subrepo: run the repo decoders when archiving

The decoders were already run by default for the main repo, so this seemed like
an oversight.

The extdiff extension has been using 'archive' since 68822b7cdd01 to support -S,
and a colleague noticed that after diffing, making changes, and closing it, the
line endings were wrong for the diff-tool modified files in the subrepository.
(Files in the parent repo were correct, with the same .hgeol settings.)  The
editor (Visual Studio in this case) reloads the file, but doesn't notice the EOL
change.  It still adds new lines with the original EOL setting, and the file
ends up inconsistent.

Without this change, the first file `cat`d in the test prints '\r (esc)' EOL,
but the second doesn't on Windows or Linux.

hgext/largefiles/overrides.py

@@ -1,1439 +1,1441 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._fileroots = set(m._files)
     m._always = 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._fileroots = set(m._files)
     m._always = 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('large')
     lfsize = lfutil.getminsize(
         ui, lfutil.islfilesrepo(repo), opts.get('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 repo.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:
                     util.unlinkpath(repo.wjoin(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:
                 util.unlinkpath(repo.wjoin(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('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):
     try:
         repo._repo.lfstatus = True
         return orig(repo, ignoreupdate)
     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)
 
             pats.update(fixpats(f, tostandin) for f in p)
         else:
             def tostandin(f):
                 if lfutil.splitstandin(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 repo[ctx.node()]:
                 m._files[i] = standin
             elif m._files[i] not in repo[ctx.node()] \
                     and repo.wvfs.isdir(standin):
                 m._files.append(standin)
 
         m._fileroots = set(m._files)
         m._always = 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' % 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, mynode, orig, fcd, fco, fca,
                       labels=None):
     if not lfutil.isstandin(orig) or fcd.isabsent() or fco.isabsent():
         return origfn(premerge, repo, mynode, orig, fcd, fco, fca,
                       labels=labels)
 
     ahash = fca.data().strip().lower()
     dhash = fcd.data().strip().lower()
     ohash = fco.data().strip().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._fileroots = set(m._files)
             origmatchfn = m.matchfn
             m.matchfn = lambda f: (lfutil.isstandin(f) and
                                 (f in manifest) and
                                 origmatchfn(lfutil.splitstandin(f)) or
                                 None)
             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.
                     util.unlinkpath(repo.wjoin(srclfile), 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, lfutil.standin(lfile))
         for lfile in s.deleted:
             if (repo.wvfs.exists(lfutil.standin(lfile))):
                 repo.wvfs.unlink(lfutil.standin(lfile))
 
         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)
 
             def tostandin(f):
                 standin = lfutil.standin(f)
                 if standin in ctx or standin in mctx:
                     return standin
                 elif standin in repo[None] or lfdirstate[f] == 'r':
                     return None
                 return f
             m._files = [tostandin(f) for f in m._files]
             m._files = [f for f in m._files if f is not None]
             m._fileroots = set(m._files)
             origmatchfn = m.matchfn
             def matchfn(f):
                 if lfutil.isstandin(f):
                     return (origmatchfn(lfutil.splitstandin(f)) 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 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", True):
         write('.hg_archival.txt', 0o644, False,
               lambda: archival.buildmetadata(ctx))
 
     for f in ctx:
         ff = ctx.flags(f)
         getdata = ctx[f].data
         if lfutil.isstandin(f):
             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')
                        % lfutil.splitstandin(f))
             else:
                 path = lfutil.splitstandin(f)
 
             f = lfutil.splitstandin(f)
 
             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):
+def hgsubrepoarchive(orig, repo, archiver, prefix, match=None, decode=True):
     if not repo._repo.lfstatus:
-        return orig(repo, archiver, prefix, match)
+        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
         if lfutil.isstandin(f):
             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')
                        % lfutil.splitstandin(f))
             else:
                 path = lfutil.splitstandin(f)
 
             f = lfutil.splitstandin(f)
 
             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)
+        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
     forget = sorted(s.modified + s.added + s.deleted + s.clean)
     forget = [f for f in forget if lfutil.standin(f) in repo[None].manifest()]
 
     for f in forget:
         if lfutil.standin(f) not in repo.dirstate and not \
                 repo.wvfs.isdir(lfutil.standin(f)):
             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:
             util.unlinkpath(repo.wjoin(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)
 
     # 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.readstandin(repo, lf, ctx.rev())
                 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)
         oldclean = set(s.clean)
         pctx = repo['.']
         for lfile in unsure + s.modified:
             lfileabs = repo.wvfs.join(lfile)
             if not repo.wvfs.exists(lfileabs):
                 continue
             lfhash = lfutil.hashrepofile(repo, lfile)
             standin = lfutil.standin(lfile)
             lfutil.writestandin(repo, standin, lfhash,
                                 lfutil.getexecutable(lfileabs))
             if (standin in pctx and
                 lfhash == lfutil.readstandin(repo, lfile, '.')):
                 oldclean.add(lfile)
         for lfile in s.added:
             lfutil.updatestandin(repo, lfutil.standin(lfile))
         # 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)
 
         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 = [lfutil.splitstandin(f) for f in files if lfutil.isstandin(f)]
     if filelist:
         lfcommands.updatelfiles(repo.ui, repo, filelist=filelist,
                                 printmessage=False, normallookup=True)
 
     return result

mercurial/archival.py

@@ -1,340 +1,340 b''
 # archival.py - revision archival for mercurial
 #
 # Copyright 2006 Vadim Gelfer <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import gzip
 import os
 import struct
 import tarfile
 import time
 import zipfile
 import zlib
 
 from .i18n import _
 
 from . import (
     cmdutil,
     encoding,
     error,
     match as matchmod,
     scmutil,
     util,
 )
 stringio = util.stringio
 
 # from unzip source code:
 _UNX_IFREG = 0x8000
 _UNX_IFLNK = 0xa000
 
 def tidyprefix(dest, kind, prefix):
     '''choose prefix to use for names in archive.  make sure prefix is
     safe for consumers.'''
 
     if prefix:
         prefix = util.normpath(prefix)
     else:
         if not isinstance(dest, str):
             raise ValueError('dest must be string if no prefix')
         prefix = os.path.basename(dest)
         lower = prefix.lower()
         for sfx in exts.get(kind, []):
             if lower.endswith(sfx):
                 prefix = prefix[:-len(sfx)]
                 break
     lpfx = os.path.normpath(util.localpath(prefix))
     prefix = util.pconvert(lpfx)
     if not prefix.endswith('/'):
         prefix += '/'
     # Drop the leading '.' path component if present, so Windows can read the
     # zip files (issue4634)
     if prefix.startswith('./'):
         prefix = prefix[2:]
     if prefix.startswith('../') or os.path.isabs(lpfx) or '/../' in prefix:
         raise error.Abort(_('archive prefix contains illegal components'))
     return prefix
 
 exts = {
     'tar': ['.tar'],
     'tbz2': ['.tbz2', '.tar.bz2'],
     'tgz': ['.tgz', '.tar.gz'],
     'zip': ['.zip'],
     }
 
 def guesskind(dest):
     for kind, extensions in exts.iteritems():
         if any(dest.endswith(ext) for ext in extensions):
             return kind
     return None
 
 def _rootctx(repo):
     # repo[0] may be hidden
     for rev in repo:
         return repo[rev]
     return repo['null']
 
 def buildmetadata(ctx):
     '''build content of .hg_archival.txt'''
     repo = ctx.repo()
     hex = ctx.hex()
     if ctx.rev() is None:
         hex = ctx.p1().hex()
         if ctx.dirty():
             hex += '+'
 
     base = 'repo: %s\nnode: %s\nbranch: %s\n' % (
         _rootctx(repo).hex(), hex, encoding.fromlocal(ctx.branch()))
 
     tags = ''.join('tag: %s\n' % t for t in ctx.tags()
                    if repo.tagtype(t) == 'global')
     if not tags:
         repo.ui.pushbuffer()
         opts = {'template': '{latesttag}\n{latesttagdistance}\n'
                             '{changessincelatesttag}',
                 'style': '', 'patch': None, 'git': None}
         cmdutil.show_changeset(repo.ui, repo, opts).show(ctx)
         ltags, dist, changessince = repo.ui.popbuffer().split('\n')
         ltags = ltags.split(':')
         tags = ''.join('latesttag: %s\n' % t for t in ltags)
         tags += 'latesttagdistance: %s\n' % dist
         tags += 'changessincelatesttag: %s\n' % changessince
 
     return base + tags
 
 class tarit(object):
     '''write archive to tar file or stream.  can write uncompressed,
     or compress with gzip or bzip2.'''
 
     class GzipFileWithTime(gzip.GzipFile):
 
         def __init__(self, *args, **kw):
             timestamp = None
             if 'timestamp' in kw:
                 timestamp = kw.pop('timestamp')
             if timestamp is None:
                 self.timestamp = time.time()
             else:
                 self.timestamp = timestamp
             gzip.GzipFile.__init__(self, *args, **kw)
 
         def _write_gzip_header(self):
             self.fileobj.write('\037\213')             # magic header
             self.fileobj.write('\010')                 # compression method
             fname = self.name
             if fname and fname.endswith('.gz'):
                 fname = fname[:-3]
             flags = 0
             if fname:
                 flags = gzip.FNAME
             self.fileobj.write(chr(flags))
             gzip.write32u(self.fileobj, long(self.timestamp))
             self.fileobj.write('\002')
             self.fileobj.write('\377')
             if fname:
                 self.fileobj.write(fname + '\000')
 
     def __init__(self, dest, mtime, kind=''):
         self.mtime = mtime
         self.fileobj = None
 
         def taropen(mode, name='', fileobj=None):
             if kind == 'gz':
                 mode = mode[0]
                 if not fileobj:
                     fileobj = open(name, mode + 'b')
                 gzfileobj = self.GzipFileWithTime(name, mode + 'b',
                                                   zlib.Z_BEST_COMPRESSION,
                                                   fileobj, timestamp=mtime)
                 self.fileobj = gzfileobj
                 return tarfile.TarFile.taropen(name, mode, gzfileobj)
             else:
                 return tarfile.open(name, mode + kind, fileobj)
 
         if isinstance(dest, str):
             self.z = taropen('w:', name=dest)
         else:
             self.z = taropen('w|', fileobj=dest)
 
     def addfile(self, name, mode, islink, data):
         i = tarfile.TarInfo(name)
         i.mtime = self.mtime
         i.size = len(data)
         if islink:
             i.type = tarfile.SYMTYPE
             i.mode = 0o777
             i.linkname = data
             data = None
             i.size = 0
         else:
             i.mode = mode
             data = stringio(data)
         self.z.addfile(i, data)
 
     def done(self):
         self.z.close()
         if self.fileobj:
             self.fileobj.close()
 
 class tellable(object):
     '''provide tell method for zipfile.ZipFile when writing to http
     response file object.'''
 
     def __init__(self, fp):
         self.fp = fp
         self.offset = 0
 
     def __getattr__(self, key):
         return getattr(self.fp, key)
 
     def write(self, s):
         self.fp.write(s)
         self.offset += len(s)
 
     def tell(self):
         return self.offset
 
 class zipit(object):
     '''write archive to zip file or stream.  can write uncompressed,
     or compressed with deflate.'''
 
     def __init__(self, dest, mtime, compress=True):
         if not isinstance(dest, str):
             try:
                 dest.tell()
             except (AttributeError, IOError):
                 dest = tellable(dest)
         self.z = zipfile.ZipFile(dest, 'w',
                                  compress and zipfile.ZIP_DEFLATED or
                                  zipfile.ZIP_STORED)
 
         # Python's zipfile module emits deprecation warnings if we try
         # to store files with a date before 1980.
         epoch = 315532800 # calendar.timegm((1980, 1, 1, 0, 0, 0, 1, 1, 0))
         if mtime < epoch:
             mtime = epoch
 
         self.mtime = mtime
         self.date_time = time.gmtime(mtime)[:6]
 
     def addfile(self, name, mode, islink, data):
         i = zipfile.ZipInfo(name, self.date_time)
         i.compress_type = self.z.compression
         # unzip will not honor unix file modes unless file creator is
         # set to unix (id 3).
         i.create_system = 3
         ftype = _UNX_IFREG
         if islink:
             mode = 0o777
             ftype = _UNX_IFLNK
         i.external_attr = (mode | ftype) << 16
         # add "extended-timestamp" extra block, because zip archives
         # without this will be extracted with unexpected timestamp,
         # if TZ is not configured as GMT
         i.extra += struct.pack('<hhBl',
                                0x5455,     # block type: "extended-timestamp"
                                1 + 4,      # size of this block
                                1,          # "modification time is present"
                                int(self.mtime)) # last modification (UTC)
         self.z.writestr(i, data)
 
     def done(self):
         self.z.close()
 
 class fileit(object):
     '''write archive as files in directory.'''
 
     def __init__(self, name, mtime):
         self.basedir = name
         self.opener = scmutil.opener(self.basedir)
 
     def addfile(self, name, mode, islink, data):
         if islink:
             self.opener.symlink(data, name)
             return
         f = self.opener(name, "w", atomictemp=True)
         f.write(data)
         f.close()
         destfile = os.path.join(self.basedir, name)
         os.chmod(destfile, mode)
 
     def done(self):
         pass
 
 archivers = {
     'files': fileit,
     'tar': tarit,
     'tbz2': lambda name, mtime: tarit(name, mtime, 'bz2'),
     'tgz': lambda name, mtime: tarit(name, mtime, 'gz'),
     'uzip': lambda name, mtime: zipit(name, mtime, False),
     'zip': zipit,
     }
 
 def archive(repo, dest, node, kind, decode=True, matchfn=None,
             prefix='', mtime=None, subrepos=False):
     '''create archive of repo as it was at node.
 
     dest can be name of directory, name of archive file, or file
     object to write archive to.
 
     kind is type of archive to create.
 
     decode tells whether to put files through decode filters from
     hgrc.
 
     matchfn is function to filter names of files to write to archive.
 
     prefix is name of path to put before every archive member.'''
 
     if kind == 'files':
         if prefix:
             raise error.Abort(_('cannot give prefix when archiving to files'))
     else:
         prefix = tidyprefix(dest, kind, prefix)
 
     def write(name, mode, islink, getdata):
         data = getdata()
         if decode:
             data = repo.wwritedata(name, data)
         archiver.addfile(prefix + name, mode, islink, data)
 
     if kind not in archivers:
         raise error.Abort(_("unknown archive type '%s'") % kind)
 
     ctx = repo[node]
     archiver = archivers[kind](dest, mtime or ctx.date()[0])
 
     if repo.ui.configbool("ui", "archivemeta", True):
         name = '.hg_archival.txt'
         if not matchfn or matchfn(name):
             write(name, 0o644, False, lambda: buildmetadata(ctx))
 
     if matchfn:
         files = [f for f in ctx.manifest().keys() if matchfn(f)]
     else:
         files = ctx.manifest().keys()
     total = len(files)
     if total:
         files.sort()
         repo.ui.progress(_('archiving'), 0, unit=_('files'), total=total)
         for i, f in enumerate(files):
             ff = ctx.flags(f)
             write(f, 'x' in ff and 0o755 or 0o644, 'l' in ff, ctx[f].data)
             repo.ui.progress(_('archiving'), i + 1, item=f,
                              unit=_('files'), total=total)
         repo.ui.progress(_('archiving'), None)
 
     if subrepos:
         for subpath in sorted(ctx.substate):
             sub = ctx.workingsub(subpath)
             submatch = matchmod.subdirmatcher(subpath, matchfn)
-            total += sub.archive(archiver, prefix, submatch)
+            total += sub.archive(archiver, prefix, submatch, decode)
 
     if total == 0:
         raise error.Abort(_('no files match the archive pattern'))
 
     archiver.done()
     return total

mercurial/subrepo.py

@@ -1,1960 +1,1964 b''
 # subrepo.py - sub-repository handling for Mercurial
 #
 # Copyright 2009-2010 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 copy
 import errno
 import hashlib
 import os
 import posixpath
 import re
 import stat
 import subprocess
 import sys
 import tarfile
 import xml.dom.minidom
 
 
 from .i18n import _
 from . import (
     cmdutil,
     config,
     encoding,
     error,
     exchange,
     filemerge,
     match as matchmod,
     node,
     pathutil,
     phases,
     pycompat,
     scmutil,
     util,
 )
 
 hg = None
 propertycache = util.propertycache
 
 nullstate = ('', '', 'empty')
 
 def _expandedabspath(path):
     '''
     get a path or url and if it is a path expand it and return an absolute path
     '''
     expandedpath = util.urllocalpath(util.expandpath(path))
     u = util.url(expandedpath)
     if not u.scheme:
         path = util.normpath(os.path.abspath(u.path))
     return path
 
 def _getstorehashcachename(remotepath):
     '''get a unique filename for the store hash cache of a remote repository'''
     return hashlib.sha1(_expandedabspath(remotepath)).hexdigest()[0:12]
 
 class SubrepoAbort(error.Abort):
     """Exception class used to avoid handling a subrepo error more than once"""
     def __init__(self, *args, **kw):
         self.subrepo = kw.pop('subrepo', None)
         self.cause = kw.pop('cause', None)
         error.Abort.__init__(self, *args, **kw)
 
 def annotatesubrepoerror(func):
     def decoratedmethod(self, *args, **kargs):
         try:
             res = func(self, *args, **kargs)
         except SubrepoAbort as ex:
             # This exception has already been handled
             raise ex
         except error.Abort as ex:
             subrepo = subrelpath(self)
             errormsg = str(ex) + ' ' + _('(in subrepo %s)') % subrepo
             # avoid handling this exception by raising a SubrepoAbort exception
             raise SubrepoAbort(errormsg, hint=ex.hint, subrepo=subrepo,
                                cause=sys.exc_info())
         return res
     return decoratedmethod
 
 def state(ctx, ui):
     """return a state dict, mapping subrepo paths configured in .hgsub
     to tuple: (source from .hgsub, revision from .hgsubstate, kind
     (key in types dict))
     """
     p = config.config()
     repo = ctx.repo()
     def read(f, sections=None, remap=None):
         if f in ctx:
             try:
                 data = ctx[f].data()
             except IOError as err:
                 if err.errno != errno.ENOENT:
                     raise
                 # handle missing subrepo spec files as removed
                 ui.warn(_("warning: subrepo spec file \'%s\' not found\n") %
                         repo.pathto(f))
                 return
             p.parse(f, data, sections, remap, read)
         else:
             raise error.Abort(_("subrepo spec file \'%s\' not found") %
                              repo.pathto(f))
     if '.hgsub' in ctx:
         read('.hgsub')
 
     for path, src in ui.configitems('subpaths'):
         p.set('subpaths', path, src, ui.configsource('subpaths', path))
 
     rev = {}
     if '.hgsubstate' in ctx:
         try:
             for i, l in enumerate(ctx['.hgsubstate'].data().splitlines()):
                 l = l.lstrip()
                 if not l:
                     continue
                 try:
                     revision, path = l.split(" ", 1)
                 except ValueError:
                     raise error.Abort(_("invalid subrepository revision "
                                        "specifier in \'%s\' line %d")
                                      % (repo.pathto('.hgsubstate'), (i + 1)))
                 rev[path] = revision
         except IOError as err:
             if err.errno != errno.ENOENT:
                 raise
 
     def remap(src):
         for pattern, repl in p.items('subpaths'):
             # Turn r'C:\foo\bar' into r'C:\\foo\\bar' since re.sub
             # does a string decode.
             repl = repl.encode('string-escape')
             # However, we still want to allow back references to go
             # through unharmed, so we turn r'\\1' into r'\1'. Again,
             # extra escapes are needed because re.sub string decodes.
             repl = re.sub(r'\\\\([0-9]+)', r'\\\1', repl)
             try:
                 src = re.sub(pattern, repl, src, 1)
             except re.error as e:
                 raise error.Abort(_("bad subrepository pattern in %s: %s")
                                  % (p.source('subpaths', pattern), e))
         return src
 
     state = {}
     for path, src in p[''].items():
         kind = 'hg'
         if src.startswith('['):
             if ']' not in src:
                 raise error.Abort(_('missing ] in subrepo source'))
             kind, src = src.split(']', 1)
             kind = kind[1:]
             src = src.lstrip() # strip any extra whitespace after ']'
 
         if not util.url(src).isabs():
             parent = _abssource(repo, abort=False)
             if parent:
                 parent = util.url(parent)
                 parent.path = posixpath.join(parent.path or '', src)
                 parent.path = posixpath.normpath(parent.path)
                 joined = str(parent)
                 # Remap the full joined path and use it if it changes,
                 # else remap the original source.
                 remapped = remap(joined)
                 if remapped == joined:
                     src = remap(src)
                 else:
                     src = remapped
 
         src = remap(src)
         state[util.pconvert(path)] = (src.strip(), rev.get(path, ''), kind)
 
     return state
 
 def writestate(repo, state):
     """rewrite .hgsubstate in (outer) repo with these subrepo states"""
     lines = ['%s %s\n' % (state[s][1], s) for s in sorted(state)
                                                 if state[s][1] != nullstate[1]]
     repo.wwrite('.hgsubstate', ''.join(lines), '')
 
 def submerge(repo, wctx, mctx, actx, overwrite, labels=None):
     """delegated from merge.applyupdates: merging of .hgsubstate file
     in working context, merging context and ancestor context"""
     if mctx == actx: # backwards?
         actx = wctx.p1()
     s1 = wctx.substate
     s2 = mctx.substate
     sa = actx.substate
     sm = {}
 
     repo.ui.debug("subrepo merge %s %s %s\n" % (wctx, mctx, actx))
 
     def debug(s, msg, r=""):
         if r:
             r = "%s:%s:%s" % r
         repo.ui.debug("  subrepo %s: %s %s\n" % (s, msg, r))
 
     for s, l in sorted(s1.iteritems()):
         a = sa.get(s, nullstate)
         ld = l # local state with possible dirty flag for compares
         if wctx.sub(s).dirty():
             ld = (l[0], l[1] + "+")
         if wctx == actx: # overwrite
             a = ld
 
         if s in s2:
             prompts = filemerge.partextras(labels)
             prompts['s'] = s
             r = s2[s]
             if ld == r or r == a: # no change or local is newer
                 sm[s] = l
                 continue
             elif ld == a: # other side changed
                 debug(s, "other changed, get", r)
                 wctx.sub(s).get(r, overwrite)
                 sm[s] = r
             elif ld[0] != r[0]: # sources differ
                 prompts['lo'] = l[0]
                 prompts['ro'] = r[0]
                 if repo.ui.promptchoice(
                     _(' subrepository sources for %(s)s differ\n'
                       'use (l)ocal%(l)s source (%(lo)s)'
                       ' or (r)emote%(o)s source (%(ro)s)?'
                       '$$ &Local $$ &Remote') % prompts, 0):
                     debug(s, "prompt changed, get", r)
                     wctx.sub(s).get(r, overwrite)
                     sm[s] = r
             elif ld[1] == a[1]: # local side is unchanged
                 debug(s, "other side changed, get", r)
                 wctx.sub(s).get(r, overwrite)
                 sm[s] = r
             else:
                 debug(s, "both sides changed")
                 srepo = wctx.sub(s)
                 prompts['sl'] = srepo.shortid(l[1])
                 prompts['sr'] = srepo.shortid(r[1])
                 option = repo.ui.promptchoice(
                     _(' subrepository %(s)s diverged (local revision: %(sl)s, '
                       'remote revision: %(sr)s)\n'
                       '(M)erge, keep (l)ocal%(l)s or keep (r)emote%(o)s?'
                       '$$ &Merge $$ &Local $$ &Remote')
                     % prompts, 0)
                 if option == 0:
                     wctx.sub(s).merge(r)
                     sm[s] = l
                     debug(s, "merge with", r)
                 elif option == 1:
                     sm[s] = l
                     debug(s, "keep local subrepo revision", l)
                 else:
                     wctx.sub(s).get(r, overwrite)
                     sm[s] = r
                     debug(s, "get remote subrepo revision", r)
         elif ld == a: # remote removed, local unchanged
             debug(s, "remote removed, remove")
             wctx.sub(s).remove()
         elif a == nullstate: # not present in remote or ancestor
             debug(s, "local added, keep")
             sm[s] = l
             continue
         else:
             if repo.ui.promptchoice(
                 _(' local%(l)s changed subrepository %(s)s'
                   ' which remote%(o)s removed\n'
                   'use (c)hanged version or (d)elete?'
                   '$$ &Changed $$ &Delete') % prompts, 0):
                 debug(s, "prompt remove")
                 wctx.sub(s).remove()
 
     for s, r in sorted(s2.items()):
         if s in s1:
             continue
         elif s not in sa:
             debug(s, "remote added, get", r)
             mctx.sub(s).get(r)
             sm[s] = r
         elif r != sa[s]:
             if repo.ui.promptchoice(
                 _(' remote%(o)s changed subrepository %(s)s'
                   ' which local%(l)s removed\n'
                   'use (c)hanged version or (d)elete?'
                   '$$ &Changed $$ &Delete') % prompts, 0) == 0:
                 debug(s, "prompt recreate", r)
                 mctx.sub(s).get(r)
                 sm[s] = r
 
     # record merged .hgsubstate
     writestate(repo, sm)
     return sm
 
 def _updateprompt(ui, sub, dirty, local, remote):
     if dirty:
         msg = (_(' subrepository sources for %s differ\n'
                  'use (l)ocal source (%s) or (r)emote source (%s)?'
                  '$$ &Local $$ &Remote')
                % (subrelpath(sub), local, remote))
     else:
         msg = (_(' subrepository sources for %s differ (in checked out '
                  'version)\n'
                  'use (l)ocal source (%s) or (r)emote source (%s)?'
                  '$$ &Local $$ &Remote')
                % (subrelpath(sub), local, remote))
     return ui.promptchoice(msg, 0)
 
 def reporelpath(repo):
     """return path to this (sub)repo as seen from outermost repo"""
     parent = repo
     while util.safehasattr(parent, '_subparent'):
         parent = parent._subparent
     return repo.root[len(pathutil.normasprefix(parent.root)):]
 
 def subrelpath(sub):
     """return path to this subrepo as seen from outermost repo"""
     return sub._relpath
 
 def _abssource(repo, push=False, abort=True):
     """return pull/push path of repo - either based on parent repo .hgsub info
     or on the top repo config. Abort or return None if no source found."""
     if util.safehasattr(repo, '_subparent'):
         source = util.url(repo._subsource)
         if source.isabs():
             return str(source)
         source.path = posixpath.normpath(source.path)
         parent = _abssource(repo._subparent, push, abort=False)
         if parent:
             parent = util.url(util.pconvert(parent))
             parent.path = posixpath.join(parent.path or '', source.path)
             parent.path = posixpath.normpath(parent.path)
             return str(parent)
     else: # recursion reached top repo
         if util.safehasattr(repo, '_subtoppath'):
             return repo._subtoppath
         if push and repo.ui.config('paths', 'default-push'):
             return repo.ui.config('paths', 'default-push')
         if repo.ui.config('paths', 'default'):
             return repo.ui.config('paths', 'default')
         if repo.shared():
             # chop off the .hg component to get the default path form
             return os.path.dirname(repo.sharedpath)
     if abort:
         raise error.Abort(_("default path for subrepository not found"))
 
 def _sanitize(ui, vfs, ignore):
     for dirname, dirs, names in vfs.walk():
         for i, d in enumerate(dirs):
             if d.lower() == ignore:
                 del dirs[i]
                 break
         if vfs.basename(dirname).lower() != '.hg':
             continue
         for f in names:
             if f.lower() == 'hgrc':
                 ui.warn(_("warning: removing potentially hostile 'hgrc' "
                           "in '%s'\n") % vfs.join(dirname))
                 vfs.unlink(vfs.reljoin(dirname, f))
 
 def subrepo(ctx, path, allowwdir=False, allowcreate=True):
     """return instance of the right subrepo class for subrepo in path"""
     # subrepo inherently violates our import layering rules
     # because it wants to make repo objects from deep inside the stack
     # so we manually delay the circular imports to not break
     # scripts that don't use our demand-loading
     global hg
     from . import hg as h
     hg = h
 
     pathutil.pathauditor(ctx.repo().root)(path)
     state = ctx.substate[path]
     if state[2] not in types:
         raise error.Abort(_('unknown subrepo type %s') % state[2])
     if allowwdir:
         state = (state[0], ctx.subrev(path), state[2])
     return types[state[2]](ctx, path, state[:2], allowcreate)
 
 def nullsubrepo(ctx, path, pctx):
     """return an empty subrepo in pctx for the extant subrepo in ctx"""
     # subrepo inherently violates our import layering rules
     # because it wants to make repo objects from deep inside the stack
     # so we manually delay the circular imports to not break
     # scripts that don't use our demand-loading
     global hg
     from . import hg as h
     hg = h
 
     pathutil.pathauditor(ctx.repo().root)(path)
     state = ctx.substate[path]
     if state[2] not in types:
         raise error.Abort(_('unknown subrepo type %s') % state[2])
     subrev = ''
     if state[2] == 'hg':
         subrev = "0" * 40
     return types[state[2]](pctx, path, (state[0], subrev), True)
 
 def newcommitphase(ui, ctx):
     commitphase = phases.newcommitphase(ui)
     substate = getattr(ctx, "substate", None)
     if not substate:
         return commitphase
     check = ui.config('phases', 'checksubrepos', 'follow')
     if check not in ('ignore', 'follow', 'abort'):
         raise error.Abort(_('invalid phases.checksubrepos configuration: %s')
                          % (check))
     if check == 'ignore':
         return commitphase
     maxphase = phases.public
     maxsub = None
     for s in sorted(substate):
         sub = ctx.sub(s)
         subphase = sub.phase(substate[s][1])
         if maxphase < subphase:
             maxphase = subphase
             maxsub = s
     if commitphase < maxphase:
         if check == 'abort':
             raise error.Abort(_("can't commit in %s phase"
                                " conflicting %s from subrepository %s") %
                              (phases.phasenames[commitphase],
                               phases.phasenames[maxphase], maxsub))
         ui.warn(_("warning: changes are committed in"
                   " %s phase from subrepository %s\n") %
                 (phases.phasenames[maxphase], maxsub))
         return maxphase
     return commitphase
 
 # subrepo classes need to implement the following abstract class:
 
 class abstractsubrepo(object):
 
     def __init__(self, ctx, path):
         """Initialize abstractsubrepo part
 
         ``ctx`` is the context referring this subrepository in the
         parent repository.
 
         ``path`` is the path to this subrepository as seen from
         innermost repository.
         """
         self.ui = ctx.repo().ui
         self._ctx = ctx
         self._path = path
 
     def storeclean(self, path):
         """
         returns true if the repository has not changed since it was last
         cloned from or pushed to a given repository.
         """
         return False
 
     def dirty(self, ignoreupdate=False):
         """returns true if the dirstate of the subrepo is dirty or does not
         match current stored state. If ignoreupdate is true, only check
         whether the subrepo has uncommitted changes in its dirstate.
         """
         raise NotImplementedError
 
     def dirtyreason(self, ignoreupdate=False):
         """return reason string if it is ``dirty()``
 
         Returned string should have enough information for the message
         of exception.
 
         This returns None, otherwise.
         """
         if self.dirty(ignoreupdate=ignoreupdate):
             return _("uncommitted changes in subrepository '%s'"
                      ) % subrelpath(self)
 
     def bailifchanged(self, ignoreupdate=False, hint=None):
         """raise Abort if subrepository is ``dirty()``
         """
         dirtyreason = self.dirtyreason(ignoreupdate=ignoreupdate)
         if dirtyreason:
             raise error.Abort(dirtyreason, hint=hint)
 
     def basestate(self):
         """current working directory base state, disregarding .hgsubstate
         state and working directory modifications"""
         raise NotImplementedError
 
     def checknested(self, path):
         """check if path is a subrepository within this repository"""
         return False
 
     def commit(self, text, user, date):
         """commit the current changes to the subrepo with the given
         log message. Use given user and date if possible. Return the
         new state of the subrepo.
         """
         raise NotImplementedError
 
     def phase(self, state):
         """returns phase of specified state in the subrepository.
         """
         return phases.public
 
     def remove(self):
         """remove the subrepo
 
         (should verify the dirstate is not dirty first)
         """
         raise NotImplementedError
 
     def get(self, state, overwrite=False):
         """run whatever commands are needed to put the subrepo into
         this state
         """
         raise NotImplementedError
 
     def merge(self, state):
         """merge currently-saved state with the new state."""
         raise NotImplementedError
 
     def push(self, opts):
         """perform whatever action is analogous to 'hg push'
 
         This may be a no-op on some systems.
         """
         raise NotImplementedError
 
     def add(self, ui, match, prefix, explicitonly, **opts):
         return []
 
     def addremove(self, matcher, prefix, opts, dry_run, similarity):
         self.ui.warn("%s: %s" % (prefix, _("addremove is not supported")))
         return 1
 
     def cat(self, match, prefix, **opts):
         return 1
 
     def status(self, rev2, **opts):
         return scmutil.status([], [], [], [], [], [], [])
 
     def diff(self, ui, diffopts, node2, match, prefix, **opts):
         pass
 
     def outgoing(self, ui, dest, opts):
         return 1
 
     def incoming(self, ui, source, opts):
         return 1
 
     def files(self):
         """return filename iterator"""
         raise NotImplementedError
 
-    def filedata(self, name):
-        """return file data"""
+    def filedata(self, name, decode):
+        """return file data, optionally passed through repo decoders"""
         raise NotImplementedError
 
     def fileflags(self, name):
         """return file flags"""
         return ''
 
     def getfileset(self, expr):
         """Resolve the fileset expression for this repo"""
         return set()
 
     def printfiles(self, ui, m, fm, fmt, subrepos):
         """handle the files command for this subrepo"""
         return 1
 
-    def archive(self, archiver, prefix, match=None):
+    def archive(self, archiver, prefix, match=None, decode=True):
         if match is not None:
             files = [f for f in self.files() if match(f)]
         else:
             files = self.files()
         total = len(files)
         relpath = subrelpath(self)
         self.ui.progress(_('archiving (%s)') % relpath, 0,
                          unit=_('files'), total=total)
         for i, name in enumerate(files):
             flags = self.fileflags(name)
             mode = 'x' in flags and 0o755 or 0o644
             symlink = 'l' in flags
             archiver.addfile(prefix + self._path + '/' + name,
-                             mode, symlink, self.filedata(name))
+                             mode, symlink, self.filedata(name, decode))
             self.ui.progress(_('archiving (%s)') % relpath, i + 1,
                              unit=_('files'), total=total)
         self.ui.progress(_('archiving (%s)') % relpath, None)
         return total
 
     def walk(self, match):
         '''
         walk recursively through the directory tree, finding all files
         matched by the match function
         '''
         pass
 
     def forget(self, match, prefix):
         return ([], [])
 
     def removefiles(self, matcher, prefix, after, force, subrepos, warnings):
         """remove the matched files from the subrepository and the filesystem,
         possibly by force and/or after the file has been removed from the
         filesystem.  Return 0 on success, 1 on any warning.
         """
         warnings.append(_("warning: removefiles not implemented (%s)")
                         % self._path)
         return 1
 
     def revert(self, substate, *pats, **opts):
         self.ui.warn(_('%s: reverting %s subrepos is unsupported\n') \
             % (substate[0], substate[2]))
         return []
 
     def shortid(self, revid):
         return revid
 
     def verify(self):
         '''verify the integrity of the repository.  Return 0 on success or
         warning, 1 on any error.
         '''
         return 0
 
     @propertycache
     def wvfs(self):
         """return vfs to access the working directory of this subrepository
         """
         return scmutil.vfs(self._ctx.repo().wvfs.join(self._path))
 
     @propertycache
     def _relpath(self):
         """return path to this subrepository as seen from outermost repository
         """
         return self.wvfs.reljoin(reporelpath(self._ctx.repo()), self._path)
 
 class hgsubrepo(abstractsubrepo):
     def __init__(self, ctx, path, state, allowcreate):
         super(hgsubrepo, self).__init__(ctx, path)
         self._state = state
         r = ctx.repo()
         root = r.wjoin(path)
         create = allowcreate and not r.wvfs.exists('%s/.hg' % path)
         self._repo = hg.repository(r.baseui, root, create=create)
 
         # Propagate the parent's --hidden option
         if r is r.unfiltered():
             self._repo = self._repo.unfiltered()
 
         self.ui = self._repo.ui
         for s, k in [('ui', 'commitsubrepos')]:
             v = r.ui.config(s, k)
             if v:
                 self.ui.setconfig(s, k, v, 'subrepo')
         # internal config: ui._usedassubrepo
         self.ui.setconfig('ui', '_usedassubrepo', 'True', 'subrepo')
         self._initrepo(r, state[0], create)
 
     def storeclean(self, path):
         with self._repo.lock():
             return self._storeclean(path)
 
     def _storeclean(self, path):
         clean = True
         itercache = self._calcstorehash(path)
         for filehash in self._readstorehashcache(path):
             if filehash != next(itercache, None):
                 clean = False
                 break
         if clean:
             # if not empty:
             # the cached and current pull states have a different size
             clean = next(itercache, None) is None
         return clean
 
     def _calcstorehash(self, remotepath):
         '''calculate a unique "store hash"
 
         This method is used to to detect when there are changes that may
         require a push to a given remote path.'''
         # sort the files that will be hashed in increasing (likely) file size
         filelist = ('bookmarks', 'store/phaseroots', 'store/00changelog.i')
         yield '# %s\n' % _expandedabspath(remotepath)
         vfs = self._repo.vfs
         for relname in filelist:
             filehash = hashlib.sha1(vfs.tryread(relname)).hexdigest()
             yield '%s = %s\n' % (relname, filehash)
 
     @propertycache
     def _cachestorehashvfs(self):
         return scmutil.vfs(self._repo.join('cache/storehash'))
 
     def _readstorehashcache(self, remotepath):
         '''read the store hash cache for a given remote repository'''
         cachefile = _getstorehashcachename(remotepath)
         return self._cachestorehashvfs.tryreadlines(cachefile, 'r')
 
     def _cachestorehash(self, remotepath):
         '''cache the current store hash
 
         Each remote repo requires its own store hash cache, because a subrepo
         store may be "clean" versus a given remote repo, but not versus another
         '''
         cachefile = _getstorehashcachename(remotepath)
         with self._repo.lock():
             storehash = list(self._calcstorehash(remotepath))
             vfs = self._cachestorehashvfs
             vfs.writelines(cachefile, storehash, mode='w', notindexed=True)
 
     def _getctx(self):
         '''fetch the context for this subrepo revision, possibly a workingctx
         '''
         if self._ctx.rev() is None:
             return self._repo[None]  # workingctx if parent is workingctx
         else:
             rev = self._state[1]
             return self._repo[rev]
 
     @annotatesubrepoerror
     def _initrepo(self, parentrepo, source, create):
         self._repo._subparent = parentrepo
         self._repo._subsource = source
 
         if create:
             lines = ['[paths]\n']
 
             def addpathconfig(key, value):
                 if value:
                     lines.append('%s = %s\n' % (key, value))
                     self.ui.setconfig('paths', key, value, 'subrepo')
 
             defpath = _abssource(self._repo, abort=False)
             defpushpath = _abssource(self._repo, True, abort=False)
             addpathconfig('default', defpath)
             if defpath != defpushpath:
                 addpathconfig('default-push', defpushpath)
 
             fp = self._repo.vfs("hgrc", "w", text=True)
             try:
                 fp.write(''.join(lines))
             finally:
                 fp.close()
 
     @annotatesubrepoerror
     def add(self, ui, match, prefix, explicitonly, **opts):
         return cmdutil.add(ui, self._repo, match,
                            self.wvfs.reljoin(prefix, self._path),
                            explicitonly, **opts)
 
     @annotatesubrepoerror
     def addremove(self, m, prefix, opts, dry_run, similarity):
         # In the same way as sub directories are processed, once in a subrepo,
         # always entry any of its subrepos.  Don't corrupt the options that will
         # be used to process sibling subrepos however.
         opts = copy.copy(opts)
         opts['subrepos'] = True
         return scmutil.addremove(self._repo, m,
                                  self.wvfs.reljoin(prefix, self._path), opts,
                                  dry_run, similarity)
 
     @annotatesubrepoerror
     def cat(self, match, prefix, **opts):
         rev = self._state[1]
         ctx = self._repo[rev]
         return cmdutil.cat(self.ui, self._repo, ctx, match, prefix, **opts)
 
     @annotatesubrepoerror
     def status(self, rev2, **opts):
         try:
             rev1 = self._state[1]
             ctx1 = self._repo[rev1]
             ctx2 = self._repo[rev2]
             return self._repo.status(ctx1, ctx2, **opts)
         except error.RepoLookupError as inst:
             self.ui.warn(_('warning: error "%s" in subrepository "%s"\n')
                          % (inst, subrelpath(self)))
             return scmutil.status([], [], [], [], [], [], [])
 
     @annotatesubrepoerror
     def diff(self, ui, diffopts, node2, match, prefix, **opts):
         try:
             node1 = node.bin(self._state[1])
             # We currently expect node2 to come from substate and be
             # in hex format
             if node2 is not None:
                 node2 = node.bin(node2)
             cmdutil.diffordiffstat(ui, self._repo, diffopts,
                                    node1, node2, match,
                                    prefix=posixpath.join(prefix, self._path),
                                    listsubrepos=True, **opts)
         except error.RepoLookupError as inst:
             self.ui.warn(_('warning: error "%s" in subrepository "%s"\n')
                           % (inst, subrelpath(self)))
 
     @annotatesubrepoerror
-    def archive(self, archiver, prefix, match=None):
+    def archive(self, archiver, prefix, match=None, decode=True):
         self._get(self._state + ('hg',))
         total = abstractsubrepo.archive(self, archiver, prefix, match)
         rev = self._state[1]
         ctx = self._repo[rev]
         for subpath in ctx.substate:
             s = subrepo(ctx, subpath, True)
             submatch = matchmod.subdirmatcher(subpath, match)
-            total += s.archive(archiver, prefix + self._path + '/', submatch)
+            total += s.archive(archiver, prefix + self._path + '/', submatch,
+                               decode)
         return total
 
     @annotatesubrepoerror
     def dirty(self, ignoreupdate=False):
         r = self._state[1]
         if r == '' and not ignoreupdate: # no state recorded
             return True
         w = self._repo[None]
         if r != w.p1().hex() and not ignoreupdate:
             # different version checked out
             return True
         return w.dirty() # working directory changed
 
     def basestate(self):
         return self._repo['.'].hex()
 
     def checknested(self, path):
         return self._repo._checknested(self._repo.wjoin(path))
 
     @annotatesubrepoerror
     def commit(self, text, user, date):
         # don't bother committing in the subrepo if it's only been
         # updated
         if not self.dirty(True):
             return self._repo['.'].hex()
         self.ui.debug("committing subrepo %s\n" % subrelpath(self))
         n = self._repo.commit(text, user, date)
         if not n:
             return self._repo['.'].hex() # different version checked out
         return node.hex(n)
 
     @annotatesubrepoerror
     def phase(self, state):
         return self._repo[state].phase()
 
     @annotatesubrepoerror
     def remove(self):
         # we can't fully delete the repository as it may contain
         # local-only history
         self.ui.note(_('removing subrepo %s\n') % subrelpath(self))
         hg.clean(self._repo, node.nullid, False)
 
     def _get(self, state):
         source, revision, kind = state
         if revision in self._repo.unfiltered():
             return True
         self._repo._subsource = source
         srcurl = _abssource(self._repo)
         other = hg.peer(self._repo, {}, srcurl)
         if len(self._repo) == 0:
             self.ui.status(_('cloning subrepo %s from %s\n')
                            % (subrelpath(self), srcurl))
             parentrepo = self._repo._subparent
             # use self._repo.vfs instead of self.wvfs to remove .hg only
             self._repo.vfs.rmtree()
             other, cloned = hg.clone(self._repo._subparent.baseui, {},
                                      other, self._repo.root,
                                      update=False)
             self._repo = cloned.local()
             self._initrepo(parentrepo, source, create=True)
             self._cachestorehash(srcurl)
         else:
             self.ui.status(_('pulling subrepo %s from %s\n')
                            % (subrelpath(self), srcurl))
             cleansub = self.storeclean(srcurl)
             exchange.pull(self._repo, other)
             if cleansub:
                 # keep the repo clean after pull
                 self._cachestorehash(srcurl)
         return False
 
     @annotatesubrepoerror
     def get(self, state, overwrite=False):
         inrepo = self._get(state)
         source, revision, kind = state
         repo = self._repo
         repo.ui.debug("getting subrepo %s\n" % self._path)
         if inrepo:
             urepo = repo.unfiltered()
             ctx = urepo[revision]
             if ctx.hidden():
                 urepo.ui.warn(
                     _('revision %s in subrepo %s is hidden\n') \
                     % (revision[0:12], self._path))
                 repo = urepo
         hg.updaterepo(repo, revision, overwrite)
 
     @annotatesubrepoerror
     def merge(self, state):
         self._get(state)
         cur = self._repo['.']
         dst = self._repo[state[1]]
         anc = dst.ancestor(cur)
 
         def mergefunc():
             if anc == cur and dst.branch() == cur.branch():
                 self.ui.debug("updating subrepo %s\n" % subrelpath(self))
                 hg.update(self._repo, state[1])
             elif anc == dst:
                 self.ui.debug("skipping subrepo %s\n" % subrelpath(self))
             else:
                 self.ui.debug("merging subrepo %s\n" % subrelpath(self))
                 hg.merge(self._repo, state[1], remind=False)
 
         wctx = self._repo[None]
         if self.dirty():
             if anc != dst:
                 if _updateprompt(self.ui, self, wctx.dirty(), cur, dst):
                     mergefunc()
             else:
                 mergefunc()
         else:
             mergefunc()
 
     @annotatesubrepoerror
     def push(self, opts):
         force = opts.get('force')
         newbranch = opts.get('new_branch')
         ssh = opts.get('ssh')
 
         # push subrepos depth-first for coherent ordering
         c = self._repo['']
         subs = c.substate # only repos that are committed
         for s in sorted(subs):
             if c.sub(s).push(opts) == 0:
                 return False
 
         dsturl = _abssource(self._repo, True)
         if not force:
             if self.storeclean(dsturl):
                 self.ui.status(
                     _('no changes made to subrepo %s since last push to %s\n')
                     % (subrelpath(self), dsturl))
                 return None
         self.ui.status(_('pushing subrepo %s to %s\n') %
             (subrelpath(self), dsturl))
         other = hg.peer(self._repo, {'ssh': ssh}, dsturl)
         res = exchange.push(self._repo, other, force, newbranch=newbranch)
 
         # the repo is now clean
         self._cachestorehash(dsturl)
         return res.cgresult
 
     @annotatesubrepoerror
     def outgoing(self, ui, dest, opts):
         if 'rev' in opts or 'branch' in opts:
             opts = copy.copy(opts)
             opts.pop('rev', None)
             opts.pop('branch', None)
         return hg.outgoing(ui, self._repo, _abssource(self._repo, True), opts)
 
     @annotatesubrepoerror
     def incoming(self, ui, source, opts):
         if 'rev' in opts or 'branch' in opts:
             opts = copy.copy(opts)
             opts.pop('rev', None)
             opts.pop('branch', None)
         return hg.incoming(ui, self._repo, _abssource(self._repo, False), opts)
 
     @annotatesubrepoerror
     def files(self):
         rev = self._state[1]
         ctx = self._repo[rev]
         return ctx.manifest().keys()
 
-    def filedata(self, name):
+    def filedata(self, name, decode):
         rev = self._state[1]
-        return self._repo[rev][name].data()
+        data = self._repo[rev][name].data()
+        if decode:
+            data = self._repo.wwritedata(name, data)
+        return data
 
     def fileflags(self, name):
         rev = self._state[1]
         ctx = self._repo[rev]
         return ctx.flags(name)
 
     @annotatesubrepoerror
     def printfiles(self, ui, m, fm, fmt, subrepos):
         # If the parent context is a workingctx, use the workingctx here for
         # consistency.
         if self._ctx.rev() is None:
             ctx = self._repo[None]
         else:
             rev = self._state[1]
             ctx = self._repo[rev]
         return cmdutil.files(ui, ctx, m, fm, fmt, subrepos)
 
     @annotatesubrepoerror
     def getfileset(self, expr):
         if self._ctx.rev() is None:
             ctx = self._repo[None]
         else:
             rev = self._state[1]
             ctx = self._repo[rev]
 
         files = ctx.getfileset(expr)
 
         for subpath in ctx.substate:
             sub = ctx.sub(subpath)
 
             try:
                 files.extend(subpath + '/' + f for f in sub.getfileset(expr))
             except error.LookupError:
                 self.ui.status(_("skipping missing subrepository: %s\n")
                                % self.wvfs.reljoin(reporelpath(self), subpath))
         return files
 
     def walk(self, match):
         ctx = self._repo[None]
         return ctx.walk(match)
 
     @annotatesubrepoerror
     def forget(self, match, prefix):
         return cmdutil.forget(self.ui, self._repo, match,
                               self.wvfs.reljoin(prefix, self._path), True)
 
     @annotatesubrepoerror
     def removefiles(self, matcher, prefix, after, force, subrepos, warnings):
         return cmdutil.remove(self.ui, self._repo, matcher,
                               self.wvfs.reljoin(prefix, self._path),
                               after, force, subrepos)
 
     @annotatesubrepoerror
     def revert(self, substate, *pats, **opts):
         # reverting a subrepo is a 2 step process:
         # 1. if the no_backup is not set, revert all modified
         #    files inside the subrepo
         # 2. update the subrepo to the revision specified in
         #    the corresponding substate dictionary
         self.ui.status(_('reverting subrepo %s\n') % substate[0])
         if not opts.get('no_backup'):
             # Revert all files on the subrepo, creating backups
             # Note that this will not recursively revert subrepos
             # We could do it if there was a set:subrepos() predicate
             opts = opts.copy()
             opts['date'] = None
             opts['rev'] = substate[1]
 
             self.filerevert(*pats, **opts)
 
         # Update the repo to the revision specified in the given substate
         if not opts.get('dry_run'):
             self.get(substate, overwrite=True)
 
     def filerevert(self, *pats, **opts):
         ctx = self._repo[opts['rev']]
         parents = self._repo.dirstate.parents()
         if opts.get('all'):
             pats = ['set:modified()']
         else:
             pats = []
         cmdutil.revert(self.ui, self._repo, ctx, parents, *pats, **opts)
 
     def shortid(self, revid):
         return revid[:12]
 
     def verify(self):
         try:
             rev = self._state[1]
             ctx = self._repo.unfiltered()[rev]
             if ctx.hidden():
                 # Since hidden revisions aren't pushed/pulled, it seems worth an
                 # explicit warning.
                 ui = self._repo.ui
                 ui.warn(_("subrepo '%s' is hidden in revision %s\n") %
                         (self._relpath, node.short(self._ctx.node())))
             return 0
         except error.RepoLookupError:
             # A missing subrepo revision may be a case of needing to pull it, so
             # don't treat this as an error.
             self._repo.ui.warn(_("subrepo '%s' not found in revision %s\n") %
                                (self._relpath, node.short(self._ctx.node())))
             return 0
 
     @propertycache
     def wvfs(self):
         """return own wvfs for efficiency and consistency
         """
         return self._repo.wvfs
 
     @propertycache
     def _relpath(self):
         """return path to this subrepository as seen from outermost repository
         """
         # Keep consistent dir separators by avoiding vfs.join(self._path)
         return reporelpath(self._repo)
 
 class svnsubrepo(abstractsubrepo):
     def __init__(self, ctx, path, state, allowcreate):
         super(svnsubrepo, self).__init__(ctx, path)
         self._state = state
         self._exe = util.findexe('svn')
         if not self._exe:
             raise error.Abort(_("'svn' executable not found for subrepo '%s'")
                              % self._path)
 
     def _svncommand(self, commands, filename='', failok=False):
         cmd = [self._exe]
         extrakw = {}
         if not self.ui.interactive():
             # Making stdin be a pipe should prevent svn from behaving
             # interactively even if we can't pass --non-interactive.
             extrakw['stdin'] = subprocess.PIPE
             # Starting in svn 1.5 --non-interactive is a global flag
             # instead of being per-command, but we need to support 1.4 so
             # we have to be intelligent about what commands take
             # --non-interactive.
             if commands[0] in ('update', 'checkout', 'commit'):
                 cmd.append('--non-interactive')
         cmd.extend(commands)
         if filename is not None:
             path = self.wvfs.reljoin(self._ctx.repo().origroot,
                                      self._path, filename)
             cmd.append(path)
         env = dict(encoding.environ)
         # Avoid localized output, preserve current locale for everything else.
         lc_all = env.get('LC_ALL')
         if lc_all:
             env['LANG'] = lc_all
             del env['LC_ALL']
         env['LC_MESSAGES'] = 'C'
         p = subprocess.Popen(cmd, bufsize=-1, close_fds=util.closefds,
                              stdout=subprocess.PIPE, stderr=subprocess.PIPE,
                               universal_newlines=True, env=env, **extrakw)
         stdout, stderr = p.communicate()
         stderr = stderr.strip()
         if not failok:
             if p.returncode:
                 raise error.Abort(stderr or 'exited with code %d'
                                   % p.returncode)
             if stderr:
                 self.ui.warn(stderr + '\n')
         return stdout, stderr
 
     @propertycache
     def _svnversion(self):
         output, err = self._svncommand(['--version', '--quiet'], filename=None)
         m = re.search(r'^(\d+)\.(\d+)', output)
         if not m:
             raise error.Abort(_('cannot retrieve svn tool version'))
         return (int(m.group(1)), int(m.group(2)))
 
     def _wcrevs(self):
         # Get the working directory revision as well as the last
         # commit revision so we can compare the subrepo state with
         # both. We used to store the working directory one.
         output, err = self._svncommand(['info', '--xml'])
         doc = xml.dom.minidom.parseString(output)
         entries = doc.getElementsByTagName('entry')
         lastrev, rev = '0', '0'
         if entries:
             rev = str(entries[0].getAttribute('revision')) or '0'
             commits = entries[0].getElementsByTagName('commit')
             if commits:
                 lastrev = str(commits[0].getAttribute('revision')) or '0'
         return (lastrev, rev)
 
     def _wcrev(self):
         return self._wcrevs()[0]
 
     def _wcchanged(self):
         """Return (changes, extchanges, missing) where changes is True
         if the working directory was changed, extchanges is
         True if any of these changes concern an external entry and missing
         is True if any change is a missing entry.
         """
         output, err = self._svncommand(['status', '--xml'])
         externals, changes, missing = [], [], []
         doc = xml.dom.minidom.parseString(output)
         for e in doc.getElementsByTagName('entry'):
             s = e.getElementsByTagName('wc-status')
             if not s:
                 continue
             item = s[0].getAttribute('item')
             props = s[0].getAttribute('props')
             path = e.getAttribute('path')
             if item == 'external':
                 externals.append(path)
             elif item == 'missing':
                 missing.append(path)
             if (item not in ('', 'normal', 'unversioned', 'external')
                 or props not in ('', 'none', 'normal')):
                 changes.append(path)
         for path in changes:
             for ext in externals:
                 if path == ext or path.startswith(ext + pycompat.ossep):
                     return True, True, bool(missing)
         return bool(changes), False, bool(missing)
 
     def dirty(self, ignoreupdate=False):
         if not self._wcchanged()[0]:
             if self._state[1] in self._wcrevs() or ignoreupdate:
                 return False
         return True
 
     def basestate(self):
         lastrev, rev = self._wcrevs()
         if lastrev != rev:
             # Last committed rev is not the same than rev. We would
             # like to take lastrev but we do not know if the subrepo
             # URL exists at lastrev.  Test it and fallback to rev it
             # is not there.
             try:
                 self._svncommand(['list', '%s@%s' % (self._state[0], lastrev)])
                 return lastrev
             except error.Abort:
                 pass
         return rev
 
     @annotatesubrepoerror
     def commit(self, text, user, date):
         # user and date are out of our hands since svn is centralized
         changed, extchanged, missing = self._wcchanged()
         if not changed:
             return self.basestate()
         if extchanged:
             # Do not try to commit externals
             raise error.Abort(_('cannot commit svn externals'))
         if missing:
             # svn can commit with missing entries but aborting like hg
             # seems a better approach.
             raise error.Abort(_('cannot commit missing svn entries'))
         commitinfo, err = self._svncommand(['commit', '-m', text])
         self.ui.status(commitinfo)
         newrev = re.search('Committed revision ([0-9]+).', commitinfo)
         if not newrev:
             if not commitinfo.strip():
                 # Sometimes, our definition of "changed" differs from
                 # svn one. For instance, svn ignores missing files
                 # when committing. If there are only missing files, no
                 # commit is made, no output and no error code.
                 raise error.Abort(_('failed to commit svn changes'))
             raise error.Abort(commitinfo.splitlines()[-1])
         newrev = newrev.groups()[0]
         self.ui.status(self._svncommand(['update', '-r', newrev])[0])
         return newrev
 
     @annotatesubrepoerror
     def remove(self):
         if self.dirty():
             self.ui.warn(_('not removing repo %s because '
                            'it has changes.\n') % self._path)
             return
         self.ui.note(_('removing subrepo %s\n') % self._path)
 
         self.wvfs.rmtree(forcibly=True)
         try:
             pwvfs = self._ctx.repo().wvfs
             pwvfs.removedirs(pwvfs.dirname(self._path))
         except OSError:
             pass
 
     @annotatesubrepoerror
     def get(self, state, overwrite=False):
         if overwrite:
             self._svncommand(['revert', '--recursive'])
         args = ['checkout']
         if self._svnversion >= (1, 5):
             args.append('--force')
         # The revision must be specified at the end of the URL to properly
         # update to a directory which has since been deleted and recreated.
         args.append('%s@%s' % (state[0], state[1]))
         status, err = self._svncommand(args, failok=True)
         _sanitize(self.ui, self.wvfs, '.svn')
         if not re.search('Checked out revision [0-9]+.', status):
             if ('is already a working copy for a different URL' in err
                 and (self._wcchanged()[:2] == (False, False))):
                 # obstructed but clean working copy, so just blow it away.
                 self.remove()
                 self.get(state, overwrite=False)
                 return
             raise error.Abort((status or err).splitlines()[-1])
         self.ui.status(status)
 
     @annotatesubrepoerror
     def merge(self, state):
         old = self._state[1]
         new = state[1]
         wcrev = self._wcrev()
         if new != wcrev:
             dirty = old == wcrev or self._wcchanged()[0]
             if _updateprompt(self.ui, self, dirty, wcrev, new):
                 self.get(state, False)
 
     def push(self, opts):
         # push is a no-op for SVN
         return True
 
     @annotatesubrepoerror
     def files(self):
         output = self._svncommand(['list', '--recursive', '--xml'])[0]
         doc = xml.dom.minidom.parseString(output)
         paths = []
         for e in doc.getElementsByTagName('entry'):
             kind = str(e.getAttribute('kind'))
             if kind != 'file':
                 continue
             name = ''.join(c.data for c
                            in e.getElementsByTagName('name')[0].childNodes
                            if c.nodeType == c.TEXT_NODE)
             paths.append(name.encode('utf-8'))
         return paths
 
-    def filedata(self, name):
+    def filedata(self, name, decode):
         return self._svncommand(['cat'], name)[0]
 
 
 class gitsubrepo(abstractsubrepo):
     def __init__(self, ctx, path, state, allowcreate):
         super(gitsubrepo, self).__init__(ctx, path)
         self._state = state
         self._abspath = ctx.repo().wjoin(path)
         self._subparent = ctx.repo()
         self._ensuregit()
 
     def _ensuregit(self):
         try:
             self._gitexecutable = 'git'
             out, err = self._gitnodir(['--version'])
         except OSError as e:
             genericerror = _("error executing git for subrepo '%s': %s")
             notfoundhint = _("check git is installed and in your PATH")
             if e.errno != errno.ENOENT:
                 raise error.Abort(genericerror % (self._path, e.strerror))
             elif pycompat.osname == 'nt':
                 try:
                     self._gitexecutable = 'git.cmd'
                     out, err = self._gitnodir(['--version'])
                 except OSError as e2:
                     if e2.errno == errno.ENOENT:
                         raise error.Abort(_("couldn't find 'git' or 'git.cmd'"
                             " for subrepo '%s'") % self._path,
                             hint=notfoundhint)
                     else:
                         raise error.Abort(genericerror % (self._path,
                             e2.strerror))
             else:
                 raise error.Abort(_("couldn't find git for subrepo '%s'")
                     % self._path, hint=notfoundhint)
         versionstatus = self._checkversion(out)
         if versionstatus == 'unknown':
             self.ui.warn(_('cannot retrieve git version\n'))
         elif versionstatus == 'abort':
             raise error.Abort(_('git subrepo requires at least 1.6.0 or later'))
         elif versionstatus == 'warning':
             self.ui.warn(_('git subrepo requires at least 1.6.0 or later\n'))
 
     @staticmethod
     def _gitversion(out):
         m = re.search(r'^git version (\d+)\.(\d+)\.(\d+)', out)
         if m:
             return (int(m.group(1)), int(m.group(2)), int(m.group(3)))
 
         m = re.search(r'^git version (\d+)\.(\d+)', out)
         if m:
             return (int(m.group(1)), int(m.group(2)), 0)
 
         return -1
 
     @staticmethod
     def _checkversion(out):
         '''ensure git version is new enough
 
         >>> _checkversion = gitsubrepo._checkversion
         >>> _checkversion('git version 1.6.0')
         'ok'
         >>> _checkversion('git version 1.8.5')
         'ok'
         >>> _checkversion('git version 1.4.0')
         'abort'
         >>> _checkversion('git version 1.5.0')
         'warning'
         >>> _checkversion('git version 1.9-rc0')
         'ok'
         >>> _checkversion('git version 1.9.0.265.g81cdec2')
         'ok'
         >>> _checkversion('git version 1.9.0.GIT')
         'ok'
         >>> _checkversion('git version 12345')
         'unknown'
         >>> _checkversion('no')
         'unknown'
         '''
         version = gitsubrepo._gitversion(out)
         # git 1.4.0 can't work at all, but 1.5.X can in at least some cases,
         # despite the docstring comment.  For now, error on 1.4.0, warn on
         # 1.5.0 but attempt to continue.
         if version == -1:
             return 'unknown'
         if version < (1, 5, 0):
             return 'abort'
         elif version < (1, 6, 0):
             return 'warning'
         return 'ok'
 
     def _gitcommand(self, commands, env=None, stream=False):
         return self._gitdir(commands, env=env, stream=stream)[0]
 
     def _gitdir(self, commands, env=None, stream=False):
         return self._gitnodir(commands, env=env, stream=stream,
                               cwd=self._abspath)
 
     def _gitnodir(self, commands, env=None, stream=False, cwd=None):
         """Calls the git command
 
         The methods tries to call the git command. versions prior to 1.6.0
         are not supported and very probably fail.
         """
         self.ui.debug('%s: git %s\n' % (self._relpath, ' '.join(commands)))
         if env is None:
             env = encoding.environ.copy()
         # disable localization for Git output (issue5176)
         env['LC_ALL'] = 'C'
         # fix for Git CVE-2015-7545
         if 'GIT_ALLOW_PROTOCOL' not in env:
             env['GIT_ALLOW_PROTOCOL'] = 'file:git:http:https:ssh'
         # unless ui.quiet is set, print git's stderr,
         # which is mostly progress and useful info
         errpipe = None
         if self.ui.quiet:
             errpipe = open(os.devnull, 'w')
         p = subprocess.Popen([self._gitexecutable] + commands, bufsize=-1,
                              cwd=cwd, env=env, close_fds=util.closefds,
                              stdout=subprocess.PIPE, stderr=errpipe)
         if stream:
             return p.stdout, None
 
         retdata = p.stdout.read().strip()
         # wait for the child to exit to avoid race condition.
         p.wait()
 
         if p.returncode != 0 and p.returncode != 1:
             # there are certain error codes that are ok
             command = commands[0]
             if command in ('cat-file', 'symbolic-ref'):
                 return retdata, p.returncode
             # for all others, abort
             raise error.Abort(_('git %s error %d in %s') %
                              (command, p.returncode, self._relpath))
 
         return retdata, p.returncode
 
     def _gitmissing(self):
         return not self.wvfs.exists('.git')
 
     def _gitstate(self):
         return self._gitcommand(['rev-parse', 'HEAD'])
 
     def _gitcurrentbranch(self):
         current, err = self._gitdir(['symbolic-ref', 'HEAD', '--quiet'])
         if err:
             current = None
         return current
 
     def _gitremote(self, remote):
         out = self._gitcommand(['remote', 'show', '-n', remote])
         line = out.split('\n')[1]
         i = line.index('URL: ') + len('URL: ')
         return line[i:]
 
     def _githavelocally(self, revision):
         out, code = self._gitdir(['cat-file', '-e', revision])
         return code == 0
 
     def _gitisancestor(self, r1, r2):
         base = self._gitcommand(['merge-base', r1, r2])
         return base == r1
 
     def _gitisbare(self):
         return self._gitcommand(['config', '--bool', 'core.bare']) == 'true'
 
     def _gitupdatestat(self):
         """This must be run before git diff-index.
         diff-index only looks at changes to file stat;
         this command looks at file contents and updates the stat."""
         self._gitcommand(['update-index', '-q', '--refresh'])
 
     def _gitbranchmap(self):
         '''returns 2 things:
         a map from git branch to revision
         a map from revision to branches'''
         branch2rev = {}
         rev2branch = {}
 
         out = self._gitcommand(['for-each-ref', '--format',
                                 '%(objectname) %(refname)'])
         for line in out.split('\n'):
             revision, ref = line.split(' ')
             if (not ref.startswith('refs/heads/') and
                 not ref.startswith('refs/remotes/')):
                 continue
             if ref.startswith('refs/remotes/') and ref.endswith('/HEAD'):
                 continue # ignore remote/HEAD redirects
             branch2rev[ref] = revision
             rev2branch.setdefault(revision, []).append(ref)
         return branch2rev, rev2branch
 
     def _gittracking(self, branches):
         'return map of remote branch to local tracking branch'
         # assumes no more than one local tracking branch for each remote
         tracking = {}
         for b in branches:
             if b.startswith('refs/remotes/'):
                 continue
             bname = b.split('/', 2)[2]
             remote = self._gitcommand(['config', 'branch.%s.remote' % bname])
             if remote:
                 ref = self._gitcommand(['config', 'branch.%s.merge' % bname])
                 tracking['refs/remotes/%s/%s' %
                          (remote, ref.split('/', 2)[2])] = b
         return tracking
 
     def _abssource(self, source):
         if '://' not in source:
             # recognize the scp syntax as an absolute source
             colon = source.find(':')
             if colon != -1 and '/' not in source[:colon]:
                 return source
         self._subsource = source
         return _abssource(self)
 
     def _fetch(self, source, revision):
         if self._gitmissing():
             source = self._abssource(source)
             self.ui.status(_('cloning subrepo %s from %s\n') %
                             (self._relpath, source))
             self._gitnodir(['clone', source, self._abspath])
         if self._githavelocally(revision):
             return
         self.ui.status(_('pulling subrepo %s from %s\n') %
                         (self._relpath, self._gitremote('origin')))
         # try only origin: the originally cloned repo
         self._gitcommand(['fetch'])
         if not self._githavelocally(revision):
             raise error.Abort(_("revision %s does not exist in subrepo %s\n") %
                                (revision, self._relpath))
 
     @annotatesubrepoerror
     def dirty(self, ignoreupdate=False):
         if self._gitmissing():
             return self._state[1] != ''
         if self._gitisbare():
             return True
         if not ignoreupdate and self._state[1] != self._gitstate():
             # different version checked out
             return True
         # check for staged changes or modified files; ignore untracked files
         self._gitupdatestat()
         out, code = self._gitdir(['diff-index', '--quiet', 'HEAD'])
         return code == 1
 
     def basestate(self):
         return self._gitstate()
 
     @annotatesubrepoerror
     def get(self, state, overwrite=False):
         source, revision, kind = state
         if not revision:
             self.remove()
             return
         self._fetch(source, revision)
         # if the repo was set to be bare, unbare it
         if self._gitisbare():
             self._gitcommand(['config', 'core.bare', 'false'])
             if self._gitstate() == revision:
                 self._gitcommand(['reset', '--hard', 'HEAD'])
                 return
         elif self._gitstate() == revision:
             if overwrite:
                 # first reset the index to unmark new files for commit, because
                 # reset --hard will otherwise throw away files added for commit,
                 # not just unmark them.
                 self._gitcommand(['reset', 'HEAD'])
                 self._gitcommand(['reset', '--hard', 'HEAD'])
             return
         branch2rev, rev2branch = self._gitbranchmap()
 
         def checkout(args):
             cmd = ['checkout']
             if overwrite:
                 # first reset the index to unmark new files for commit, because
                 # the -f option will otherwise throw away files added for
                 # commit, not just unmark them.
                 self._gitcommand(['reset', 'HEAD'])
                 cmd.append('-f')
             self._gitcommand(cmd + args)
             _sanitize(self.ui, self.wvfs, '.git')
 
         def rawcheckout():
             # no branch to checkout, check it out with no branch
             self.ui.warn(_('checking out detached HEAD in subrepo %s\n') %
                           self._relpath)
             self.ui.warn(_('check out a git branch if you intend '
                             'to make changes\n'))
             checkout(['-q', revision])
 
         if revision not in rev2branch:
             rawcheckout()
             return
         branches = rev2branch[revision]
         firstlocalbranch = None
         for b in branches:
             if b == 'refs/heads/master':
                 # master trumps all other branches
                 checkout(['refs/heads/master'])
                 return
             if not firstlocalbranch and not b.startswith('refs/remotes/'):
                 firstlocalbranch = b
         if firstlocalbranch:
             checkout([firstlocalbranch])
             return
 
         tracking = self._gittracking(branch2rev.keys())
         # choose a remote branch already tracked if possible
         remote = branches[0]
         if remote not in tracking:
             for b in branches:
                 if b in tracking:
                     remote = b
                     break
 
         if remote not in tracking:
             # create a new local tracking branch
             local = remote.split('/', 3)[3]
             checkout(['-b', local, remote])
         elif self._gitisancestor(branch2rev[tracking[remote]], remote):
             # When updating to a tracked remote branch,
             # if the local tracking branch is downstream of it,
             # a normal `git pull` would have performed a "fast-forward merge"
             # which is equivalent to updating the local branch to the remote.
             # Since we are only looking at branching at update, we need to
             # detect this situation and perform this action lazily.
             if tracking[remote] != self._gitcurrentbranch():
                 checkout([tracking[remote]])
             self._gitcommand(['merge', '--ff', remote])
             _sanitize(self.ui, self.wvfs, '.git')
         else:
             # a real merge would be required, just checkout the revision
             rawcheckout()
 
     @annotatesubrepoerror
     def commit(self, text, user, date):
         if self._gitmissing():
             raise error.Abort(_("subrepo %s is missing") % self._relpath)
         cmd = ['commit', '-a', '-m', text]
         env = encoding.environ.copy()
         if user:
             cmd += ['--author', user]
         if date:
             # git's date parser silently ignores when seconds < 1e9
             # convert to ISO8601
             env['GIT_AUTHOR_DATE'] = util.datestr(date,
                                                   '%Y-%m-%dT%H:%M:%S %1%2')
         self._gitcommand(cmd, env=env)
         # make sure commit works otherwise HEAD might not exist under certain
         # circumstances
         return self._gitstate()
 
     @annotatesubrepoerror
     def merge(self, state):
         source, revision, kind = state
         self._fetch(source, revision)
         base = self._gitcommand(['merge-base', revision, self._state[1]])
         self._gitupdatestat()
         out, code = self._gitdir(['diff-index', '--quiet', 'HEAD'])
 
         def mergefunc():
             if base == revision:
                 self.get(state) # fast forward merge
             elif base != self._state[1]:
                 self._gitcommand(['merge', '--no-commit', revision])
             _sanitize(self.ui, self.wvfs, '.git')
 
         if self.dirty():
             if self._gitstate() != revision:
                 dirty = self._gitstate() == self._state[1] or code != 0
                 if _updateprompt(self.ui, self, dirty,
                                  self._state[1][:7], revision[:7]):
                     mergefunc()
         else:
             mergefunc()
 
     @annotatesubrepoerror
     def push(self, opts):
         force = opts.get('force')
 
         if not self._state[1]:
             return True
         if self._gitmissing():
             raise error.Abort(_("subrepo %s is missing") % self._relpath)
         # if a branch in origin contains the revision, nothing to do
         branch2rev, rev2branch = self._gitbranchmap()
         if self._state[1] in rev2branch:
             for b in rev2branch[self._state[1]]:
                 if b.startswith('refs/remotes/origin/'):
                     return True
         for b, revision in branch2rev.iteritems():
             if b.startswith('refs/remotes/origin/'):
                 if self._gitisancestor(self._state[1], revision):
                     return True
         # otherwise, try to push the currently checked out branch
         cmd = ['push']
         if force:
             cmd.append('--force')
 
         current = self._gitcurrentbranch()
         if current:
             # determine if the current branch is even useful
             if not self._gitisancestor(self._state[1], current):
                 self.ui.warn(_('unrelated git branch checked out '
                                 'in subrepo %s\n') % self._relpath)
                 return False
             self.ui.status(_('pushing branch %s of subrepo %s\n') %
                            (current.split('/', 2)[2], self._relpath))
             ret = self._gitdir(cmd + ['origin', current])
             return ret[1] == 0
         else:
             self.ui.warn(_('no branch checked out in subrepo %s\n'
                            'cannot push revision %s\n') %
                           (self._relpath, self._state[1]))
             return False
 
     @annotatesubrepoerror
     def add(self, ui, match, prefix, explicitonly, **opts):
         if self._gitmissing():
             return []
 
         (modified, added, removed,
          deleted, unknown, ignored, clean) = self.status(None, unknown=True,
                                                          clean=True)
 
         tracked = set()
         # dirstates 'amn' warn, 'r' is added again
         for l in (modified, added, deleted, clean):
             tracked.update(l)
 
         # Unknown files not of interest will be rejected by the matcher
         files = unknown
         files.extend(match.files())
 
         rejected = []
 
         files = [f for f in sorted(set(files)) if match(f)]
         for f in files:
             exact = match.exact(f)
             command = ["add"]
             if exact:
                 command.append("-f") #should be added, even if ignored
             if ui.verbose or not exact:
                 ui.status(_('adding %s\n') % match.rel(f))
 
             if f in tracked:  # hg prints 'adding' even if already tracked
                 if exact:
                     rejected.append(f)
                 continue
             if not opts.get('dry_run'):
                 self._gitcommand(command + [f])
 
         for f in rejected:
             ui.warn(_("%s already tracked!\n") % match.abs(f))
 
         return rejected
 
     @annotatesubrepoerror
     def remove(self):
         if self._gitmissing():
             return
         if self.dirty():
             self.ui.warn(_('not removing repo %s because '
                            'it has changes.\n') % self._relpath)
             return
         # we can't fully delete the repository as it may contain
         # local-only history
         self.ui.note(_('removing subrepo %s\n') % self._relpath)
         self._gitcommand(['config', 'core.bare', 'true'])
         for f, kind in self.wvfs.readdir():
             if f == '.git':
                 continue
             if kind == stat.S_IFDIR:
                 self.wvfs.rmtree(f)
             else:
                 self.wvfs.unlink(f)
 
-    def archive(self, archiver, prefix, match=None):
+    def archive(self, archiver, prefix, match=None, decode=True):
         total = 0
         source, revision = self._state
         if not revision:
             return total
         self._fetch(source, revision)
 
         # Parse git's native archive command.
         # This should be much faster than manually traversing the trees
         # and objects with many subprocess calls.
         tarstream = self._gitcommand(['archive', revision], stream=True)
         tar = tarfile.open(fileobj=tarstream, mode='r|')
         relpath = subrelpath(self)
         self.ui.progress(_('archiving (%s)') % relpath, 0, unit=_('files'))
         for i, info in enumerate(tar):
             if info.isdir():
                 continue
             if match and not match(info.name):
                 continue
             if info.issym():
                 data = info.linkname
             else:
                 data = tar.extractfile(info).read()
             archiver.addfile(prefix + self._path + '/' + info.name,
                              info.mode, info.issym(), data)
             total += 1
             self.ui.progress(_('archiving (%s)') % relpath, i + 1,
                              unit=_('files'))
         self.ui.progress(_('archiving (%s)') % relpath, None)
         return total
 
 
     @annotatesubrepoerror
     def cat(self, match, prefix, **opts):
         rev = self._state[1]
         if match.anypats():
             return 1 #No support for include/exclude yet
 
         if not match.files():
             return 1
 
         for f in match.files():
             output = self._gitcommand(["show", "%s:%s" % (rev, f)])
             fp = cmdutil.makefileobj(self._subparent, opts.get('output'),
                                      self._ctx.node(),
                                      pathname=self.wvfs.reljoin(prefix, f))
             fp.write(output)
             fp.close()
         return 0
 
 
     @annotatesubrepoerror
     def status(self, rev2, **opts):
         rev1 = self._state[1]
         if self._gitmissing() or not rev1:
             # if the repo is missing, return no results
             return scmutil.status([], [], [], [], [], [], [])
         modified, added, removed = [], [], []
         self._gitupdatestat()
         if rev2:
             command = ['diff-tree', '--no-renames', '-r', rev1, rev2]
         else:
             command = ['diff-index', '--no-renames', rev1]
         out = self._gitcommand(command)
         for line in out.split('\n'):
             tab = line.find('\t')
             if tab == -1:
                 continue
             status, f = line[tab - 1], line[tab + 1:]
             if status == 'M':
                 modified.append(f)
             elif status == 'A':
                 added.append(f)
             elif status == 'D':
                 removed.append(f)
 
         deleted, unknown, ignored, clean = [], [], [], []
 
         command = ['status', '--porcelain', '-z']
         if opts.get('unknown'):
             command += ['--untracked-files=all']
         if opts.get('ignored'):
             command += ['--ignored']
         out = self._gitcommand(command)
 
         changedfiles = set()
         changedfiles.update(modified)
         changedfiles.update(added)
         changedfiles.update(removed)
         for line in out.split('\0'):
             if not line:
                 continue
             st = line[0:2]
             #moves and copies show 2 files on one line
             if line.find('\0') >= 0:
                 filename1, filename2 = line[3:].split('\0')
             else:
                 filename1 = line[3:]
                 filename2 = None
 
             changedfiles.add(filename1)
             if filename2:
                 changedfiles.add(filename2)
 
             if st == '??':
                 unknown.append(filename1)
             elif st == '!!':
                 ignored.append(filename1)
 
         if opts.get('clean'):
             out = self._gitcommand(['ls-files'])
             for f in out.split('\n'):
                 if not f in changedfiles:
                     clean.append(f)
 
         return scmutil.status(modified, added, removed, deleted,
                               unknown, ignored, clean)
 
     @annotatesubrepoerror
     def diff(self, ui, diffopts, node2, match, prefix, **opts):
         node1 = self._state[1]
         cmd = ['diff', '--no-renames']
         if opts['stat']:
             cmd.append('--stat')
         else:
             # for Git, this also implies '-p'
             cmd.append('-U%d' % diffopts.context)
 
         gitprefix = self.wvfs.reljoin(prefix, self._path)
 
         if diffopts.noprefix:
             cmd.extend(['--src-prefix=%s/' % gitprefix,
                         '--dst-prefix=%s/' % gitprefix])
         else:
             cmd.extend(['--src-prefix=a/%s/' % gitprefix,
                         '--dst-prefix=b/%s/' % gitprefix])
 
         if diffopts.ignorews:
             cmd.append('--ignore-all-space')
         if diffopts.ignorewsamount:
             cmd.append('--ignore-space-change')
         if self._gitversion(self._gitcommand(['--version'])) >= (1, 8, 4) \
                 and diffopts.ignoreblanklines:
             cmd.append('--ignore-blank-lines')
 
         cmd.append(node1)
         if node2:
             cmd.append(node2)
 
         output = ""
         if match.always():
             output += self._gitcommand(cmd) + '\n'
         else:
             st = self.status(node2)[:3]
             files = [f for sublist in st for f in sublist]
             for f in files:
                 if match(f):
                     output += self._gitcommand(cmd + ['--', f]) + '\n'
 
         if output.strip():
             ui.write(output)
 
     @annotatesubrepoerror
     def revert(self, substate, *pats, **opts):
         self.ui.status(_('reverting subrepo %s\n') % substate[0])
         if not opts.get('no_backup'):
             status = self.status(None)
             names = status.modified
             for name in names:
                 bakname = scmutil.origpath(self.ui, self._subparent, name)
                 self.ui.note(_('saving current version of %s as %s\n') %
                         (name, bakname))
                 self.wvfs.rename(name, bakname)
 
         if not opts.get('dry_run'):
             self.get(substate, overwrite=True)
         return []
 
     def shortid(self, revid):
         return revid[:7]
 
 types = {
     'hg': hgsubrepo,
     'svn': svnsubrepo,
     'git': gitsubrepo,
     }

tests/test-eol.t

@@ -1,545 +1,561 b''
 Test EOL extension
 
   $ cat >> $HGRCPATH <<EOF
   > [diff]
   > git = True
   > EOF
 
 Set up helpers
 
   $ cat > switch-eol.py <<EOF
   > import sys
   > try:
   >     import os, msvcrt
   >     msvcrt.setmode(sys.stdin.fileno(), os.O_BINARY)
   >     msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
   > except ImportError:
   >     pass
   > (old, new) = sys.argv[1] == 'LF' and ('\n', '\r\n') or ('\r\n', '\n')
   > print "%% switching encoding from %r to %r" % (old, new)
   > for path in sys.argv[2:]:
   >     data = file(path, 'rb').read()
   >     data = data.replace(old, new)
   >     file(path, 'wb').write(data)
   > EOF
 
   $ seteol () {
   >     if [ $1 = "LF" ]; then
   >         EOL='\n'
   >     else
   >         EOL='\r\n'
   >     fi
   > }
 
   $ makerepo () {
   >     seteol $1
   >     echo "% setup $1 repository"
   >     hg init repo
   >     cd repo
   >     cat > .hgeol <<EOF
   > [repository]
   > native = $1
   > [patterns]
   > mixed.txt = BIN
   > **.txt = native
   > EOF
   >     printf "first${EOL}second${EOL}third${EOL}" > a.txt
   >     hg commit --addremove -m 'checkin'
   >     echo
   >     cd ..
   > }
 
   $ dotest () {
   >     seteol $1
   >     echo "% hg clone repo repo-$1"
   >     hg clone --noupdate repo repo-$1
   >     cd repo-$1
   >     cat > .hg/hgrc <<EOF
   > [extensions]
   > eol =
   > [eol]
   > native = $1
   > EOF
   >     hg update
   >     echo '% a.txt'
   >     cat a.txt
   >     echo '% hg cat a.txt'
   >     hg cat a.txt
   >     printf "fourth${EOL}" >> a.txt
   >     echo '% a.txt'
   >     cat a.txt
   >     hg diff
   >     python ../switch-eol.py $1 a.txt
   >     echo '% hg diff only reports a single changed line:'
   >     hg diff
   >     echo "% reverting back to $1 format"
   >     hg revert a.txt
   >     cat a.txt
   >     printf "first\r\nsecond\n" > mixed.txt
   >     hg add mixed.txt
   >     echo "% hg commit of inconsistent .txt file marked as binary (should work)"
   >     hg commit -m 'binary file'
   >     echo "% hg commit of inconsistent .txt file marked as native (should fail)"
   >     printf "first\nsecond\r\nthird\nfourth\r\n" > a.txt
   >     hg commit -m 'inconsistent file'
   >     echo "% hg commit --config eol.only-consistent=False (should work)"
   >     hg commit --config eol.only-consistent=False -m 'inconsistent file'
   >     echo "% hg commit of binary .txt file marked as native (binary files always okay)"
   >     printf "first${EOL}\0${EOL}third${EOL}" > a.txt
   >     hg commit -m 'binary file'
   >     cd ..
   >     rm -r repo-$1
   > }
 
   $ makemixedrepo () {
   >     echo
   >     echo "# setup $1 repository"
   >     hg init mixed
   >     cd mixed
   >     printf "foo\r\nbar\r\nbaz\r\n" > win.txt
   >     printf "foo\nbar\nbaz\n" > unix.txt
   >     #printf "foo\r\nbar\nbaz\r\n" > mixed.txt
   >     hg commit --addremove -m 'created mixed files'
   >     echo "# setting repository-native EOLs to $1"
   >     cat > .hgeol <<EOF
   > [repository]
   > native = $1
   > [patterns]
   > **.txt = native
   > EOF
   >     hg commit --addremove -m 'added .hgeol'
   >     cd ..
   > }
 
   $ testmixed () {
   >     echo
   >     echo "% hg clone mixed mixed-$1"
   >     hg clone mixed mixed-$1
   >     cd mixed-$1
   >     echo '% hg status (eol extension not yet activated)'
   >     hg status
   >     cat > .hg/hgrc <<EOF
   > [extensions]
   > eol =
   > [eol]
   > native = $1
   > EOF
   >     echo '% hg status (eol activated)'
   >     hg status
   >     echo '% hg commit'
   >     hg commit -m 'synchronized EOLs'
   >     echo '% hg status'
   >     hg status
   >     cd ..
   >     rm -r mixed-$1
   > }
 
 Basic tests
 
   $ makerepo LF
   % setup LF repository
   adding .hgeol
   adding a.txt
   
   $ dotest LF
   % hg clone repo repo-LF
   2 files updated, 0 files merged, 0 files removed, 0 files unresolved
   % a.txt
   first
   second
   third
   % hg cat a.txt
   first
   second
   third
   % a.txt
   first
   second
   third
   fourth
   diff --git a/a.txt b/a.txt
   --- a/a.txt
   +++ b/a.txt
   @@ -1,3 +1,4 @@
    first
    second
    third
   +fourth
   % switching encoding from '\n' to '\r\n'
   % hg diff only reports a single changed line:
   diff --git a/a.txt b/a.txt
   --- a/a.txt
   +++ b/a.txt
   @@ -1,3 +1,4 @@
    first
    second
    third
   +fourth
   % reverting back to LF format
   first
   second
   third
   % hg commit of inconsistent .txt file marked as binary (should work)
   % hg commit of inconsistent .txt file marked as native (should fail)
   abort: inconsistent newline style in a.txt
   
   % hg commit --config eol.only-consistent=False (should work)
   % hg commit of binary .txt file marked as native (binary files always okay)
   $ dotest CRLF
   % hg clone repo repo-CRLF
   2 files updated, 0 files merged, 0 files removed, 0 files unresolved
   % a.txt
   first\r (esc)
   second\r (esc)
   third\r (esc)
   % hg cat a.txt
   first
   second
   third
   % a.txt
   first\r (esc)
   second\r (esc)
   third\r (esc)
   fourth\r (esc)
   diff --git a/a.txt b/a.txt
   --- a/a.txt
   +++ b/a.txt
   @@ -1,3 +1,4 @@
    first
    second
    third
   +fourth
   % switching encoding from '\r\n' to '\n'
   % hg diff only reports a single changed line:
   diff --git a/a.txt b/a.txt
   --- a/a.txt
   +++ b/a.txt
   @@ -1,3 +1,4 @@
    first
    second
    third
   +fourth
   % reverting back to CRLF format
   first\r (esc)
   second\r (esc)
   third\r (esc)
   % hg commit of inconsistent .txt file marked as binary (should work)
   % hg commit of inconsistent .txt file marked as native (should fail)
   abort: inconsistent newline style in a.txt
   
   % hg commit --config eol.only-consistent=False (should work)
   % hg commit of binary .txt file marked as native (binary files always okay)
   $ rm -r repo
   $ makerepo CRLF
   % setup CRLF repository
   adding .hgeol
   adding a.txt
   
   $ dotest LF
   % hg clone repo repo-LF
   2 files updated, 0 files merged, 0 files removed, 0 files unresolved
   % a.txt
   first
   second
   third
   % hg cat a.txt
   first\r (esc)
   second\r (esc)
   third\r (esc)
   % a.txt
   first
   second
   third
   fourth
   diff --git a/a.txt b/a.txt
   --- a/a.txt
   +++ b/a.txt
   @@ -1,3 +1,4 @@
    first\r (esc)
    second\r (esc)
    third\r (esc)
   +fourth\r (esc)
   % switching encoding from '\n' to '\r\n'
   % hg diff only reports a single changed line:
   diff --git a/a.txt b/a.txt
   --- a/a.txt
   +++ b/a.txt
   @@ -1,3 +1,4 @@
    first\r (esc)
    second\r (esc)
    third\r (esc)
   +fourth\r (esc)
   % reverting back to LF format
   first
   second
   third
   % hg commit of inconsistent .txt file marked as binary (should work)
   % hg commit of inconsistent .txt file marked as native (should fail)
   abort: inconsistent newline style in a.txt
   
   % hg commit --config eol.only-consistent=False (should work)
   % hg commit of binary .txt file marked as native (binary files always okay)
   $ dotest CRLF
   % hg clone repo repo-CRLF
   2 files updated, 0 files merged, 0 files removed, 0 files unresolved
   % a.txt
   first\r (esc)
   second\r (esc)
   third\r (esc)
   % hg cat a.txt
   first\r (esc)
   second\r (esc)
   third\r (esc)
   % a.txt
   first\r (esc)
   second\r (esc)
   third\r (esc)
   fourth\r (esc)
   diff --git a/a.txt b/a.txt
   --- a/a.txt
   +++ b/a.txt
   @@ -1,3 +1,4 @@
    first\r (esc)
    second\r (esc)
    third\r (esc)
   +fourth\r (esc)
   % switching encoding from '\r\n' to '\n'
   % hg diff only reports a single changed line:
   diff --git a/a.txt b/a.txt
   --- a/a.txt
   +++ b/a.txt
   @@ -1,3 +1,4 @@
    first\r (esc)
    second\r (esc)
    third\r (esc)
   +fourth\r (esc)
   % reverting back to CRLF format
   first\r (esc)
   second\r (esc)
   third\r (esc)
   % hg commit of inconsistent .txt file marked as binary (should work)
   % hg commit of inconsistent .txt file marked as native (should fail)
   abort: inconsistent newline style in a.txt
   
   % hg commit --config eol.only-consistent=False (should work)
   % hg commit of binary .txt file marked as native (binary files always okay)
   $ rm -r repo
 
 Mixed tests
 
   $ makemixedrepo LF
   
   # setup LF repository
   adding unix.txt
   adding win.txt
   # setting repository-native EOLs to LF
   adding .hgeol
   $ testmixed LF
   
   % hg clone mixed mixed-LF
   updating to branch default
   3 files updated, 0 files merged, 0 files removed, 0 files unresolved
   % hg status (eol extension not yet activated)
   % hg status (eol activated)
   M win.txt
   % hg commit
   % hg status
   $ testmixed CRLF
   
   % hg clone mixed mixed-CRLF
   updating to branch default
   3 files updated, 0 files merged, 0 files removed, 0 files unresolved
   % hg status (eol extension not yet activated)
   % hg status (eol activated)
   M win.txt
   % hg commit
   % hg status
   $ rm -r mixed
   $ makemixedrepo CRLF
   
   # setup CRLF repository
   adding unix.txt
   adding win.txt
   # setting repository-native EOLs to CRLF
   adding .hgeol
   $ testmixed LF
   
   % hg clone mixed mixed-LF
   updating to branch default
   3 files updated, 0 files merged, 0 files removed, 0 files unresolved
   % hg status (eol extension not yet activated)
   % hg status (eol activated)
   M unix.txt
   % hg commit
   % hg status
   $ testmixed CRLF
   
   % hg clone mixed mixed-CRLF
   updating to branch default
   3 files updated, 0 files merged, 0 files removed, 0 files unresolved
   % hg status (eol extension not yet activated)
   % hg status (eol activated)
   M unix.txt
   % hg commit
   % hg status
   $ rm -r mixed
 
   $ echo '[extensions]' >> $HGRCPATH
   $ echo 'eol =' >> $HGRCPATH
 
 #if unix-permissions
 
 Test issue2569 -- eol extension takes write lock on reading:
 
   $ hg init repo
   $ cd repo
   $ touch .hgeol
   $ hg status
   ? .hgeol
   $ chmod -R -w .hg
   $ sleep 1
   $ touch .hgeol
   $ hg status --traceback
   ? .hgeol
   $ chmod -R u+w .hg
   $ cd ..
 
 #endif
 
 Test cleverencode: and cleverdecode: aliases for win32text extension
 
   $ cat <<EOF >> $HGRCPATH
   > [encode]
   > **.txt = cleverencode:
   > [decode]
   > **.txt = cleverdecode:
   > EOF
 
   $ hg init win32compat
   $ cd win32compat
   $ printf "foo\r\nbar\r\nbaz\r\n" > win.txt
   $ printf "foo\nbar\nbaz\n" > unix.txt
   $ hg add
   adding unix.txt
   adding win.txt
   $ hg commit -m checkin
 
 Check that both files have LF line-endings in the repository:
 
   $ hg cat win.txt
   foo
   bar
   baz
   $ hg cat unix.txt
   foo
   bar
   baz
 
 Test handling of a broken .hgeol file:
 
   $ touch .hgeol
   $ hg add .hgeol
   $ hg commit -m 'clean version'
   $ echo "bad" > .hgeol
   $ hg status
   warning: ignoring .hgeol file due to parse error at .hgeol:1: bad
   M .hgeol
   $ hg revert .hgeol
   warning: ignoring .hgeol file due to parse error at .hgeol:1: bad
   $ hg status
   ? .hgeol.orig
 
 Test eol.only-consistent can be specified in .hgeol
 
   $ cd $TESTTMP
   $ hg init only-consistent
   $ cd only-consistent
   $ printf "first\nsecond\r\n" > a.txt
   $ hg add a.txt
   $ cat > .hgeol << EOF
   > [eol]
   > only-consistent = True
   > EOF
   $ hg commit -m 'inconsistent'
   abort: inconsistent newline style in a.txt
   
   [255]
   $ cat > .hgeol << EOF
   > [eol]
   > only-consistent = False
   > EOF
   $ hg commit -m 'consistent'
 
+  $ hg init subrepo
+  $ hg -R subrepo pull -qu .
+  $ echo "subrepo = subrepo" > .hgsub
+  $ hg ci -Am "add subrepo"
+  adding .hgeol
+  adding .hgsub
+  $ hg archive -S ../archive
+  $ find ../archive/* | sort
+  ../archive/a.txt
+  ../archive/subrepo
+  ../archive/subrepo/a.txt
+  $ cat ../archive/a.txt ../archive/subrepo/a.txt
+  first\r (esc)
+  second\r (esc)
+  first\r (esc)
+  second\r (esc)
 
 Test trailing newline
 
   $ cat >> $HGRCPATH <<EOF
   > [extensions]
   > eol=
   > EOF
 
 setup repository
 
   $ cd $TESTTMP
   $ hg init trailing
   $ cd trailing
   $ cat > .hgeol <<EOF
   > [patterns]
   > **.txt = native
   > [eol]
   > fix-trailing-newline = False
   > EOF
 
 add text without trailing newline
 
   $ printf "first\nsecond" > a.txt
   $ hg commit --addremove -m 'checking in'
   adding .hgeol
   adding a.txt
   $ rm a.txt
   $ hg update -C -q
   $ cat a.txt
   first
   second (no-eol)
 
   $ cat > .hgeol <<EOF
   > [patterns]
   > **.txt = native
   > [eol]
   > fix-trailing-newline = True
   > EOF
   $ printf "third\nfourth" > a.txt
   $ hg commit -m 'checking in with newline fix'
   $ rm a.txt
   $ hg update -C -q
   $ cat a.txt
   third
   fourth
 
 append a line without trailing newline
 
   $ printf "fifth" >> a.txt
   $ hg commit -m 'adding another line line'
   $ rm a.txt
   $ hg update -C -q
   $ cat a.txt
   third
   fourth
   fifth
 
 amend of changesets with renamed/deleted files expose new code paths
 
   $ hg mv a.txt b.txt
   $ hg ci --amend -q
   $ hg diff -c.
   diff --git a/a.txt b/b.txt
   rename from a.txt
   rename to b.txt
   --- a/a.txt
   +++ b/b.txt
   @@ -1,2 +1,3 @@
    third
    fourth
   +fifth
 
   $ cd ..