write(chunk)
         err = 0
     return err
 
 def mergeupdate(orig, repo, node, branchmerge, force,
                 *args, **kwargs):
     matcher = kwargs.get('matcher', None)
     # note if this is a partial update
     partial = matcher and not matcher.always()
     with repo.wlock():
         # branch |       |         |
         #  merge | force | partial | action
         # -------+-------+---------+--------------
         #    x   |   x   |    x    | linear-merge
         #    o   |   x   |    x    | branch-merge
         #    x   |   o   |    x    | overwrite (as clean update)
         #    o   |   o   |    x    | force-branch-merge (*1)
         #    x   |   x   |    o    |   (*)
         #    o   |   x   |    o    |   (*)
         #    x   |   o   |    o    | overwrite (as revert)
         #    o   |   o   |    o    |   (*)
         #
         # (*) don't care
         # (*1) deprecated, but used internally (e.g: "rebase --collapse")
 
         lfdirstate = lfutil.openlfdirstate(repo.ui, repo)
         unsure, s = lfdirstate.status(matchmod.always(repo.root,
                                                     repo.getcwd()),
-                                      [], False, True, False)
+                                      subrepos=[], ignored=False,
+                                      clean=True, unknown=False)
         oldclean = set(s.clean)
         pctx = repo['.']
         dctx = repo[node]
         for lfile in unsure + s.modified:
             lfileabs = repo.wvfs.join(lfile)
             if not repo.wvfs.exists(lfileabs):
                 continue
             lfhash = lfutil.hashfile(lfileabs)
             standin = lfutil.standin(lfile)
             lfutil.writestandin(repo, standin, lfhash,
                                 lfutil.getexecutable(lfileabs))
             if (standin in pctx and
                 lfhash == lfutil.readasstandin(pctx[standin])):
                 oldclean.add(lfile)
         for lfile in s.added:
             fstandin = lfutil.standin(lfile)
             if fstandin not in dctx:
                 # in this case, content of standin file is meaningless
                 # (in dctx, lfile is unknown, or normal file)
                 continue
             lfutil.updatestandin(repo, lfile, fstandin)
         # mark all clean largefiles as dirty, just in case the update gets
         # interrupted before largefiles and lfdirstate are synchronized
         for lfile in oldclean:
             lfdirstate.normallookup(lfile)
         lfdirstate.write()
 
         oldstandins = lfutil.getstandinsstate(repo)
         # Make sure the merge runs on disk, not in-memory. largefiles is not a
         # good candidate for in-memory merge (large files, custom dirstate,
         # matcher usage).
         kwargs['wc'] = repo[None]
         result = orig(repo, node, branchmerge, force, *args, **kwargs)
 
         newstandins = lfutil.getstandinsstate(repo)
         filelist = lfutil.getlfilestoupdate(oldstandins, newstandins)
 
         # to avoid leaving all largefiles as dirty and thus rehash them, mark
         # all the ones that didn't change as clean
         for lfile in oldclean.difference(filelist):
             lfdirstate.normal(lfile)
         lfdirstate.write()
 
         if branchmerge or force or partial:
             filelist.extend(s.deleted + s.removed)
 
         lfcommands.updatelfiles(repo.ui, repo, filelist=filelist,
                                 normallookup=partial)
 
         return result
 
 def scmutilmarktouched(orig, repo, files, *args, **kwargs):
     result = orig(repo, files, *args, **kwargs)
 
     filelist = []
     for f in files:
         lf = lfutil.splitstandin(f)
         if lf is not None:
             filelist.append(lf)
     if filelist:
         lfcommands.updatelfiles(repo.ui, repo, filelist=filelist,
                                 printmessage=False, normallookup=True)
 
     return result

hgext/largefiles/reposetup.py

@@ -1,391 +1,393 b''
 # Copyright 2009-2010 Gregory P. Ward
 # Copyright 2009-2010 Intelerad Medical Systems Incorporated
 # Copyright 2010-2011 Fog Creek Software
 # Copyright 2010-2011 Unity Technologies
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 '''setup for largefiles repositories: reposetup'''
 from __future__ import absolute_import
 
 import copy
 
 from mercurial.i18n import _
 
 from mercurial import (
     error,
     localrepo,
     match as matchmod,
     scmutil,
 )
 
 from . import (
     lfcommands,
     lfutil,
 )
 
 def reposetup(ui, repo):
     # wire repositories should be given new wireproto functions
     # by "proto.wirereposetup()" via "hg.wirepeersetupfuncs"
     if not repo.local():
         return
 
     class lfilesrepo(repo.__class__):
         # the mark to examine whether "repo" object enables largefiles or not
         _largefilesenabled = True
 
         lfstatus = False
         def status_nolfiles(self, *args, **kwargs):
             return super(lfilesrepo, self).status(*args, **kwargs)
 
         # When lfstatus is set, return a context that gives the names
         # of largefiles instead of their corresponding standins and
         # identifies the largefiles as always binary, regardless of
         # their actual contents.
         def __getitem__(self, changeid):
             ctx = super(lfilesrepo, self).__getitem__(changeid)
             if self.lfstatus:
                 class lfilesctx(ctx.__class__):
                     def files(self):
                         filenames = super(lfilesctx, self).files()
                         return [lfutil.splitstandin(f) or f for f in filenames]
                     def manifest(self):
                         man1 = super(lfilesctx, self).manifest()
                         class lfilesmanifest(man1.__class__):
                             def __contains__(self, filename):
                                 orig = super(lfilesmanifest, self).__contains__
                                 return (orig(filename) or
                                         orig(lfutil.standin(filename)))
                         man1.__class__ = lfilesmanifest
                         return man1
                     def filectx(self, path, fileid=None, filelog=None):
                         orig = super(lfilesctx, self).filectx
                         try:
                             if filelog is not None:
                                 result = orig(path, fileid, filelog)
                             else:
                                 result = orig(path, fileid)
                         except error.LookupError:
                             # Adding a null character will cause Mercurial to
                             # identify this as a binary file.
                             if filelog is not None:
                                 result = orig(lfutil.standin(path), fileid,
                                               filelog)
                             else:
                                 result = orig(lfutil.standin(path), fileid)
                             olddata = result.data
                             result.data = lambda: olddata() + '\0'
                         return result
                 ctx.__class__ = lfilesctx
             return ctx
 
         # Figure out the status of big files and insert them into the
         # appropriate list in the result. Also removes standin files
         # from the listing. Revert to the original status if
         # self.lfstatus is False.
         # XXX large file status is buggy when used on repo proxy.
         # XXX this needs to be investigated.
         @localrepo.unfilteredmethod
         def status(self, node1='.', node2=None, match=None, ignored=False,
                 clean=False, unknown=False, listsubrepos=False):
             listignored, listclean, listunknown = ignored, clean, unknown
             orig = super(lfilesrepo, self).status
             if not self.lfstatus:
                 return orig(node1, node2, match, listignored, listclean,
                             listunknown, listsubrepos)
 
             # some calls in this function rely on the old version of status
             self.lfstatus = False
             ctx1 = self[node1]
             ctx2 = self[node2]
             working = ctx2.rev() is None
             parentworking = working and ctx1 == self['.']
 
             if match is None:
                 match = matchmod.always(self.root, self.getcwd())
 
             wlock = None
             try:
                 try:
                     # updating the dirstate is optional
                     # so we don't wait on the lock
                     wlock = self.wlock(False)
                 except error.LockError:
                     pass
 
                 # First check if paths or patterns were specified on the
                 # command line.  If there were, and they don't match any
                 # largefiles, we should just bail here and let super
                 # handle it -- thus gaining a big performance boost.
                 lfdirstate = lfutil.openlfdirstate(ui, self)
                 if not match.always():
                     for f in lfdirstate:
                         if match(f):
                             break
                     else:
                         return orig(node1, node2, match, listignored, listclean,
                                     listunknown, listsubrepos)
 
                 # Create a copy of match that matches standins instead
                 # of largefiles.
                 def tostandins(files):
                     if not working:
                         return files
                     newfiles = []
                     dirstate = self.dirstate
                     for f in files:
                         sf = lfutil.standin(f)
                         if sf in dirstate:
                             newfiles.append(sf)
                         elif sf in dirstate.dirs():
                             # Directory entries could be regular or
                             # standin, check both
                             newfiles.extend((f, sf))
                         else:
                             newfiles.append(f)
                     return newfiles
 
                 m = copy.copy(match)
                 m._files = tostandins(m._files)
 
                 result = orig(node1, node2, m, ignored, clean, unknown,
                               listsubrepos)
                 if working:
 
                     def sfindirstate(f):
                         sf = lfutil.standin(f)
                         dirstate = self.dirstate
                         return sf in dirstate or sf in dirstate.dirs()
 
                     match._files = [f for f in match._files
                                     if sfindirstate(f)]
                     # Don't waste time getting the ignored and unknown
                     # files from lfdirstate
-                    unsure, s = lfdirstate.status(match, [], False, listclean,
-                                                  False)
+                    unsure, s = lfdirstate.status(match, subrepos=[],
+                                                  ignored=False,
+                                                  clean=listclean,
+                                                  unknown=False)
                     (modified, added, removed, deleted, clean) = (
                         s.modified, s.added, s.removed, s.deleted, s.clean)
                     if parentworking:
                         for lfile in unsure:
                             standin = lfutil.standin(lfile)
                             if standin not in ctx1:
                                 # from second parent
                                 modified.append(lfile)
                             elif lfutil.readasstandin(ctx1[standin]) \
                                     != lfutil.hashfile(self.wjoin(lfile)):
                                 modified.append(lfile)
                             else:
                                 if listclean:
                                     clean.append(lfile)
                                 lfdirstate.normal(lfile)
                     else:
                         tocheck = unsure + modified + added + clean
                         modified, added, clean = [], [], []
                         checkexec = self.dirstate._checkexec
 
                         for lfile in tocheck:
                             standin = lfutil.standin(lfile)
                             if standin in ctx1:
                                 abslfile = self.wjoin(lfile)
                                 if ((lfutil.readasstandin(ctx1[standin]) !=
                                      lfutil.hashfile(abslfile)) or
                                     (checkexec and
                                      ('x' in ctx1.flags(standin)) !=
                                      bool(lfutil.getexecutable(abslfile)))):
                                     modified.append(lfile)
                                 elif listclean:
                                     clean.append(lfile)
                             else:
                                 added.append(lfile)
 
                         # at this point, 'removed' contains largefiles
                         # marked as 'R' in the working context.
                         # then, largefiles not managed also in the target
                         # context should be excluded from 'removed'.
                         removed = [lfile for lfile in removed
                                    if lfutil.standin(lfile) in ctx1]
 
                     # Standins no longer found in lfdirstate have been deleted
                     for standin in ctx1.walk(lfutil.getstandinmatcher(self)):
                         lfile = lfutil.splitstandin(standin)
                         if not match(lfile):
                             continue
                         if lfile not in lfdirstate:
                             deleted.append(lfile)
                             # Sync "largefile has been removed" back to the
                             # standin. Removing a file as a side effect of
                             # running status is gross, but the alternatives (if
                             # any) are worse.
                             self.wvfs.unlinkpath(standin, ignoremissing=True)
 
                     # Filter result lists
                     result = list(result)
 
                     # Largefiles are not really removed when they're
                     # still in the normal dirstate. Likewise, normal
                     # files are not really removed if they are still in
                     # lfdirstate. This happens in merges where files
                     # change type.
                     removed = [f for f in removed
                                if f not in self.dirstate]
                     result[2] = [f for f in result[2]
                                  if f not in lfdirstate]
 
                     lfiles = set(lfdirstate._map)
                     # Unknown files
                     result[4] = set(result[4]).difference(lfiles)
                     # Ignored files
                     result[5] = set(result[5]).difference(lfiles)
                     # combine normal files and largefiles
                     normals = [[fn for fn in filelist
                                 if not lfutil.isstandin(fn)]
                                for filelist in result]
                     lfstatus = (modified, added, removed, deleted, [], [],
                                 clean)
                     result = [sorted(list1 + list2)
                               for (list1, list2) in zip(normals, lfstatus)]
                 else: # not against working directory
                     result = [[lfutil.splitstandin(f) or f for f in items]
                               for items in result]
 
                 if wlock:
                     lfdirstate.write()
 
             finally:
                 if wlock:
                     wlock.release()
 
             self.lfstatus = True
             return scmutil.status(*result)
 
         def commitctx(self, ctx, *args, **kwargs):
             node = super(lfilesrepo, self).commitctx(ctx, *args, **kwargs)
             class lfilesctx(ctx.__class__):
                 def markcommitted(self, node):
                     orig = super(lfilesctx, self).markcommitted
                     return lfutil.markcommitted(orig, self, node)
             ctx.__class__ = lfilesctx
             return node
 
         # Before commit, largefile standins have not had their
         # contents updated to reflect the hash of their largefile.
         # Do that here.
         def commit(self, text="", user=None, date=None, match=None,
                 force=False, editor=False, extra=None):
             if extra is None:
                 extra = {}
             orig = super(lfilesrepo, self).commit
 
             with self.wlock():
                 lfcommithook = self._lfcommithooks[-1]
                 match = lfcommithook(self, match)
                 result = orig(text=text, user=user, date=date, match=match,
                                 force=force, editor=editor, extra=extra)
                 return result
 
         def push(self, remote, force=False, revs=None, newbranch=False):
             if remote.local():
                 missing = set(self.requirements) - remote.local().supported
                 if missing:
                     msg = _("required features are not"
                             " supported in the destination:"
                             " %s") % (', '.join(sorted(missing)))
                     raise error.Abort(msg)
             return super(lfilesrepo, self).push(remote, force=force, revs=revs,
                 newbranch=newbranch)
 
         # TODO: _subdirlfs should be moved into "lfutil.py", because
         # it is referred only from "lfutil.updatestandinsbymatch"
         def _subdirlfs(self, files, lfiles):
             '''
             Adjust matched file list
             If we pass a directory to commit whose only committable files
             are largefiles, the core commit code aborts before finding
             the largefiles.
             So we do the following:
             For directories that only have largefiles as matches,
             we explicitly add the largefiles to the match list and remove
             the directory.
             In other cases, we leave the match list unmodified.
             '''
             actualfiles = []
             dirs = []
             regulars = []
 
             for f in files:
                 if lfutil.isstandin(f + '/'):
                     raise error.Abort(
                         _('file "%s" is a largefile standin') % f,
                         hint=('commit the largefile itself instead'))
                 # Scan directories
                 if self.wvfs.isdir(f):
                     dirs.append(f)
                 else:
                     regulars.append(f)
 
             for f in dirs:
                 matcheddir = False
                 d = self.dirstate.normalize(f) + '/'
                 # Check for matched normal files
                 for mf in regulars:
                     if self.dirstate.normalize(mf).startswith(d):
                         actualfiles.append(f)
                         matcheddir = True
                         break
                 if not matcheddir:
                     # If no normal match, manually append
                     # any matching largefiles
                     for lf in lfiles:
                         if self.dirstate.normalize(lf).startswith(d):
                             actualfiles.append(lf)
                             if not matcheddir:
                                 # There may still be normal files in the dir, so
                                 # add a directory to the list, which
                                 # forces status/dirstate to walk all files and
                                 # call the match function on the matcher, even
                                 # on case sensitive filesystems.
                                 actualfiles.append('.')
                                 matcheddir = True
                 # Nothing in dir, so readd it
                 # and let commit reject it
                 if not matcheddir:
                     actualfiles.append(f)
 
             # Always add normal files
             actualfiles += regulars
             return actualfiles
 
     repo.__class__ = lfilesrepo
 
     # stack of hooks being executed before committing.
     # only last element ("_lfcommithooks[-1]") is used for each committing.
     repo._lfcommithooks = [lfutil.updatestandinsbymatch]
 
     # Stack of status writer functions taking "*msg, **opts" arguments
     # like "ui.status()". Only last element ("_lfstatuswriters[-1]")
     # is used to write status out.
     repo._lfstatuswriters = [ui.status]
 
     def prepushoutgoinghook(pushop):
         """Push largefiles for pushop before pushing revisions."""
         lfrevs = pushop.lfrevs
         if lfrevs is None:
             lfrevs = pushop.outgoing.missing
         if lfrevs:
             toupload = set()
             addfunc = lambda fn, lfhash: toupload.add(lfhash)
             lfutil.getlfilestoupload(pushop.repo, lfrevs,
                                      addfunc)
             lfcommands.uploadlfiles(ui, pushop.repo, pushop.remote, toupload)
     repo.prepushoutgoinghooks.add("largefiles", prepushoutgoinghook)
 
     def checkrequireslfiles(ui, repo, **kwargs):
         if 'largefiles' not in repo.requirements and any(
                 lfutil.shortname+'/' in f[0] for f in repo.store.datafiles()):
             repo.requirements.add('largefiles')
             repo._writerequirements()
 
     ui.setconfig('hooks', 'changegroup.lfiles', checkrequireslfiles,
                  'largefiles')
     ui.setconfig('hooks', 'commit.lfiles', checkrequireslfiles, 'largefiles')

mercurial/context.py

@@ -1,2607 +1,2606 b''
 # context.py - changeset and file context objects for mercurial
 #
 # Copyright 2006, 2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import errno
 import os
 import re
 import stat
 
 from .i18n import _
 from .node import (
     addednodeid,
     bin,
     hex,
     modifiednodeid,
     nullid,
     nullrev,
     short,
     wdirid,
     wdirnodes,
     wdirrev,
 )
 from . import (
     encoding,
     error,
     fileset,
     match as matchmod,
     mdiff,
     obsolete as obsmod,
     patch,
     pathutil,
     phases,
     pycompat,
     repoview,
     revlog,
     scmutil,
     sparse,
     subrepo,
     util,
 )
 
 propertycache = util.propertycache
 
 nonascii = re.compile(r'[^\x21-\x7f]').search
 
 class basectx(object):
     """A basectx object represents the common logic for its children:
     changectx: read-only context that is already present in the repo,
     workingctx: a context that represents the working directory and can
                 be committed,
     memctx: a context that represents changes in-memory and can also
             be committed."""
     def __new__(cls, repo, changeid='', *args, **kwargs):
         if isinstance(changeid, basectx):
             return changeid
 
         o = super(basectx, cls).__new__(cls)
 
         o._repo = repo
         o._rev = nullrev
         o._node = nullid
 
         return o
 
     def __bytes__(self):
         return short(self.node())
 
     __str__ = encoding.strmethod(__bytes__)
 
     def __int__(self):
         return self.rev()
 
     def __repr__(self):
         return r"<%s %s>" % (type(self).__name__, str(self))
 
     def __eq__(self, other):
         try:
             return type(self) == type(other) and self._rev == other._rev
         except AttributeError:
             return False
 
     def __ne__(self, other):
         return not (self == other)
 
     def __contains__(self, key):
         return key in self._manifest
 
     def __getitem__(self, key):
         return self.filectx(key)
 
     def __iter__(self):
         return iter(self._manifest)
 
     def _buildstatusmanifest(self, status):
         """Builds a manifest that includes the given status results, if this is
         a working copy context. For non-working copy contexts, it just returns
         the normal manifest."""
         return self.manifest()
 
     def _matchstatus(self, other, match):
         """This internal method provides a way for child objects to override the
         match operator.
         """
         return match
 
     def _buildstatus(self, other, s, match, listignored, listclean,
                      listunknown):
         """build a status with respect to another context"""
         # Load earliest manifest first for caching reasons. More specifically,
         # if you have revisions 1000 and 1001, 1001 is probably stored as a
         # delta against 1000. Thus, if you read 1000 first, we'll reconstruct
         # 1000 and cache it so that when you read 1001, we just need to apply a
         # delta to what's in the cache. So that's one full reconstruction + one
         # delta application.
         mf2 = None
         if self.rev() is not None and self.rev() < other.rev():
             mf2 = self._buildstatusmanifest(s)
         mf1 = other._buildstatusmanifest(s)
         if mf2 is None:
             mf2 = self._buildstatusmanifest(s)
 
         modified, added = [], []
         removed = []
         clean = []
         deleted, unknown, ignored = s.deleted, s.unknown, s.ignored
         deletedset = set(deleted)
         d = mf1.diff(mf2, match=match, clean=listclean)
         for fn, value in d.iteritems():
             if fn in deletedset:
                 continue
             if value is None:
                 clean.append(fn)
                 continue
             (node1, flag1), (node2, flag2) = value
             if node1 is None:
                 added.append(fn)
             elif node2 is None:
                 removed.append(fn)
             elif flag1 != flag2:
                 modified.append(fn)
             elif node2 not in wdirnodes:
                 # When comparing files between two commits, we save time by
                 # not comparing the file contents when the nodeids differ.
                 # Note that this means we incorrectly report a reverted change
                 # to a file as a modification.
                 modified.append(fn)
             elif self[fn].cmp(other[fn]):
                 modified.append(fn)
             else:
                 clean.append(fn)
 
         if removed:
             # need to filter files if they are already reported as removed
             unknown = [fn for fn in unknown if fn not in mf1 and
                                                (not match or match(fn))]
             ignored = [fn for fn in ignored if fn not in mf1 and
                                                (not match or match(fn))]
             # if they're deleted, don't report them as removed
             removed = [fn for fn in removed if fn not in deletedset]
 
         return scmutil.status(modified, added, removed, deleted, unknown,
                               ignored, clean)
 
     @propertycache
     def substate(self):
         return subrepo.state(self, self._repo.ui)
 
     def subrev(self, subpath):
         return self.substate[subpath][1]
 
     def rev(self):
         return self._rev
     def node(self):
         return self._node
     def hex(self):
         return hex(self.node())
     def manifest(self):
         return self._manifest
     def manifestctx(self):
         return self._manifestctx
     def repo(self):
         return self._repo
     def phasestr(self):
         return phases.phasenames[self.phase()]
     def mutable(self):
         return self.phase() > phases.public
 
     def getfileset(self, expr):
         return fileset.getfileset(self, expr)
 
     def obsolete(self):
         """True if the changeset is obsolete"""
         return self.rev() in obsmod.getrevs(self._repo, 'obsolete')
 
     def extinct(self):
         """True if the changeset is extinct"""
         return self.rev() in obsmod.getrevs(self._repo, 'extinct')
 
     def unstable(self):
         msg = ("'context.unstable' is deprecated, "
                "use 'context.orphan'")
         self._repo.ui.deprecwarn(msg, '4.4')
         return self.orphan()
 
     def orphan(self):
         """True if the changeset is not obsolete but it's ancestor are"""
         return self.rev() in obsmod.getrevs(self._repo, 'orphan')
 
     def bumped(self):
         msg = ("'context.bumped' is deprecated, "
                "use 'context.phasedivergent'")
         self._repo.ui.deprecwarn(msg, '4.4')
         return self.phasedivergent()
 
     def phasedivergent(self):
         """True if the changeset try to be a successor of a public changeset
 
         Only non-public and non-obsolete changesets may be bumped.
         """
         return self.rev() in obsmod.getrevs(self._repo, 'phasedivergent')
 
     def divergent(self):
         msg = ("'context.divergent' is deprecated, "
                "use 'context.contentdivergent'")
         self._repo.ui.deprecwarn(msg, '4.4')
         return self.contentdivergent()
 
     def contentdivergent(self):
         """Is a successors of a changeset with multiple possible successors set
 
         Only non-public and non-obsolete changesets may be divergent.
         """
         return self.rev() in obsmod.getrevs(self._repo, 'contentdivergent')
 
     def troubled(self):
         msg = ("'context.troubled' is deprecated, "
                "use 'context.isunstable'")
         self._repo.ui.deprecwarn(msg, '4.4')
         return self.isunstable()
 
     def isunstable(self):
         """True if the changeset is either unstable, bumped or divergent"""
         return self.orphan() or self.phasedivergent() or self.contentdivergent()
 
     def troubles(self):
         """Keep the old version around in order to avoid breaking extensions
         about different return values.
         """
         msg = ("'context.troubles' is deprecated, "
                "use 'context.instabilities'")
         self._repo.ui.deprecwarn(msg, '4.4')
 
         troubles = []
         if self.orphan():
             troubles.append('orphan')
         if self.phasedivergent():
             troubles.append('bumped')
         if self.contentdivergent():
             troubles.append('divergent')
         return troubles
 
     def instabilities(self):
         """return the list of instabilities affecting this changeset.
 
         Instabilities are returned as strings. possible values are:
         - orphan,
         - phase-divergent,
         - content-divergent.
         """
         instabilities = []
         if self.orphan():
             instabilities.append('orphan')
         if self.phasedivergent():
             instabilities.append('phase-divergent')
         if self.contentdivergent():
             instabilities.append('content-divergent')
         return instabilities
 
     def parents(self):
         """return contexts for each parent changeset"""
         return self._parents
 
     def p1(self):
         return self._parents[0]
 
     def p2(self):
         parents = self._parents
         if len(parents) == 2:
             return parents[1]
         return changectx(self._repo, nullrev)
 
     def _fileinfo(self, path):
         if r'_manifest' in self.__dict__:
             try:
                 return self._manifest[path], self._manifest.flags(path)
             except KeyError:
                 raise error.ManifestLookupError(self._node, path,
                                                 _('not found in manifest'))
         if r'_manifestdelta' in self.__dict__ or path in self.files():
             if path in self._manifestdelta:
                 return (self._manifestdelta[path],
                         self._manifestdelta.flags(path))
         mfl = self._repo.manifestlog
         try:
             node, flag = mfl[self._changeset.manifest].find(path)
         except KeyError:
             raise error.ManifestLookupError(self._node, path,
                                             _('not found in manifest'))
 
         return node, flag
 
     def filenode(self, path):
         return self._fileinfo(path)[0]
 
     def flags(self, path):
         try:
             return self._fileinfo(path)[1]
         except error.LookupError:
             return ''
 
     def sub(self, path, allowcreate=True):
         '''return a subrepo for the stored revision of path, never wdir()'''
         return subrepo.subrepo(self, path, allowcreate=allowcreate)
 
     def nullsub(self, path, pctx):
         return subrepo.nullsubrepo(self, path, pctx)
 
     def workingsub(self, path):
         '''return a subrepo for the stored revision, or wdir if this is a wdir
         context.
         '''
         return subrepo.subrepo(self, path, allowwdir=True)
 
     def match(self, pats=None, include=None, exclude=None, default='glob',
               listsubrepos=False, badfn=None):
         r = self._repo
         return matchmod.match(r.root, r.getcwd(), pats,
                               include, exclude, default,
                               auditor=r.nofsauditor, ctx=self,
                               listsubrepos=listsubrepos, badfn=badfn)
 
     def diff(self, ctx2=None, match=None, **opts):
         """Returns a diff generator for the given contexts and matcher"""
         if ctx2 is None:
             ctx2 = self.p1()
         if ctx2 is not None:
             ctx2 = self._repo[ctx2]
         diffopts = patch.diffopts(self._repo.ui, opts)
         return patch.diff(self._repo, ctx2, self, match=match, opts=diffopts)
 
     def dirs(self):
         return self._manifest.dirs()
 
     def hasdir(self, dir):
         return self._manifest.hasdir(dir)
 
     def status(self, other=None, match=None, listignored=False,
                listclean=False, listunknown=False, listsubrepos=False):
         """return status of files between two nodes or node and working
         directory.
 
         If other is None, compare this node with working directory.
 
         returns (modified, added, removed, deleted, unknown, ignored, clean)
         """
 
         ctx1 = self
         ctx2 = self._repo[other]
 
         # This next code block is, admittedly, fragile logic that tests for
         # reversing the contexts and wouldn't need to exist if it weren't for
         # the fast (and common) code path of comparing the working directory
         # with its first parent.
         #
         # What we're aiming for here is the ability to call:
         #
         # workingctx.status(parentctx)
         #
         # If we always built the manifest for each context and compared those,
         # then we'd be done. But the special case of the above call means we
         # just copy the manifest of the parent.
         reversed = False
         if (not isinstance(ctx1, changectx)
             and isinstance(ctx2, changectx)):
             reversed = True
             ctx1, ctx2 = ctx2, ctx1
 
         match = match or matchmod.always(self._repo.root, self._repo.getcwd())
         match = ctx2._matchstatus(ctx1, match)
         r = scmutil.status([], [], [], [], [], [], [])
         r = ctx2._buildstatus(ctx1, r, match, listignored, listclean,
                               listunknown)
 
         if reversed:
             # Reverse added and removed. Clear deleted, unknown and ignored as
             # these make no sense to reverse.
             r = scmutil.status(r.modified, r.removed, r.added, [], [], [],
                                r.clean)
 
         if listsubrepos:
             for subpath, sub in scmutil.itersubrepos(ctx1, ctx2):
                 try:
                     rev2 = ctx2.subrev(subpath)
                 except KeyError:
                     # A subrepo that existed in node1 was deleted between
                     # node1 and node2 (inclusive). Thus, ctx2's substate
                     # won't contain that subpath. The best we can do ignore it.
                     rev2 = None
                 submatch = matchmod.subdirmatcher(subpath, match)
                 s = sub.status(rev2, match=submatch, ignored=listignored,
                                clean=listclean, unknown=listunknown,
                                listsubrepos=True)
                 for rfiles, sfiles in zip(r, s):
                     rfiles.extend("%s/%s" % (subpath, f) for f in sfiles)
 
         for l in r:
             l.sort()
 
         return r
 
 def _filterederror(repo, changeid):
     """build an exception to be raised about a filtered changeid
 
     This is extracted in a function to help extensions (eg: evolve) to
     experiment with various message variants."""
     if repo.filtername.startswith('visible'):
         msg = _("hidden revision '%s'") % changeid
         hint = _('use --hidden to access hidden revisions')
         return error.FilteredRepoLookupError(msg, hint=hint)
     msg = _("filtered revision '%s' (not in '%s' subset)")
     msg %= (changeid, repo.filtername)
     return error.FilteredRepoLookupError(msg)
 
 class changectx(basectx):
     """A changecontext object makes access to data related to a particular
     changeset convenient. It represents a read-only context already present in
     the repo."""
     def __init__(self, repo, changeid=''):
         """changeid is a revision number, node, or tag"""
 
         # since basectx.__new__ already took care of copying the object, we
         # don't need to do anything in __init__, so we just exit here
         if isinstance(changeid, basectx):
             return
 
         if changeid == '':
             changeid = '.'
         self._repo = repo
 
         try:
             if isinstance(changeid, int):
                 self._node = repo.changelog.node(changeid)
                 self._rev = changeid
                 return
             if not pycompat.ispy3 and isinstance(changeid, long):
                 changeid = str(changeid)
             if changeid == 'null':
                 self._node = nullid
                 self._rev = nullrev
                 return
             if changeid == 'tip':
                 self._node = repo.changelog.tip()
                 self._rev = repo.changelog.rev(self._node)
                 return
             if changeid == '.' or changeid == repo.dirstate.p1():
                 # this is a hack to delay/avoid loading obsmarkers
                 # when we know that '.' won't be hidden
                 self._node = repo.dirstate.p1()
                 self._rev = repo.unfiltered().changelog.rev(self._node)
                 return
             if len(changeid) == 20:
                 try:
                     self._node = changeid
                     self._rev = repo.changelog.rev(changeid)
                     return
                 except error.FilteredRepoLookupError:
                     raise
                 except LookupError:
                     pass
 
             try:
                 r = int(changeid)
                 if '%d' % r != changeid:
                     raise ValueError
                 l = len(repo.changelog)
                 if r < 0:
                     r += l
                 if r < 0 or r >= l and r != wdirrev:
                     raise ValueError
                 self._rev = r
                 self._node = repo.changelog.node(r)
                 return
             except error.FilteredIndexError:
                 raise
             except (ValueError, OverflowError, IndexError):
                 pass
 
             if len(changeid) == 40:
                 try:
                     self._node = bin(changeid)
                     self._rev = repo.changelog.rev(self._node)
                     return
                 except error.FilteredLookupError:
                     raise
                 except (TypeError, LookupError):
                     pass
 
             # lookup bookmarks through the name interface
             try:
                 self._node = repo.names.singlenode(repo, changeid)
                 self._rev = repo.changelog.rev(self._node)
                 return
             except KeyError:
                 pass
             except error.FilteredRepoLookupError:
                 raise
             except error.RepoLookupError:
                 pass
 
             self._node = repo.unfiltered().changelog._partialmatch(changeid)
             if self._node is not None:
                 self._rev = repo.changelog.rev(self._node)
                 return
 
             # lookup failed
             # check if it might have come from damaged dirstate
             #
             # XXX we could avoid the unfiltered if we had a recognizable
             # exception for filtered changeset access
             if changeid in repo.unfiltered().dirstate.parents():
                 msg = _("working directory has unknown parent '%s'!")
                 raise error.Abort(msg % short(changeid))
             try:
                 if len(changeid) == 20 and nonascii(changeid):
                     changeid = hex(changeid)
             except TypeError:
                 pass
         except (error.FilteredIndexError, error.FilteredLookupError,
                 error.FilteredRepoLookupError):
             raise _filterederror(repo, changeid)
         except IndexError:
             pass
         raise error.RepoLookupError(
             _("unknown revision '%s'") % changeid)
 
     def __hash__(self):
         try:
             return hash(self._rev)
         except AttributeError:
             return id(self)
 
     def __nonzero__(self):
         return self._rev != nullrev
 
     __bool__ = __nonzero__
 
     @propertycache
     def _changeset(self):
         return self._repo.changelog.changelogrevision(self.rev())
 
     @propertycache
     def _manifest(self):
         return self._manifestctx.read()
 
     @property
     def _manifestctx(self):
         return self._repo.manifestlog[self._changeset.manifest]
 
     @propertycache
     def _manifestdelta(self):
         return self._manifestctx.readdelta()
 
     @propertycache
     def _parents(self):
         repo = self._repo
         p1, p2 = repo.changelog.parentrevs(self._rev)
         if p2 == nullrev:
             return [changectx(repo, p1)]
         return [changectx(repo, p1), changectx(repo, p2)]
 
     def changeset(self):
         c = self._changeset
         return (
             c.manifest,
             c.user,
             c.date,
             c.files,
             c.description,
             c.extra,
         )
     def manifestnode(self):
         return self._changeset.manifest
 
     def user(self):
         return self._changeset.user
     def date(self):
         return self._changeset.date
     def files(self):
         return self._changeset.files
     def description(self):
         return self._changeset.description
     def branch(self):
         return encoding.tolocal(self._changeset.extra.get("branch"))
     def closesbranch(self):
         return 'close' in self._changeset.extra
     def extra(self):
         return self._changeset.extra
     def tags(self):
         return self._repo.nodetags(self._node)
     def bookmarks(self):
         return self._repo.nodebookmarks(self._node)
     def phase(self):
         return self._repo._phasecache.phase(self._repo, self._rev)
     def hidden(self):
         return self._rev in repoview.filterrevs(self._repo, 'visible')
 
     def children(self):
         """return contexts for each child changeset"""
         c = self._repo.changelog.children(self._node)
         return [changectx(self._repo, x) for x in c]
 
     def ancestors(self):
         for a in self._repo.changelog.ancestors([self._rev]):
             yield changectx(self._repo, a)
 
     def descendants(self):
         for d in self._repo.changelog.descendants([self._rev]):
             yield changectx(self._repo, d)
 
     def filectx(self, path, fileid=None, filelog=None):
         """get a file context from this changeset"""
         if fileid is None:
             fileid = self.filenode(path)
         return filectx(self._repo, path, fileid=fileid,
                        changectx=self, filelog=filelog)
 
     def ancestor(self, c2, warn=False):
         """return the "best" ancestor context of self and c2
 
         If there are multiple candidates, it will show a message and check
         merge.preferancestor configuration before falling back to the
         revlog ancestor."""
         # deal with workingctxs
         n2 = c2._node
         if n2 is None:
             n2 = c2._parents[0]._node
         cahs = self._repo.changelog.commonancestorsheads(self._node, n2)
         if not cahs:
             anc = nullid
         elif len(cahs) == 1:
             anc = cahs[0]
         else:
             # experimental config: merge.preferancestor
             for r in self._repo.ui.configlist('merge', 'preferancestor', ['*']):
                 try:
                     ctx = changectx(self._repo, r)
                 except error.RepoLookupError:
                     continue
                 anc = ctx.node()
                 if anc in cahs:
                     break
             else:
                 anc = self._repo.changelog.ancestor(self._node, n2)
             if warn:
                 self._repo.ui.status(
                     (_("note: using %s as ancestor of %s and %s\n") %
                      (short(anc), short(self._node), short(n2))) +
                     ''.join(_("      alternatively, use --config "
                               "merge.preferancestor=%s\n") %
                             short(n) for n in sorted(cahs) if n != anc))
         return changectx(self._repo, anc)
 
     def descendant(self, other):
         """True if other is descendant of this changeset"""
         return self._repo.changelog.descendant(self._rev, other._rev)
 
     def walk(self, match):
         '''Generates matching file names.'''
 
         # Wrap match.bad method to have message with nodeid
         def bad(fn, msg):
             # The manifest doesn't know about subrepos, so don't complain about
             # paths into valid subrepos.
             if any(fn == s or fn.startswith(s + '/')
                    for s in self.substate):
                 return
             match.bad(fn, _('no such file in rev %s') % self)
 
         m = matchmod.badmatch(match, bad)
         return self._manifest.walk(m)
 
     def matches(self, match):
         return self.walk(match)
 
 class basefilectx(object):
     """A filecontext object represents the common logic for its children:
     filectx: read-only access to a filerevision that is already present
              in the repo,
     workingfilectx: a filecontext that represents files from the working
                     directory,
     memfilectx: a filecontext that represents files in-memory,
     overlayfilectx: duplicate another filecontext with some fields overridden.
     """
     @propertycache
     def _filelog(self):
         return self._repo.file(self._path)
 
     @propertycache
     def _changeid(self):
         if r'_changeid' in self.__dict__:
             return self._changeid
         elif r'_changectx' in self.__dict__:
             return self._changectx.rev()
         elif r'_descendantrev' in self.__dict__:
             # this file context was created from a revision with a known
             # descendant, we can (lazily) correct for linkrev aliases
             return self._adjustlinkrev(self._descendantrev)
         else:
             return self._filelog.linkrev(self._filerev)
 
     @propertycache
     def _filenode(self):
         if r'_fileid' in self.__dict__:
             return self._filelog.lookup(self._fileid)
         else:
             return self._changectx.filenode(self._path)
 
     @propertycache
     def _filerev(self):
         return self._filelog.rev(self._filenode)
 
     @propertycache
     def _repopath(self):
         return self._path
 
     def __nonzero__(self):
         try:
             self._filenode
             return True
         except error.LookupError:
             # file is missing
             return False
 
     __bool__ = __nonzero__
 
     def __bytes__(self):
         try:
             return "%s@%s" % (self.path(), self._changectx)
         except error.LookupError:
             return "%s@???" % self.path()
 
     __str__ = encoding.strmethod(__bytes__)
 
     def __repr__(self):
         return "<%s %s>" % (type(self).__name__, str(self))
 
     def __hash__(self):
         try:
             return hash((self._path, self._filenode))
         except AttributeError:
             return id(self)
 
     def __eq__(self, other):
         try:
             return (type(self) == type(other) and self._path == other._path
                     and self._filenode == other._filenode)
         except AttributeError:
             return False
 
     def __ne__(self, other):
         return not (self == other)
 
     def filerev(self):
         return self._filerev
     def filenode(self):
         return self._filenode
     @propertycache
     def _flags(self):
         return self._changectx.flags(self._path)
     def flags(self):
         return self._flags
     def filelog(self):
         return self._filelog
     def rev(self):
         return self._changeid
     def linkrev(self):
         return self._filelog.linkrev(self._filerev)
     def node(self):
         return self._changectx.node()
     def hex(self):
         return self._changectx.hex()
     def user(self):
         return self._changectx.user()
     def date(self):
         return self._changectx.date()
     def files(self):
         return self._changectx.files()
     def description(self):
         return self._changectx.description()
     def branch(self):
         return self._changectx.branch()
     def extra(self):
         return self._changectx.extra()
     def phase(self):
         return self._changectx.phase()
     def phasestr(self):
         return self._changectx.phasestr()
     def manifest(self):
         return self._changectx.manifest()
     def changectx(self):
         return self._changectx
     def renamed(self):
         return self._copied
     def repo(self):
         return self._repo
     def size(self):
         return len(self.data())
 
     def path(self):
         return self._path
 
     def isbinary(self):
         try:
             return util.binary(self.data())
         except IOError:
             return False
     def isexec(self):
         return 'x' in self.flags()
     def islink(self):
         return 'l' in self.flags()
 
     def isabsent(self):
         """whether this filectx represents a file not in self._changectx
 
         This is mainly for merge code to detect change/delete conflicts. This is
         expected to be True for all subclasses of basectx."""
         return False
 
     _customcmp = False
     def cmp(self, fctx):
         """compare with other file context
 
         returns True if different than fctx.
         """
         if fctx._customcmp:
             return fctx.cmp(self)
 
         if (fctx._filenode is None
             and (self._repo._encodefilterpats
                  # if file data starts with '\1\n', empty metadata block is
                  # prepended, which adds 4 bytes to filelog.size().
                  or self.size() - 4 == fctx.size())
             or self.size() == fctx.size()):
             return self._filelog.cmp(self._filenode, fctx.data())
 
         return True
 
     def _adjustlinkrev(self, srcrev, inclusive=False):
         """return the first ancestor of <srcrev> introducing <fnode>
 
         If the linkrev of the file revision does not point to an ancestor of
         srcrev, we'll walk down the ancestors until we find one introducing
         this file revision.
 
         :srcrev: the changeset revision we search ancestors from
         :inclusive: if true, the src revision will also be checked
         """
         repo = self._repo
         cl = repo.unfiltered().changelog
         mfl = repo.manifestlog
         # fetch the linkrev
         lkr = self.linkrev()
         # hack to reuse ancestor computation when searching for renames
         memberanc = getattr(self, '_ancestrycontext', None)
         iteranc = None
         if srcrev is None:
             # wctx case, used by workingfilectx during mergecopy
             revs = [p.rev() for p in self._repo[None].parents()]
             inclusive = True # we skipped the real (revless) source
         else:
             revs = [srcrev]
         if memberanc is None:
             memberanc = iteranc = cl.ancestors(revs, lkr,
                                                inclusive=inclusive)
         # check if this linkrev is an ancestor of srcrev
         if lkr not in memberanc:
             if iteranc is None:
                 iteranc = cl.ancestors(revs, lkr, inclusive=inclusive)
             fnode = self._filenode
             path = self._path
             for a in iteranc:
                 ac = cl.read(a) # get changeset data (we avoid object creation)
                 if path in ac[3]: # checking the 'files' field.
                     # The file has been touched, check if the content is
                     # similar to the one we search for.
                     if fnode == mfl[ac[0]].readfast().get(path):
                         return a
             # In theory, we should never get out of that loop without a result.
             # But if manifest uses a buggy file revision (not children of the
             # one it replaces) we could. Such a buggy situation will likely
             # result is crash somewhere else at to some point.
         return lkr
 
     def introrev(self):
         """return the rev of the changeset which introduced this file revision
 
         This method is different from linkrev because it take into account the
         changeset the filectx was created from. It ensures the returned
         revision is one of its ancestors. This prevents bugs from
         'linkrev-shadowing' when a file revision is used by multiple
         changesets.
         """
         lkr = self.linkrev()
         attrs = vars(self)
         noctx = not ('_changeid' in attrs or '_changectx' in attrs)
         if noctx or self.rev() == lkr:
             return self.linkrev()
         return self._adjustlinkrev(self.rev(), inclusive=True)
 
     def _parentfilectx(self, path, fileid, filelog):
         """create parent filectx keeping ancestry info for _adjustlinkrev()"""
         fctx = filectx(self._repo, path, fileid=fileid, filelog=filelog)
         if '_changeid' in vars(self) or '_changectx' in vars(self):
             # If self is associated with a changeset (probably explicitly
             # fed), ensure the created filectx is associated with a
             # changeset that is an ancestor of self.changectx.
             # This lets us later use _adjustlinkrev to get a correct link.
             fctx._descendantrev = self.rev()
             fctx._ancestrycontext = getattr(self, '_ancestrycontext', None)
         elif '_descendantrev' in vars(self):
             # Otherwise propagate _descendantrev if we have one associated.
             fctx._descendantrev = self._descendantrev
             fctx._ancestrycontext = getattr(self, '_ancestrycontext', None)
         return fctx
 
     def parents(self):
         _path = self._path
         fl = self._filelog
         parents = self._filelog.parents(self._filenode)
         pl = [(_path, node, fl) for node in parents if node != nullid]
 
         r = fl.renamed(self._filenode)
         if r:
             # - In the simple rename case, both parent are nullid, pl is empty.
             # - In case of merge, only one of the parent is null id and should
             # be replaced with the rename information. This parent is -always-
             # the first one.
             #
             # As null id have always been filtered out in the previous list
             # comprehension, inserting to 0 will always result in "replacing
             # first nullid parent with rename information.
             pl.insert(0, (r[0], r[1], self._repo.file(r[0])))
 
         return [self._parentfilectx(path, fnode, l) for path, fnode, l in pl]
 
     def p1(self):
         return self.parents()[0]
 
     def p2(self):
         p = self.parents()
         if len(p) == 2:
             return p[1]
         return filectx(self._repo, self._path, fileid=-1, filelog=self._filelog)
 
     def annotate(self, follow=False, linenumber=False, skiprevs=None,
                  diffopts=None):
         '''returns a list of tuples of ((ctx, number), line) for each line
         in the file, where ctx is the filectx of the node where
         that line was last changed; if linenumber parameter is true, number is
         the line number at the first appearance in the managed file, otherwise,
         number has a fixed value of False.
         '''
 
         def lines(text):
             if text.endswith("\n"):
                 return text.count("\n")
             return text.count("\n") + int(bool(text))
 
         if linenumber:
             def decorate(text, rev):
                 return ([(rev, i) for i in xrange(1, lines(text) + 1)], text)
         else:
             def decorate(text, rev):
                 return ([(rev, False)] * lines(text), text)
 
         getlog = util.lrucachefunc(lambda x: self._repo.file(x))
 
         def parents(f):
             # Cut _descendantrev here to mitigate the penalty of lazy linkrev
             # adjustment. Otherwise, p._adjustlinkrev() would walk changelog
             # from the topmost introrev (= srcrev) down to p.linkrev() if it
             # isn't an ancestor of the srcrev.
             f._changeid
             pl = f.parents()
 
             # Don't return renamed parents if we aren't following.
             if not follow:
                 pl = [p for p in pl if p.path() == f.path()]
 
             # renamed filectx won't have a filelog yet, so set it
             # from the cache to save time
             for p in pl:
                 if not '_filelog' in p.__dict__:
                     p._filelog = getlog(p.path())
 
             return pl
 
         # use linkrev to find the first changeset where self appeared
         base = self
         introrev = self.introrev()
         if self.rev() != introrev:
             base = self.filectx(self.filenode(), changeid=introrev)
         if getattr(base, '_ancestrycontext', None) is None:
             cl = self._repo.changelog
             if introrev is None:
                 # wctx is not inclusive, but works because _ancestrycontext
                 # is used to test filelog revisions
                 ac = cl.ancestors([p.rev() for p in base.parents()],
                                   inclusive=True)
             else:
                 ac = cl.ancestors([introrev], inclusive=True)
             base._ancestrycontext = ac
 
         # This algorithm would prefer to be recursive, but Python is a
         # bit recursion-hostile. Instead we do an iterative
         # depth-first search.
 
         # 1st DFS pre-calculates pcache and needed
         visit = [base]
         pcache = {}
         needed = {base: 1}
         while visit:
             f = visit.pop()
             if f in pcache:
                 continue
             pl = parents(f)
             pcache[f] = pl
             for p in pl:
                 needed[p] = needed.get(p, 0) + 1
                 if p not in pcache:
                     visit.append(p)
 
         # 2nd DFS does the actual annotate
         visit[:] = [base]
         hist = {}
         while visit:
             f = visit[-1]
             if f in hist:
                 visit.pop()
                 continue
 
             ready = True
             pl = pcache[f]
             for p in pl:
                 if p not in hist:
                     ready = False
                     visit.append(p)
             if ready:
                 visit.pop()
                 curr = decorate(f.data(), f)
                 skipchild = False
                 if skiprevs is not None:
                     skipchild = f._changeid in skiprevs
                 curr = _annotatepair([hist[p] for p in pl], f, curr, skipchild,
                                      diffopts)
                 for p in pl:
                     if needed[p] == 1:
                         del hist[p]
                         del needed[p]
                     else:
                         needed[p] -= 1
 
                 hist[f] = curr
                 del pcache[f]
 
         return zip(hist[base][0], hist[base][1].splitlines(True))
 
     def ancestors(self, followfirst=False):
         visit = {}
         c = self
         if followfirst:
             cut = 1
         else:
             cut = None
 
         while True:
             for parent in c.parents()[:cut]:
                 visit[(parent.linkrev(), parent.filenode())] = parent
             if not visit:
                 break
             c = visit.pop(max(visit))
             yield c
 
     def decodeddata(self):
         """Returns `data()` after running repository decoding filters.
 
         This is often equivalent to how the data would be expressed on disk.
         """
         return self._repo.wwritedata(self.path(), self.data())
 
 def _annotatepair(parents, childfctx, child, skipchild, diffopts):
     r'''
     Given parent and child fctxes and annotate data for parents, for all lines
     in either parent that match the child, annotate the child with the parent's
     data.
 
     Additionally, if `skipchild` is True, replace all other lines with parent
     annotate data as well such that child is never blamed for any lines.
 
     >>> oldfctx = b'old'
     >>> p1fctx, p2fctx, childfctx = b'p1', b'p2', b'c'
     >>> olddata = b'a\nb\n'
     >>> p1data = b'a\nb\nc\n'
     >>> p2data = b'a\nc\nd\n'
     >>> childdata = b'a\nb2\nc\nc2\nd\n'
     >>> diffopts = mdiff.diffopts()
 
     >>> def decorate(text, rev):
     ...     return ([(rev, i) for i in xrange(1, text.count(b'\n') + 1)], text)
 
     Basic usage:
 
     >>> oldann = decorate(olddata, oldfctx)
     >>> p1ann = decorate(p1data, p1fctx)
     >>> p1ann = _annotatepair([oldann], p1fctx, p1ann, False, diffopts)
     >>> p1ann[0]
     [('old', 1), ('old', 2), ('p1', 3)]
     >>> p2ann = decorate(p2data, p2fctx)
     >>> p2ann = _annotatepair([oldann], p2fctx, p2ann, False, diffopts)
     >>> p2ann[0]
     [('old', 1), ('p2', 2), ('p2', 3)]
 
     Test with multiple parents (note the difference caused by ordering):
 
     >>> childann = decorate(childdata, childfctx)
     >>> childann = _annotatepair([p1ann, p2ann], childfctx, childann, False,
     ...                          diffopts)
     >>> childann[0]
     [('old', 1), ('c', 2), ('p2', 2), ('c', 4), ('p2', 3)]
 
     >>> childann = decorate(childdata, childfctx)
     >>> childann = _annotatepair([p2ann, p1ann], childfctx, childann, False,
     ...                          diffopts)
     >>> childann[0]
     [('old', 1), ('c', 2), ('p1', 3), ('c', 4), ('p2', 3)]
 
     Test with skipchild (note the difference caused by ordering):
 
     >>> childann = decorate(childdata, childfctx)
     >>> childann = _annotatepair([p1ann, p2ann], childfctx, childann, True,
     ...                          diffopts)
     >>> childann[0]
     [('old', 1), ('old', 2), ('p2', 2), ('p2', 2), ('p2', 3)]
 
     >>> childann = decorate(childdata, childfctx)
     >>> childann = _annotatepair([p2ann, p1ann], childfctx, childann, True,
     ...                          diffopts)
     >>> childann[0]
     [('old', 1), ('old', 2), ('p1', 3), ('p1', 3), ('p2', 3)]
     '''
     pblocks = [(parent, mdiff.allblocks(parent[1], child[1], opts=diffopts))
                for parent in parents]
 
     if skipchild:
         # Need to iterate over the blocks twice -- make it a list
         pblocks = [(p, list(blocks)) for (p, blocks) in pblocks]
     # Mercurial currently prefers p2 over p1 for annotate.
     # TODO: change this?
     for parent, blocks in pblocks:
         for (a1, a2, b1, b2), t in blocks:
             # Changed blocks ('!') or blocks made only of blank lines ('~')
             # belong to the child.
             if t == '=':
                 child[0][b1:b2] = parent[0][a1:a2]
 
     if skipchild:
         # Now try and match up anything that couldn't be matched,
         # Reversing pblocks maintains bias towards p2, matching above
         # behavior.
         pblocks.reverse()
 
         # The heuristics are:
         # * Work on blocks of changed lines (effectively diff hunks with -U0).
         # This could potentially be smarter but works well enough.
         # * For a non-matching section, do a best-effort fit. Match lines in
         #   diff hunks 1:1, dropping lines as necessary.
         # * Repeat the last line as a last resort.
 
         # First, replace as much as possible without repeating the last line.
         remaining = [(parent, []) for parent, _blocks in pblocks]
         for idx, (parent, blocks) in enumerate(pblocks):
             for (a1, a2, b1, b2), _t in blocks:
                 if a2 - a1 >= b2 - b1:
                     for bk in xrange(b1, b2):
                         if child[0][bk][0] == childfctx:
                             ak = min(a1 + (bk - b1), a2 - 1)
                             child[0][bk] = parent[0][ak]
                 else:
                     remaining[idx][1].append((a1, a2, b1, b2))
 
         # Then, look at anything left, which might involve repeating the last
         # line.
         for parent, blocks in remaining:
             for a1, a2, b1, b2 in blocks:
                 for bk in xrange(b1, b2):
                     if child[0][bk][0] == childfctx:
                         ak = min(a1 + (bk - b1), a2 - 1)
                         child[0][bk] = parent[0][ak]
     return child
 
 class filectx(basefilectx):
     """A filecontext object makes access to data related to a particular
        filerevision convenient."""
     def __init__(self, repo, path, changeid=None, fileid=None,
                  filelog=None, changectx=None):
         """changeid can be a changeset revision, node, or tag.
            fileid can be a file revision or node."""
         self._repo = repo
         self._path = path
 
         assert (changeid is not None
                 or fileid is not None
                 or changectx is not None), \
                 ("bad args: changeid=%r, fileid=%r, changectx=%r"
                  % (changeid, fileid, changectx))
 
         if filelog is not None:
             self._filelog = filelog
 
         if changeid is not None:
             self._changeid = changeid
         if changectx is not None:
             self._changectx = changectx
         if fileid is not None:
             self._fileid = fileid
 
     @propertycache
     def _changectx(self):
         try:
             return changectx(self._repo, self._changeid)
         except error.FilteredRepoLookupError:
             # Linkrev may point to any revision in the repository.  When the
             # repository is filtered this may lead to `filectx` trying to build
             # `changectx` for filtered revision. In such case we fallback to
             # creating `changectx` on the unfiltered version of the reposition.
             # This fallback should not be an issue because `changectx` from
             # `filectx` are not used in complex operations that care about
             # filtering.
             #
             # This fallback is a cheap and dirty fix that prevent several
             # crashes. It does not ensure the behavior is correct. However the
             # behavior was not correct before filtering either and "incorrect
             # behavior" is seen as better as "crash"
             #
             # Linkrevs have several serious troubles with filtering that are
             # complicated to solve. Proper handling of the issue here should be
             # considered when solving linkrev issue are on the table.
             return changectx(self._repo.unfiltered(), self._changeid)
 
     def filectx(self, fileid, changeid=None):
         '''opens an arbitrary revision of the file without
         opening a new filelog'''
         return filectx(self._repo, self._path, fileid=fileid,
                        filelog=self._filelog, changeid=changeid)
 
     def rawdata(self):
         return self._filelog.revision(self._filenode, raw=True)
 
     def rawflags(self):
         """low-level revlog flags"""
         return self._filelog.flags(self._filerev)
 
     def data(self):
         try:
             return self._filelog.read(self._filenode)
         except error.CensoredNodeError:
             if self._repo.ui.config("censor", "policy") == "ignore":
                 return ""
             raise error.Abort(_("censored node: %s") % short(self._filenode),
                              hint=_("set censor.policy to ignore errors"))
 
     def size(self):
         return self._filelog.size(self._filerev)
 
     @propertycache
     def _copied(self):
         """check if file was actually renamed in this changeset revision
 
         If rename logged in file revision, we report copy for changeset only
         if file revisions linkrev points back to the changeset in question
         or both changeset parents contain different file revisions.
         """
 
         renamed = self._filelog.renamed(self._filenode)
         if not renamed:
             return renamed
 
         if self.rev() == self.linkrev():
             return renamed
 
         name = self.path()
         fnode = self._filenode
         for p in self._changectx.parents():
             try:
                 if fnode == p.filenode(name):
                     return None
             except error.LookupError:
                 pass
         return renamed
 
     def children(self):
         # hard for renames
         c = self._filelog.children(self._filenode)
         return [filectx(self._repo, self._path, fileid=x,
                         filelog=self._filelog) for x in c]
 
 class committablectx(basectx):
     """A committablectx object provides common functionality for a context that
     wants the ability to commit, e.g. workingctx or memctx."""
     def __init__(self, repo, text="", user=None, date=None, extra=None,
                  changes=None):
         self._repo = repo
         self._rev = None
         self._node = None
         self._text = text
         if date:
             self._date = util.parsedate(date)
         if user:
             self._user = user
         if changes:
             self._status = changes
 
         self._extra = {}
         if extra:
             self._extra = extra.copy()
         if 'branch' not in self._extra:
             try:
                 branch = encoding.fromlocal(self._repo.dirstate.branch())
             except UnicodeDecodeError:
                 raise error.Abort(_('branch name not in UTF-8!'))
             self._extra['branch'] = branch
         if self._extra['branch'] == '':
             self._extra['branch'] = 'default'
 
     def __bytes__(self):
         return bytes(self._parents[0]) + "+"
 
     __str__ = encoding.strmethod(__bytes__)
 
     def __nonzero__(self):
         return True
 
     __bool__ = __nonzero__
 
     def _buildflagfunc(self):
         # Create a fallback function for getting file flags when the
         # filesystem doesn't support them
 
         copiesget = self._repo.dirstate.copies().get
         parents = self.parents()
         if len(parents) < 2:
             # when we have one parent, it's easy: copy from parent
             man = parents[0].manifest()
             def func(f):
                 f = copiesget(f, f)
                 return man.flags(f)
         else:
             # merges are tricky: we try to reconstruct the unstored
             # result from the merge (issue1802)
             p1, p2 = parents
             pa = p1.ancestor(p2)
             m1, m2, ma = p1.manifest(), p2.manifest(), pa.manifest()
 
             def func(f):
                 f = copiesget(f, f) # may be wrong for merges with copies
                 fl1, fl2, fla = m1.flags(f), m2.flags(f), ma.flags(f)
                 if fl1 == fl2:
                     return fl1
                 if fl1 == fla:
                     return fl2
                 if fl2 == fla:
                     return fl1
                 return '' # punt for conflicts
 
         return func
 
     @propertycache
     def _flagfunc(self):
         return self._repo.dirstate.flagfunc(self._buildflagfunc)
 
     @propertycache
     def _status(self):
         return self._repo.status()
 
     @propertycache
     def _user(self):
         return self._repo.ui.username()
 
     @propertycache
     def _date(self):
         ui = self._repo.ui
         date = ui.configdate('devel', 'default-date')
         if date is None:
             date = util.makedate()
         return date
 
     def subrev(self, subpath):
         return None
 
     def manifestnode(self):
         return None
     def user(self):
         return self._user or self._repo.ui.username()
     def date(self):
         return self._date
     def description(self):
         return self._text
     def files(self):
         return sorted(self._status.modified + self._status.added +
                       self._status.removed)
 
     def modified(self):
         return self._status.modified
     def added(self):
         return self._status.added
     def removed(self):
         return self._status.removed
     def deleted(self):
         return self._status.deleted
     def branch(self):
         return encoding.tolocal(self._extra['branch'])
     def closesbranch(self):
         return 'close' in self._extra
     def extra(self):
         return self._extra
 
     def tags(self):
         return []
 
     def bookmarks(self):
         b = []
         for p in self.parents():
             b.extend(p.bookmarks())
         return b
 
     def phase(self):
         phase = phases.draft # default phase to draft
         for p in self.parents():
             phase = max(phase, p.phase())
         return phase
 
     def hidden(self):
         return False
 
     def children(self):
         return []
 
     def flags(self, path):
         if r'_manifest' in self.__dict__:
             try:
                 return self._manifest.flags(path)
             except KeyError:
                 return ''
 
         try:
             return self._flagfunc(path)
         except OSError:
             return ''
 
     def ancestor(self, c2):
         """return the "best" ancestor context of self and c2"""
         return self._parents[0].ancestor(c2) # punt on two parents for now
 
     def walk(self, match):
         '''Generates matching file names.'''
         return sorted(self._repo.dirstate.walk(match,
                                                subrepos=sorted(self.substate),
                                                unknown=True, ignored=False))
 
     def matches(self, match):
         return sorted(self._repo.dirstate.matches(match))
 
     def ancestors(self):
         for p in self._parents:
             yield p
         for a in self._repo.changelog.ancestors(
             [p.rev() for p in self._parents]):
             yield changectx(self._repo, a)
 
     def markcommitted(self, node):
         """Perform post-commit cleanup necessary after committing this ctx
 
         Specifically, this updates backing stores this working context
         wraps to reflect the fact that the changes reflected by this
         workingctx have been committed.  For example, it marks
         modified and added files as normal in the dirstate.
 
         """
 
         with self._repo.dirstate.parentchange():
             for f in self.modified() + self.added():
                 self._repo.dirstate.normal(f)
             for f in self.removed():
                 self._repo.dirstate.drop(f)
             self._repo.dirstate.setparents(node)
 
         # write changes out explicitly, because nesting wlock at
         # runtime may prevent 'wlock.release()' in 'repo.commit()'
         # from immediately doing so for subsequent changing files
         self._repo.dirstate.write(self._repo.currenttransaction())
 
     def dirty(self, missing=False, merge=True, branch=True):
         return False
 
 class workingctx(committablectx):
     """A workingctx object makes access to data related to
     the current working directory convenient.
     date - any valid date string or (unixtime, offset), or None.
     user - username string, or None.
     extra - a dictionary of extra values, or None.
     changes - a list of file lists as returned by localrepo.status()
                or None to use the repository status.
     """
     def __init__(self, repo, text="", user=None, date=None, extra=None,
                  changes=None):
         super(workingctx, self).__init__(repo, text, user, date, extra, changes)
 
     def __iter__(self):
         d = self._repo.dirstate
         for f in d:
             if d[f] != 'r':
                 yield f
 
     def __contains__(self, key):
         return self._repo.dirstate[key] not in "?r"
 
     def hex(self):
         return hex(wdirid)
 
     @propertycache
     def _parents(self):
         p = self._repo.dirstate.parents()
         if p[1] == nullid:
             p = p[:-1]
         return [changectx(self._repo, x) for x in p]
 
     def filectx(self, path, filelog=None):
         """get a file context from the working directory"""
         return workingfilectx(self._repo, path, workingctx=self,
                               filelog=filelog)
 
     def dirty(self, missing=False, merge=True, branch=True):
         "check whether a working directory is modified"
         # check subrepos first
         for s in sorted(self.substate):
             if self.sub(s).dirty(missing=missing):
                 return True
         # check current working dir
         return ((merge and self.p2()) or
                 (branch and self.branch() != self.p1().branch()) or
                 self.modified() or self.added() or self.removed() or
                 (missing and self.deleted()))
 
     def add(self, list, prefix=""):
         with self._repo.wlock():
             ui, ds = self._repo.ui, self._repo.dirstate
             uipath = lambda f: ds.pathto(pathutil.join(prefix, f))
             rejected = []
             lstat = self._repo.wvfs.lstat
             for f in list:
                 # ds.pathto() returns an absolute file when this is invoked from
                 # the keyword extension.  That gets flagged as non-portable on
                 # Windows, since it contains the drive letter and colon.
                 scmutil.checkportable(ui, os.path.join(prefix, f))
                 try:
                     st = lstat(f)
                 except OSError:
                     ui.warn(_("%s does not exist!\n") % uipath(f))
                     rejected.append(f)
                     continue
                 if st.st_size > 10000000:
                     ui.warn(_("%s: up to %d MB of RAM may be required "
                               "to manage this file\n"
                               "(use 'hg revert %s' to cancel the "
                               "pending addition)\n")
                             % (f, 3 * st.st_size // 1000000, uipath(f)))
                 if not (stat.S_ISREG(st.st_mode) or stat.S_ISLNK(st.st_mode)):
                     ui.warn(_("%s not added: only files and symlinks "
                               "supported currently\n") % uipath(f))
                     rejected.append(f)
                 elif ds[f] in 'amn':
                     ui.warn(_("%s already tracked!\n") % uipath(f))
                 elif ds[f] == 'r':
                     ds.normallookup(f)
                 else:
                     ds.add(f)
             return rejected
 
     def forget(self, files, prefix=""):
         with self._repo.wlock():
             ds = self._repo.dirstate
             uipath = lambda f: ds.pathto(pathutil.join(prefix, f))
             rejected = []
             for f in files:
                 if f not in self._repo.dirstate:
                     self._repo.ui.warn(_("%s not tracked!\n") % uipath(f))
                     rejected.append(f)
                 elif self._repo.dirstate[f] != 'a':
                     self._repo.dirstate.remove(f)
                 else:
                     self._repo.dirstate.drop(f)
             return rejected
 
     def undelete(self, list):
         pctxs = self.parents()
         with self._repo.wlock():
             ds = self._repo.dirstate
             for f in list:
                 if self._repo.dirstate[f] != 'r':
                     self._repo.ui.warn(_("%s not removed!\n") % ds.pathto(f))
                 else:
                     fctx = f in pctxs[0] and pctxs[0][f] or pctxs[1][f]
                     t = fctx.data()
                     self._repo.wwrite(f, t, fctx.flags())
                     self._repo.dirstate.normal(f)
 
     def copy(self, source, dest):
         try:
             st = self._repo.wvfs.lstat(dest)
         except OSError as err:
             if err.errno != errno.ENOENT:
                 raise
             self._repo.ui.warn(_("%s does not exist!\n")
                                % self._repo.dirstate.pathto(dest))
             return
         if not (stat.S_ISREG(st.st_mode) or stat.S_ISLNK(st.st_mode)):
             self._repo.ui.warn(_("copy failed: %s is not a file or a "
                                  "symbolic link\n")
                                % self._repo.dirstate.pathto(dest))
         else:
             with self._repo.wlock():
                 if self._repo.dirstate[dest] in '?':
                     self._repo.dirstate.add(dest)
                 elif self._repo.dirstate[dest] in 'r':
                     self._repo.dirstate.normallookup(dest)
                 self._repo.dirstate.copy(source, dest)
 
     def match(self, pats=None, include=None, exclude=None, default='glob',
               listsubrepos=False, badfn=None):
         r = self._repo
 
         # Only a case insensitive filesystem needs magic to translate user input
         # to actual case in the filesystem.
         icasefs = not util.fscasesensitive(r.root)
         return matchmod.match(r.root, r.getcwd(), pats, include, exclude,
                               default, auditor=r.auditor, ctx=self,
                               listsubrepos=listsubrepos, badfn=badfn,
                               icasefs=icasefs)
 
     def flushall(self):
         pass # For overlayworkingfilectx compatibility.
 
     def _filtersuspectsymlink(self, files):
         if not files or self._repo.dirstate._checklink:
             return files
 
         # Symlink placeholders may get non-symlink-like contents
         # via user error or dereferencing by NFS or Samba servers,
         # so we filter out any placeholders that don't look like a
         # symlink
         sane = []
         for f in files:
             if self.flags(f) == 'l':
                 d = self[f].data()
                 if d == '' or len(d) >= 1024 or '\n' in d or util.binary(d):
                     self._repo.ui.debug('ignoring suspect symlink placeholder'
                                         ' "%s"\n' % f)
                     continue
             sane.append(f)
         return sane
 
     def _checklookup(self, files):
         # check for any possibly clean files
         if not files:
             return [], [], []
 
         modified = []
         deleted = []
         fixup = []
         pctx = self._parents[0]
         # do a full compare of any files that might have changed
         for f in sorted(files):
             try:
                 # This will return True for a file that got replaced by a
                 # directory in the interim, but fixing that is pretty hard.
                 if (f not in pctx or self.flags(f) != pctx.flags(f)
                     or pctx[f].cmp(self[f])):
                     modified.append(f)
                 else:
                     fixup.append(f)
             except (IOError, OSError):
                 # A file become inaccessible in between? Mark it as deleted,
                 # matching dirstate behavior (issue5584).
                 # The dirstate has more complex behavior around whether a
                 # missing file matches a directory, etc, but we don't need to
                 # bother with that: if f has made it to this point, we're sure
                 # it's in the dirstate.
                 deleted.append(f)
 
         return modified, deleted, fixup
 
     def _poststatusfixup(self, status, fixup):
         """update dirstate for files that are actually clean"""
         poststatus = self._repo.postdsstatus()
         if fixup or poststatus:
             try:
                 oldid = self._repo.dirstate.identity()
 
                 # updating the dirstate is optional
                 # so we don't wait on the lock
                 # wlock can invalidate the dirstate, so cache normal _after_
                 # taking the lock
                 with self._repo.wlock(False):
                     if self._repo.dirstate.identity() == oldid:
                         if fixup:
                             normal = self._repo.dirstate.normal
                             for f in fixup:
                                 normal(f)
                             # write changes out explicitly, because nesting
                             # wlock at runtime may prevent 'wlock.release()'
                             # after this block from doing so for subsequent
                             # changing files
                             tr = self._repo.currenttransaction()
                             self._repo.dirstate.write(tr)
 
                         if poststatus:
                             for ps in poststatus:
                                 ps(self, status)
                     else:
                         # in this case, writing changes out breaks
                         # consistency, because .hg/dirstate was
                         # already changed simultaneously after last
                         # caching (see also issue5584 for detail)
                         self._repo.ui.debug('skip updating dirstate: '
                                             'identity mismatch\n')
             except error.LockError:
                 pass
             finally:
                 # Even if the wlock couldn't be grabbed, clear out the list.
                 self._repo.clearpostdsstatus()
 
     def _dirstatestatus(self, match, ignored=False, clean=False, unknown=False):
         '''Gets the status from the dirstate -- internal use only.'''
-        listignored, listclean, listunknown = ignored, clean, unknown
         subrepos = []
         if '.hgsub' in self:
             subrepos = sorted(self.substate)
-        cmp, s = self._repo.dirstate.status(match, subrepos, listignored,
-                                            listclean, listunknown)
+        cmp, s = self._repo.dirstate.status(match, subrepos, ignored=ignored,
+                                            clean=clean, unknown=unknown)
 
         # check for any possibly clean files
         fixup = []
         if cmp:
             modified2, deleted2, fixup = self._checklookup(cmp)
             s.modified.extend(modified2)
             s.deleted.extend(deleted2)
 
-            if fixup and listclean:
+            if fixup and clean:
                 s.clean.extend(fixup)
 
         self._poststatusfixup(s, fixup)
 
         if match.always():
             # cache for performance
             if s.unknown or s.ignored or s.clean:
                 # "_status" is cached with list*=False in the normal route
                 self._status = scmutil.status(s.modified, s.added, s.removed,
                                               s.deleted, [], [], [])
             else:
                 self._status = s
 
         return s
 
     @propertycache
     def _manifest(self):
         """generate a manifest corresponding to the values in self._status
 
         This reuse the file nodeid from parent, but we use special node
         identifiers for added and modified files. This is used by manifests
         merge to see that files are different and by update logic to avoid
         deleting newly added files.
         """
         return self._buildstatusmanifest(self._status)
 
     def _buildstatusmanifest(self, status):
         """Builds a manifest that includes the given status results."""
         parents = self.parents()
 
         man = parents[0].manifest().copy()
 
         ff = self._flagfunc
         for i, l in ((addednodeid, status.added),
                      (modifiednodeid, status.modified)):
             for f in l:
                 man[f] = i
                 try:
                     man.setflag(f, ff(f))
                 except OSError:
                     pass
 
         for f in status.deleted + status.removed:
             if f in man:
                 del man[f]
 
         return man
 
     def _buildstatus(self, other, s, match, listignored, listclean,
                      listunknown):
         """build a status with respect to another context
 
         This includes logic for maintaining the fast path of status when
         comparing the working directory against its parent, which is to skip
         building a new manifest if self (working directory) is not comparing
         against its parent (repo['.']).
         """
         s = self._dirstatestatus(match, listignored, listclean, listunknown)
         # Filter out symlinks that, in the case of FAT32 and NTFS filesystems,
         # might have accidentally ended up with the entire contents of the file
         # they are supposed to be linking to.
         s.modified[:] = self._filtersuspectsymlink(s.modified)
         if other != self._repo['.']:
             s = super(workingctx, self)._buildstatus(other, s, match,
                                                      listignored, listclean,
                                                      listunknown)
         return s
 
     def _matchstatus(self, other, match):
         """override the match method with a filter for directory patterns
 
         We use inheritance to customize the match.bad method only in cases of
         workingctx since it belongs only to the working directory when
         comparing against the parent changeset.
 
         If we aren't comparing against the working directory's parent, then we
         just use the default match object sent to us.
         """
         if other != self._repo['.']:
             def bad(f, msg):
                 # 'f' may be a directory pattern from 'match.files()',
                 # so 'f not in ctx1' is not enough
                 if f not in other and not other.hasdir(f):
                     self._repo.ui.warn('%s: %s\n' %
                                        (self._repo.dirstate.pathto(f), msg))
             match.bad = bad
         return match
 
     def markcommitted(self, node):
         super(workingctx, self).markcommitted(node)
 
         sparse.aftercommit(self._repo, node)
 
 class committablefilectx(basefilectx):
     """A committablefilectx provides common functionality for a file context
     that wants the ability to commit, e.g. workingfilectx or memfilectx."""
     def __init__(self, repo, path, filelog=None, ctx=None):
         self._repo = repo
         self._path = path
         self._changeid = None
         self._filerev = self._filenode = None
 
         if filelog is not None:
             self._filelog = filelog
         if ctx:
             self._changectx = ctx
 
     def __nonzero__(self):
         return True
 
     __bool__ = __nonzero__
 
     def linkrev(self):
         # linked to self._changectx no matter if file is modified or not
         return self.rev()
 
     def parents(self):
         '''return parent filectxs, following copies if necessary'''
         def filenode(ctx, path):
             return ctx._manifest.get(path, nullid)
 
         path = self._path
         fl = self._filelog
         pcl = self._changectx._parents
         renamed = self.renamed()
 
         if renamed:
             pl = [renamed + (None,)]
         else:
             pl = [(path, filenode(pcl[0], path), fl)]
 
         for pc in pcl[1:]:
             pl.append((path, filenode(pc, path), fl))
 
         return [self._parentfilectx(p, fileid=n, filelog=l)
                 for p, n, l in pl if n != nullid]
 
     def children(self):
         return []
 
 class workingfilectx(committablefilectx):
     """A workingfilectx object makes access to data related to a particular
        file in the working directory convenient."""
     def __init__(self, repo, path, filelog=None, workingctx=None):
         super(workingfilectx, self).__init__(repo, path, filelog, workingctx)
 
     @propertycache
     def _changectx(self):
         return workingctx(self._repo)
 
     def data(self):
         return self._repo.wread(self._path)
     def renamed(self):
         rp = self._repo.dirstate.copied(self._path)
         if not rp:
             return None
         return rp, self._changectx._parents[0]._manifest.get(rp, nullid)
 
     def size(self):
         return self._repo.wvfs.lstat(self._path).st_size
     def date(self):
         t, tz = self._changectx.date()
         try:
             return (self._repo.wvfs.lstat(self._path).st_mtime, tz)
         except OSError as err:
             if err.errno != errno.ENOENT:
                 raise
             return (t, tz)
 
     def exists(self):
         return self._repo.wvfs.exists(self._path)
 
     def lexists(self):
         return self._repo.wvfs.lexists(self._path)
 
     def audit(self):
         return self._repo.wvfs.audit(self._path)
 
     def cmp(self, fctx):
         """compare with other file context
 
         returns True if different than fctx.
         """
         # fctx should be a filectx (not a workingfilectx)
         # invert comparison to reuse the same code path
         return fctx.cmp(self)
 
     def remove(self, ignoremissing=False):
         """wraps unlink for a repo's working directory"""
         self._repo.wvfs.unlinkpath(self._path, ignoremissing=ignoremissing)
 
     def write(self, data, flags, backgroundclose=False):
         """wraps repo.wwrite"""
         self._repo.wwrite(self._path, data, flags,
                           backgroundclose=backgroundclose)
 
     def clearunknown(self):
         """Removes conflicting items in the working directory so that
         ``write()`` can be called successfully.
         """
         wvfs = self._repo.wvfs
         if wvfs.isdir(self._path) and not wvfs.islink(self._path):
             wvfs.removedirs(self._path)
 
     def setflags(self, l, x):
         self._repo.wvfs.setflags(self._path, l, x)
 
 class overlayworkingctx(workingctx):
     """Wraps another mutable context with a write-back cache that can be flushed
     at a later time.
 
     self._cache[path] maps to a dict with keys: {
         'exists': bool?
         'date': date?
         'data': str?
         'flags': str?
     }
     If `exists` is True, `flags` must be non-None and 'date' is non-None. If it
     is `False`, the file was deleted.
     """
 
     def __init__(self, repo, wrappedctx):
         super(overlayworkingctx, self).__init__(repo)
         self._repo = repo
         self._wrappedctx = wrappedctx
         self._clean()
 
     def data(self, path):
         if self.isdirty(path):
             if self._cache[path]['exists']:
                 if self._cache[path]['data']:
                     return self._cache[path]['data']
                 else:
                     # Must fallback here, too, because we only set flags.
                     return self._wrappedctx[path].data()
             else:
                 raise error.ProgrammingError("No such file or directory: %s" %
                                              self._path)
         else:
             return self._wrappedctx[path].data()
 
     def filedate(self, path):
         if self.isdirty(path):
             return self._cache[path]['date']
         else:
             return self._wrappedctx[path].date()
 
     def flags(self, path):
         if self.isdirty(path):
             if self._cache[path]['exists']:
                 return self._cache[path]['flags']
             else:
                 raise error.ProgrammingError("No such file or directory: %s" %
                                              self._path)
         else:
             return self._wrappedctx[path].flags()
 
     def write(self, path, data, flags=''):
         if data is None:
             raise error.ProgrammingError("data must be non-None")
         self._markdirty(path, exists=True, data=data, date=util.makedate(),
                         flags=flags)
 
     def setflags(self, path, l, x):
         self._markdirty(path, exists=True, date=util.makedate(),
                         flags=(l and 'l' or '') + (x and 'x' or ''))
 
     def remove(self, path):
         self._markdirty(path, exists=False)
 
     def exists(self, path):
         """exists behaves like `lexists`, but needs to follow symlinks and
         return False if they are broken.
         """
         if self.isdirty(path):
             # If this path exists and is a symlink, "follow" it by calling
             # exists on the destination path.
             if (self._cache[path]['exists'] and
                         'l' in self._cache[path]['flags']):
                 return self.exists(self._cache[path]['data'].strip())
             else:
                 return self._cache[path]['exists']
         return self._wrappedctx[path].exists()
 
     def lexists(self, path):
         """lexists returns True if the path exists"""
         if self.isdirty(path):
             return self._cache[path]['exists']
         return self._wrappedctx[path].lexists()
 
     def size(self, path):
         if self.isdirty(path):
             if self._cache[path]['exists']:
                 return len(self._cache[path]['data'])
             else:
                 raise error.ProgrammingError("No such file or directory: %s" %
                                              self._path)
         return self._wrappedctx[path].size()
 
     def flushall(self):
         for path in self._writeorder:
             entry = self._cache[path]
             if entry['exists']:
                 self._wrappedctx[path].clearunknown()
                 if entry['data'] is not None:
                     if entry['flags'] is None:
                         raise error.ProgrammingError('data set but not flags')
                     self._wrappedctx[path].write(
                         entry['data'],
                         entry['flags'])
                 else:
                     self._wrappedctx[path].setflags(
                         'l' in entry['flags'],
                         'x' in entry['flags'])
             else:
                 self._wrappedctx[path].remove(path)
         self._clean()
 
     def isdirty(self, path):
         return path in self._cache
 
     def _clean(self):
         self._cache = {}
         self._writeorder = []
 
     def _markdirty(self, path, exists, data=None, date=None, flags=''):
         if path not in self._cache:
             self._writeorder.append(path)
 
         self._cache[path] = {
             'exists': exists,
             'data': data,
             'date': date,
             'flags': flags,
         }
 
     def filectx(self, path, filelog=None):
         return overlayworkingfilectx(self._repo, path, parent=self,
                                      filelog=filelog)
 
 class overlayworkingfilectx(workingfilectx):
     """Wrap a ``workingfilectx`` but intercepts all writes into an in-memory
     cache, which can be flushed through later by calling ``flush()``."""
 
     def __init__(self, repo, path, filelog=None, parent=None):
         super(overlayworkingfilectx, self).__init__(repo, path, filelog,
                                                     parent)
         self._repo = repo
         self._parent = parent
         self._path = path
 
     def ctx(self):
         return self._parent
 
     def data(self):
         return self._parent.data(self._path)
 
     def date(self):
         return self._parent.filedate(self._path)
 
     def exists(self):
         return self.lexists()
 
     def lexists(self):
         return self._parent.exists(self._path)
 
     def renamed(self):
         # Copies are currently tracked in the dirstate as before. Straight copy
         # from workingfilectx.
         rp = self._repo.dirstate.copied(self._path)
         if not rp:
             return None
         return rp, self._changectx._parents[0]._manifest.get(rp, nullid)
 
     def size(self):
         return self._parent.size(self._path)
 
     def audit(self):
         pass
 
     def flags(self):
         return self._parent.flags(self._path)
 
     def setflags(self, islink, isexec):
         return self._parent.setflags(self._path, islink, isexec)
 
     def write(self, data, flags, backgroundclose=False):
         return self._parent.write(self._path, data, flags)
 
     def remove(self, ignoremissing=False):
         return self._parent.remove(self._path)
 
 class workingcommitctx(workingctx):
     """A workingcommitctx object makes access to data related to
     the revision being committed convenient.
 
     This hides changes in the working directory, if they aren't
     committed in this context.
     """
     def __init__(self, repo, changes,
                  text="", user=None, date=None, extra=None):
         super(workingctx, self).__init__(repo, text, user, date, extra,
                                          changes)
 
     def _dirstatestatus(self, match, ignored=False, clean=False, unknown=False):
         """Return matched files only in ``self._status``
 
         Uncommitted files appear "clean" via this context, even if
         they aren't actually so in the working directory.
         """
         if clean:
             clean = [f for f in self._manifest if f not in self._changedset]
         else:
             clean = []
         return scmutil.status([f for f in self._status.modified if match(f)],
                               [f for f in self._status.added if match(f)],
                               [f for f in self._status.removed if match(f)],
                               [], [], [], clean)
 
     @propertycache
     def _changedset(self):
         """Return the set of files changed in this context
         """
         changed = set(self._status.modified)
         changed.update(self._status.added)
         changed.update(self._status.removed)
         return changed
 
 def makecachingfilectxfn(func):
     """Create a filectxfn that caches based on the path.
 
     We can't use util.cachefunc because it uses all arguments as the cache
     key and this creates a cycle since the arguments include the repo and
     memctx.
     """
     cache = {}
 
     def getfilectx(repo, memctx, path):
         if path not in cache:
             cache[path] = func(repo, memctx, path)
         return cache[path]
 
     return getfilectx
 
 def memfilefromctx(ctx):
     """Given a context return a memfilectx for ctx[path]
 
     This is a convenience method for building a memctx based on another
     context.
     """
     def getfilectx(repo, memctx, path):
         fctx = ctx[path]
         # this is weird but apparently we only keep track of one parent
         # (why not only store that instead of a tuple?)
         copied = fctx.renamed()
         if copied:
             copied = copied[0]
         return memfilectx(repo, path, fctx.data(),
                           islink=fctx.islink(), isexec=fctx.isexec(),
                           copied=copied, memctx=memctx)
 
     return getfilectx
 
 def memfilefrompatch(patchstore):
     """Given a patch (e.g. patchstore object) return a memfilectx
 
     This is a convenience method for building a memctx based on a patchstore.
     """
     def getfilectx(repo, memctx, path):
         data, mode, copied = patchstore.getfile(path)
         if data is None:
             return None
         islink, isexec = mode
         return memfilectx(repo, path, data, islink=islink,
                           isexec=isexec, copied=copied,
                           memctx=memctx)
 
     return getfilectx
 
 class memctx(committablectx):
     """Use memctx to perform in-memory commits via localrepo.commitctx().
 
     Revision information is supplied at initialization time while
     related files data and is made available through a callback
     mechanism.  'repo' is the current localrepo, 'parents' is a
     sequence of two parent revisions identifiers (pass None for every
     missing parent), 'text' is the commit message and 'files' lists
     names of files touched by the revision (normalized and relative to
     repository root).
 
     filectxfn(repo, memctx, path) is a callable receiving the
     repository, the current memctx object and the normalized path of
     requested file, relative to repository root. It is fired by the
     commit function for every file in 'files', but calls order is
     undefined. If the file is available in the revision being
     committed (updated or added), filectxfn returns a memfilectx
     object. If the file was removed, filectxfn return None for recent
     Mercurial. Moved files are represented by marking the source file
     removed and the new file added with copy information (see
     memfilectx).
 
     user receives the committer name and defaults to current
     repository username, date is the commit date in any format
     supported by util.parsedate() and defaults to current date, extra
     is a dictionary of metadata or is left empty.
     """
 
     # Mercurial <= 3.1 expects the filectxfn to raise IOError for missing files.
     # Extensions that need to retain compatibility across Mercurial 3.1 can use
     # this field to determine what to do in filectxfn.
     _returnnoneformissingfiles = True
 
     def __init__(self, repo, parents, text, files, filectxfn, user=None,
                  date=None, extra=None, branch=None, editor=False):
         super(memctx, self).__init__(repo, text, user, date, extra)
         self._rev = None
         self._node = None
         parents = [(p or nullid) for p in parents]
         p1, p2 = parents
         self._parents = [changectx(self._repo, p) for p in (p1, p2)]
         files = sorted(set(files))
         self._files = files
         if branch is not None:
             self._extra['branch'] = encoding.fromlocal(branch)
         self.substate = {}
 
         if isinstance(filectxfn, patch.filestore):
             filectxfn = memfilefrompatch(filectxfn)
         elif not callable(filectxfn):
             # if store is not callable, wrap it in a function
             filectxfn = memfilefromctx(filectxfn)
 
         # memoizing increases performance for e.g. vcs convert scenarios.
         self._filectxfn = makecachingfilectxfn(filectxfn)
 
         if editor:
             self._text = editor(self._repo, self, [])
             self._repo.savecommitmessage(self._text)
 
     def filectx(self, path, filelog=None):
         """get a file context from the working directory
 
         Returns None if file doesn't exist and should be removed."""
         return self._filectxfn(self._repo, self, path)
 
     def commit(self):
         """commit context to the repo"""
         return self._repo.commitctx(self)
 
     @propertycache
     def _manifest(self):
         """generate a manifest based on the return values of filectxfn"""
 
         # keep this simple for now; just worry about p1
         pctx = self._parents[0]
         man = pctx.manifest().copy()
 
         for f in self._status.modified:
             p1node = nullid
             p2node = nullid
             p = pctx[f].parents() # if file isn't in pctx, check p2?
             if len(p) > 0:
                 p1node = p[0].filenode()
                 if len(p) > 1:
                     p2node = p[1].filenode()
             man[f] = revlog.hash(self[f].data(), p1node, p2node)
 
         for f in self._status.added:
             man[f] = revlog.hash(self[f].data(), nullid, nullid)
 
         for f in self._status.removed:
             if f in man:
                 del man[f]
 
         return man
 
     @propertycache
     def _status(self):
         """Calculate exact status from ``files`` specified at construction
         """
         man1 = self.p1().manifest()
         p2 = self._parents[1]
         # "1 < len(self._parents)" can't be used for checking
         # existence of the 2nd parent, because "memctx._parents" is
         # explicitly initialized by the list, of which length is 2.
         if p2.node() != nullid:
             man2 = p2.manifest()
             managing = lambda f: f in man1 or f in man2
         else:
             managing = lambda f: f in man1
 
         modified, added, removed = [], [], []
         for f in self._files:
             if not managing(f):
                 added.append(f)
             elif self[f]:
                 modified.append(f)
             else:
                 removed.append(f)
 
         return scmutil.status(modified, added, removed, [], [], [], [])
 
 class memfilectx(committablefilectx):
     """memfilectx represents an in-memory file to commit.
 
     See memctx and committablefilectx for more details.
     """
     def __init__(self, repo, path, data, islink=False,
                  isexec=False, copied=None, memctx=None):
         """
         path is the normalized file path relative to repository root.
         data is the file content as a string.
         islink is True if the file is a symbolic link.
         isexec is True if the file is executable.
         copied is the source file path if current file was copied in the
         revision being committed, or None."""
         super(memfilectx, self).__init__(repo, path, None, memctx)
         self._data = data
         self._flags = (islink and 'l' or '') + (isexec and 'x' or '')
         self._copied = None
         if copied:
             self._copied = (copied, nullid)
 
     def data(self):
         return self._data
 
     def remove(self, ignoremissing=False):
         """wraps unlink for a repo's working directory"""
         # need to figure out what to do here
         del self._changectx[self._path]
 
     def write(self, data, flags):
         """wraps repo.wwrite"""
         self._data = data
 
 class overlayfilectx(committablefilectx):
     """Like memfilectx but take an original filectx and optional parameters to
     override parts of it. This is useful when fctx.data() is expensive (i.e.
     flag processor is expensive) and raw data, flags, and filenode could be
     reused (ex. rebase or mode-only amend a REVIDX_EXTSTORED file).
     """
 
     def __init__(self, originalfctx, datafunc=None, path=None, flags=None,
                  copied=None, ctx=None):
         """originalfctx: filecontext to duplicate
 
         datafunc: None or a function to override data (file content). It is a
         function to be lazy. path, flags, copied, ctx: None or overridden value
 
         copied could be (path, rev), or False. copied could also be just path,
         and will be converted to (path, nullid). This simplifies some callers.
         """
 
         if path is None:
             path = originalfctx.path()
         if ctx is None:
             ctx = originalfctx.changectx()
             ctxmatch = lambda: True
         else:
             ctxmatch = lambda: ctx == originalfctx.changectx()
 
         repo = originalfctx.repo()
         flog = originalfctx.filelog()
         super(overlayfilectx, self).__init__(repo, path, flog, ctx)
 
         if copied is None:
             copied = originalfctx.renamed()
             copiedmatch = lambda: True
         else:
             if copied and not isinstance(copied, tuple):
                 # repo._filecommit will recalculate copyrev so nullid is okay
                 copied = (copied, nullid)
             copiedmatch = lambda: copied == originalfctx.renamed()
 
         # When data, copied (could affect data), ctx (could affect filelog
         # parents) are not overridden, rawdata, rawflags, and filenode may be
         # reused (repo._filecommit should double check filelog parents).
         #
         # path, flags are not hashed in filelog (but in manifestlog) so they do
         # not affect reusable here.
         #
         # If ctx or copied is overridden to a same value with originalfctx,
         # still consider it's reusable. originalfctx.renamed() may be a bit
         # expensive so it's not called unless necessary. Assuming datafunc is
         # always expensive, do not call it for this "reusable" test.
         reusable = datafunc is None and ctxmatch() and copiedmatch()
 
         if datafunc is None:
             datafunc = originalfctx.data
         if flags is None:
             flags = originalfctx.flags()
 
         self._datafunc = datafunc
         self._flags = flags
         self._copied = copied
 
         if reusable:
             # copy extra fields from originalfctx
             attrs = ['rawdata', 'rawflags', '_filenode', '_filerev']
             for attr in attrs:
                 if util.safehasattr(originalfctx, attr):
                     setattr(self, attr, getattr(originalfctx, attr))
 
     def data(self):
         return self._datafunc()
 
 class metadataonlyctx(committablectx):
     """Like memctx but it's reusing the manifest of different commit.
     Intended to be used by lightweight operations that are creating
     metadata-only changes.
 
     Revision information is supplied at initialization time.  'repo' is the
     current localrepo, 'ctx' is original revision which manifest we're reuisng
     'parents' is a sequence of two parent revisions identifiers (pass None for
     every missing parent), 'text' is the commit.
 
     user receives the committer name and defaults to current repository
     username, date is the commit date in any format supported by
     util.parsedate() and defaults to current date, extra is a dictionary of
     metadata or is left empty.
     """
     def __new__(cls, repo, originalctx, *args, **kwargs):
         return super(metadataonlyctx, cls).__new__(cls, repo)
 
     def __init__(self, repo, originalctx, parents=None, text=None, user=None,
                  date=None, extra=None, editor=False):
         if text is None:
             text = originalctx.description()
         super(metadataonlyctx, self).__init__(repo, text, user, date, extra)
         self._rev = None
         self._node = None
         self._originalctx = originalctx
         self._manifestnode = originalctx.manifestnode()
         if parents is None:
             parents = originalctx.parents()
         else:
             parents = [repo[p] for p in parents if p is not None]
         parents = parents[:]
         while len(parents) < 2:
             parents.append(repo[nullid])
         p1, p2 = self._parents = parents
 
         # sanity check to ensure that the reused manifest parents are
         # manifests of our commit parents
         mp1, mp2 = self.manifestctx().parents
         if p1 != nullid and p1.manifestnode() != mp1:
             raise RuntimeError('can\'t reuse the manifest: '
                                'its p1 doesn\'t match the new ctx p1')
         if p2 != nullid and p2.manifestnode() != mp2:
             raise RuntimeError('can\'t reuse the manifest: '
                                'its p2 doesn\'t match the new ctx p2')
 
         self._files = originalctx.files()
         self.substate = {}
 
         if editor:
             self._text = editor(self._repo, self, [])
             self._repo.savecommitmessage(self._text)
 
     def manifestnode(self):
         return self._manifestnode
 
     @property
     def _manifestctx(self):
         return self._repo.manifestlog[self._manifestnode]
 
     def filectx(self, path, filelog=None):
         return self._originalctx.filectx(path, filelog=filelog)
 
     def commit(self):
         """commit context to the repo"""
         return self._repo.commitctx(self)
 
     @property
     def _manifest(self):
         return self._originalctx.manifest()
 
     @propertycache
     def _status(self):
         """Calculate exact status from ``files`` specified in the ``origctx``
         and parents manifests.
         """
         man1 = self.p1().manifest()
         p2 = self._parents[1]
         # "1 < len(self._parents)" can't be used for checking
         # existence of the 2nd parent, because "metadataonlyctx._parents" is
         # explicitly initialized by the list, of which length is 2.
         if p2.node() != nullid:
             man2 = p2.manifest()
             managing = lambda f: f in man1 or f in man2
         else:
             managing = lambda f: f in man1
 
         modified, added, removed = [], [], []
         for f in self._files:
             if not managing(f):
                 added.append(f)
             elif f in self:
                 modified.append(f)
             else:
                 removed.append(f)
 
         return scmutil.status(modified, added, removed, [], [], [], [])
 
 class arbitraryfilectx(object):
     """Allows you to use filectx-like functions on a file in an arbitrary
     location on disk, possibly not in the working directory.
     """
     def __init__(self, path):
         self._path = path
 
     def cmp(self, otherfilectx):
         return self.data() != otherfilectx.data()
 
     def path(self):
         return self._path
 
     def flags(self):
         return ''
 
     def data(self):
         return util.readfile(self._path)
 
     def decodeddata(self):
         with open(self._path, "rb") as f:
             return f.read()
 
     def remove(self):
         util.unlink(self._path)
 
     def write(self, data, flags):
         assert not flags
         with open(self._path, "w") as f:
             f.write(data)