archive: call the storage prefetch hook

mercurial/archival.py

@@ -1,358 +1,360 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,
     error,
     formatter,
     match as matchmod,
     util,
     vfs as vfsmod,
 )
 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']
 
 # {tags} on ctx includes local tags and 'tip', with no current way to limit
 # that to global tags.  Therefore, use {latesttag} as a substitute when
 # the distance is 0, since that will be the list of global tags on ctx.
 _defaultmetatemplate = br'''
 repo: {root}
 node: {ifcontains(rev, revset("wdir()"), "{p1node}{dirty}", "{node}")}
 branch: {branch|utf8}
 {ifeq(latesttagdistance, 0, join(latesttag % "tag: {tag}", "\n"),
       separate("\n",
                join(latesttag % "latesttag: {tag}", "\n"),
                "latesttagdistance: {latesttagdistance}",
                "changessincelatesttag: {changessincelatesttag}"))}
 '''[1:]  # drop leading '\n'
 
 def buildmetadata(ctx):
     '''build content of .hg_archival.txt'''
     repo = ctx.repo()
 
     opts = {
         'template': repo.ui.config('experimental', 'archivemetatemplate',
                                    _defaultmetatemplate)
     }
 
     out = util.stringio()
 
     fm = formatter.formatter(repo.ui, out, 'archive', opts)
     fm.startitem()
     fm.context(ctx=ctx)
     fm.data(root=_rootctx(repo).hex())
 
     if ctx.rev() is None:
         dirty = ''
         if ctx.dirty(missing=True):
             dirty = '+'
         fm.data(dirty=dirty)
     fm.end()
 
     return out.getvalue()
 
 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(r'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 = vfsmod.vfs(self.basedir)
         self.mtime = mtime
 
     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)
         if self.mtime is not None:
             os.utime(destfile, (self.mtime, self.mtime))
 
     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.
 
     mtime is the modified time, in seconds, or None to use the changeset time.
 
     subrepos tells whether to include subrepos.
     '''
 
     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"):
         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()
+        cmdutil._prefetchfiles(repo, ctx, files)
         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, decode)
 
     if total == 0:
         raise error.Abort(_('no files match the archive pattern'))
 
     archiver.done()
     return total

mercurial/subrepo.py

@@ -1,2151 +1,2155 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,
     logcmdutil,
     match as matchmod,
     node,
     pathutil,
     phases,
     pycompat,
     scmutil,
     util,
     vfs as vfsmod,
 )
 
 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 node.hex(hashlib.sha1(_expandedabspath(remotepath)).digest())[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(r'subrepo', None)
         self.cause = kw.pop(r'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 subrepository "%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 = util.escapestr(repl)
             # 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(br'\\\\([0-9]+)', br'\\\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 subrepository 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))
 
     promptssrc = filemerge.partextras(labels)
     for s, l in sorted(s1.iteritems()):
         prompts = None
         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
 
         prompts = promptssrc.copy()
         prompts['s'] = s
         if s in s2:
             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()):
         prompts = None
         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]:
             prompts = promptssrc.copy()
             prompts['s'] = 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 precommit(ui, wctx, status, match, force=False):
     """Calculate .hgsubstate changes that should be applied before committing
 
     Returns (subs, commitsubs, newstate) where
     - subs: changed subrepos (including dirty ones)
     - commitsubs: dirty subrepos which the caller needs to commit recursively
     - newstate: new state dict which the caller must write to .hgsubstate
 
     This also updates the given status argument.
     """
     subs = []
     commitsubs = set()
     newstate = wctx.substate.copy()
 
     # only manage subrepos and .hgsubstate if .hgsub is present
     if '.hgsub' in wctx:
         # we'll decide whether to track this ourselves, thanks
         for c in status.modified, status.added, status.removed:
             if '.hgsubstate' in c:
                 c.remove('.hgsubstate')
 
         # compare current state to last committed state
         # build new substate based on last committed state
         oldstate = wctx.p1().substate
         for s in sorted(newstate.keys()):
             if not match(s):
                 # ignore working copy, use old state if present
                 if s in oldstate:
                     newstate[s] = oldstate[s]
                     continue
                 if not force:
                     raise error.Abort(
                         _("commit with new subrepo %s excluded") % s)
             dirtyreason = wctx.sub(s).dirtyreason(True)
             if dirtyreason:
                 if not ui.configbool('ui', 'commitsubrepos'):
                     raise error.Abort(dirtyreason,
                         hint=_("use --subrepos for recursive commit"))
                 subs.append(s)
                 commitsubs.add(s)
             else:
                 bs = wctx.sub(s).basestate()
                 newstate[s] = (newstate[s][0], bs, newstate[s][2])
                 if oldstate.get(s, (None, None, None))[1] != bs:
                     subs.append(s)
 
         # check for removed subrepos
         for p in wctx.parents():
             r = [s for s in p.substate if s not in newstate]
             subs += [s for s in r if match(s)]
         if subs:
             if (not match('.hgsub') and
                 '.hgsub' in (wctx.modified() + wctx.added())):
                 raise error.Abort(_("can't commit subrepos without .hgsub"))
             status.modified.insert(0, '.hgsubstate')
 
     elif '.hgsub' in status.removed:
         # clean up .hgsubstate when .hgsub is removed
         if ('.hgsubstate' in wctx and
             '.hgsubstate' not in (status.modified + status.added +
                                   status.removed)):
             status.removed.insert(0, '.hgsubstate')
 
     return subs, commitsubs, newstate
 
 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 bytes(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 bytes(parent)
     else: # recursion reached top repo
         path = None
         if util.safehasattr(repo, '_subtoppath'):
             path = repo._subtoppath
         elif push and repo.ui.config('paths', 'default-push'):
             path = repo.ui.config('paths', 'default-push')
         elif repo.ui.config('paths', 'default'):
             path = repo.ui.config('paths', 'default')
         elif repo.shared():
             # chop off the .hg component to get the default path form.  This has
             # already run through vfsmod.vfs(..., realpath=True), so it doesn't
             # have problems with 'C:'
             return os.path.dirname(repo.sharedpath)
         if path:
             # issue5770: 'C:\' and 'C:' are not equivalent paths.  The former is
             # as expected: an absolute path to the root of the C: drive.  The
             # latter is a relative path, and works like so:
             #
             #   C:\>cd C:\some\path
             #   C:\>D:
             #   D:\>python -c "import os; print os.path.abspath('C:')"
             #   C:\some\path
             #
             #   D:\>python -c "import os; print os.path.abspath('C:relative')"
             #   C:\some\path\relative
             if util.hasdriveletter(path):
                 if len(path) == 2 or path[2:3] not in br'\/':
                     path = os.path.abspath(path)
             return path
 
     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 _auditsubrepopath(repo, path):
     # auditor doesn't check if the path itself is a symlink
     pathutil.pathauditor(repo.root)(path)
     if repo.wvfs.islink(path):
         raise error.Abort(_("subrepo '%s' traverses symbolic link") % path)
 
 SUBREPO_ALLOWED_DEFAULTS = {
     'hg': True,
     'git': False,
     'svn': False,
 }
 
 def _checktype(ui, kind):
     # subrepos.allowed is a master kill switch. If disabled, subrepos are
     # disabled period.
     if not ui.configbool('subrepos', 'allowed', True):
         raise error.Abort(_('subrepos not enabled'),
                           hint=_("see 'hg help config.subrepos' for details"))
 
     default = SUBREPO_ALLOWED_DEFAULTS.get(kind, False)
     if not ui.configbool('subrepos', '%s:allowed' % kind, default):
         raise error.Abort(_('%s subrepos not allowed') % kind,
                           hint=_("see 'hg help config.subrepos' for details"))
 
     if kind not in types:
         raise error.Abort(_('unknown subrepo type %s') % kind)
 
 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
 
     repo = ctx.repo()
     _auditsubrepopath(repo, path)
     state = ctx.substate[path]
     _checktype(repo.ui, 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
 
     repo = ctx.repo()
     _auditsubrepopath(repo, path)
     state = ctx.substate[path]
     _checktype(repo.ui, 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')
     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 addwebdirpath(self, serverpath, webconf):
         """Add the hgwebdir entries for this subrepo, and any of its subrepos.
 
         ``serverpath`` is the path component of the URL for this repo.
 
         ``webconf`` is the dictionary of hgwebdir entries.
         """
         pass
 
     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, missing=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.  If missing
         is true, check for deleted files.
         """
         raise NotImplementedError
 
     def dirtyreason(self, ignoreupdate=False, missing=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, missing=missing):
             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,
                                        missing=True)
         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, fm, fntemplate, 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, 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, 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, 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
         '''
 
     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 unshare(self):
         '''
         convert this repository from shared to normal storage.
         '''
 
     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 vfsmod.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)
 
     @annotatesubrepoerror
     def addwebdirpath(self, serverpath, webconf):
         cmdutil.addwebdirpath(self._repo, subrelpath(self), webconf)
 
     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 = node.hex(hashlib.sha1(vfs.tryread(relname)).digest())
             yield '%s = %s\n' % (relname, filehash)
 
     @propertycache
     def _cachestorehashvfs(self):
         return vfsmod.vfs(self._repo.vfs.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='wb', 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)
 
             self._repo.vfs.write('hgrc', util.tonativeeol(''.join(lines)))
 
     @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, fm, fntemplate, prefix, **opts):
         rev = self._state[1]
         ctx = self._repo[rev]
         return cmdutil.cat(self.ui, self._repo, ctx, match, fm, fntemplate,
                            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)
             logcmdutil.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, decode=True):
         self._get(self._state + ('hg',))
-        total = abstractsubrepo.archive(self, archiver, prefix, match)
+        files = self.files()
+        if match:
+            files = [f for f in files if match(f)]
         rev = self._state[1]
         ctx = self._repo[rev]
+        cmdutil._prefetchfiles(self._repo, ctx, files)
+        total = abstractsubrepo.archive(self, archiver, prefix, match)
         for subpath in ctx.substate:
             s = subrepo(ctx, subpath, True)
             submatch = matchmod.subdirmatcher(subpath, match)
             total += s.archive(archiver, prefix + self._path + '/', submatch,
                                decode)
         return total
 
     @annotatesubrepoerror
     def dirty(self, ignoreupdate=False, missing=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(missing=missing) # 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
         parentrepo = self._repo._subparent
 
         if revision in self._repo.unfiltered():
             # Allow shared subrepos tracked at null to setup the sharedpath
             if len(self._repo) != 0 or not parentrepo.shared():
                 return True
         self._repo._subsource = source
         srcurl = _abssource(self._repo)
         other = hg.peer(self._repo, {}, srcurl)
         if len(self._repo) == 0:
             # use self._repo.vfs instead of self.wvfs to remove .hg only
             self._repo.vfs.rmtree()
             if parentrepo.shared():
                 self.ui.status(_('sharing subrepo %s from %s\n')
                                % (subrelpath(self), srcurl))
                 shared = hg.share(self._repo._subparent.baseui,
                                   other, self._repo.root,
                                   update=False, bookmarks=False)
                 self._repo = shared.local()
             else:
                 self.ui.status(_('cloning subrepo %s from %s\n')
                                % (subrelpath(self), srcurl))
                 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 subrepository "%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 subrepository "%s"\n'
                               % subrelpath(self))
                 hg.update(self._repo, state[1])
             elif anc == dst:
                 self.ui.debug('skipping subrepository "%s"\n'
                               % subrelpath(self))
             else:
                 self.ui.debug('merging subrepository "%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, decode):
         rev = self._state[1]
         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(r'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[r'date'] = None
             opts[r'rev'] = substate[1]
 
             self.filerevert(*pats, **opts)
 
         # Update the repo to the revision specified in the given substate
         if not opts.get(r'dry_run'):
             self.get(substate, overwrite=True)
 
     def filerevert(self, *pats, **opts):
         ctx = self._repo[opts[r'rev']]
         parents = self._repo.dirstate.parents()
         if opts.get(r'all'):
             pats = ['set:modified()']
         else:
             pats = []
         cmdutil.revert(self.ui, self._repo, ctx, parents, *pats, **opts)
 
     def shortid(self, revid):
         return revid[:12]
 
     @annotatesubrepoerror
     def unshare(self):
         # 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
 
         # Nothing prevents a user from sharing in a repo, and then making that a
         # subrepo.  Alternately, the previous unshare attempt may have failed
         # part way through.  So recurse whether or not this layer is shared.
         if self._repo.shared():
             self.ui.status(_("unsharing subrepo '%s'\n") % self._relpath)
 
         hg.unshare(self.ui, self._repo)
 
     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[r'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(br'^(\d+)\.(\d+)', output)
         if not m:
             raise error.Abort(_('cannot retrieve svn tool version'))
         return (int(m.group(1)), int(m.group(2)))
 
     def _svnmissing(self):
         return not self.wvfs.exists('.svn')
 
     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)
 
     @annotatesubrepoerror
     def dirty(self, ignoreupdate=False, missing=False):
         if self._svnmissing():
             return self._state[1] != ''
         wcchanged = self._wcchanged()
         changed = wcchanged[0] or (missing and wcchanged[2])
         if not changed:
             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]))
 
         # SEC: check that the ssh url is safe
         util.checksafessh(state[0])
 
         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, 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, encoding.strtolocal(e.strerror)))
             elif pycompat.iswindows:
                 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,
                             encoding.strtolocal(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(br'^git version (\d+)\.(\d+)\.(\d+)', out)
         if m:
             return (int(m.group(1)), int(m.group(2)), int(m.group(3)))
 
         m = re.search(br'^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(b'git version 1.6.0')
         'ok'
         >>> _checkversion(b'git version 1.8.5')
         'ok'
         >>> _checkversion(b'git version 1.4.0')
         'abort'
         >>> _checkversion(b'git version 1.5.0')
         'warning'
         >>> _checkversion(b'git version 1.9-rc0')
         'ok'
         >>> _checkversion(b'git version 1.9.0.265.g81cdec2')
         'ok'
         >>> _checkversion(b'git version 1.9.0.GIT')
         'ok'
         >>> _checkversion(b'git version 12345')
         'unknown'
         >>> _checkversion(b'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')
         if self.ui._colormode and len(commands) and commands[0] == "diff":
             # insert the argument in the front,
             # the end of git diff arguments is used for paths
             commands.insert(1, '--color')
         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():
             # SEC: check for safe ssh url
             util.checksafessh(source)
 
             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 subrepository '
                                 '"%s"\n') % (revision, self._relpath))
 
     @annotatesubrepoerror
     def dirty(self, ignoreupdate=False, missing=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 '
                            'subrepository "%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 subrepository "%s"\n') % self._relpath)
                 return False
             self.ui.status(_('pushing branch %s of subrepository "%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 subrepository "%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(r'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, 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, fm, fntemplate, prefix, **opts):
         rev = self._state[1]
         if match.anypats():
             return 1 #No support for include/exclude yet
 
         if not match.files():
             return 1
 
         # TODO: add support for non-plain formatter (see cmdutil.cat())
         for f in match.files():
             output = self._gitcommand(["show", "%s:%s" % (rev, f)])
             fp = cmdutil.makefileobj(self._subparent, fntemplate,
                                      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(r'unknown'):
             command += ['--untracked-files=all']
         if opts.get(r'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(r'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[r'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(r'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(r'dry_run'):
             self.get(substate, overwrite=True)
         return []
 
     def shortid(self, revid):
         return revid[:7]
 
 types = {
     'hg': hgsubrepo,
     'svn': svnsubrepo,
     'git': gitsubrepo,
     }

tests/test-lfs-test-server.t

@@ -1,231 +1,259 b''
 #require lfs-test-server
 
   $ LFS_LISTEN="tcp://:$HGPORT"
   $ LFS_HOST="localhost:$HGPORT"
   $ LFS_PUBLIC=1
   $ export LFS_LISTEN LFS_HOST LFS_PUBLIC
 #if no-windows
   $ lfs-test-server &> lfs-server.log &
   $ echo $! >> $DAEMON_PIDS
 #else
   $ cat >> $TESTTMP/spawn.py <<EOF
   > import os
   > import subprocess
   > import sys
   > 
   > for path in os.environ["PATH"].split(os.pathsep):
   >     exe = os.path.join(path, 'lfs-test-server.exe')
   >     if os.path.exists(exe):
   >         with open('lfs-server.log', 'wb') as out:
   >             p = subprocess.Popen(exe, stdout=out, stderr=out)
   >             sys.stdout.write('%s\n' % p.pid)
   >             sys.exit(0)
   > sys.exit(1)
   > EOF
   $ $PYTHON $TESTTMP/spawn.py >> $DAEMON_PIDS
 #endif
 
   $ cat >> $HGRCPATH <<EOF
   > [extensions]
   > lfs=
   > [lfs]
   > url=http://foo:bar@$LFS_HOST/
   > track=all()
   > EOF
 
   $ hg init repo1
   $ cd repo1
   $ echo THIS-IS-LFS > a
   $ hg commit -m a -A a
 
 A push can be serviced directly from the usercache if it isn't in the local
 store.
 
   $ hg init ../repo2
   $ mv .hg/store/lfs .hg/store/lfs_
   $ hg push ../repo2 -v
   pushing to ../repo2
   searching for changes
   lfs: uploading 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b (12 bytes)
   lfs: processed: 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b
   lfs: uploaded 1 files (12 bytes)
   1 changesets found
   uncompressed size of bundle content:
        * (changelog) (glob)
        * (manifests) (glob)
        *  a (glob)
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 1 changes to 1 files
   calling hook pretxnchangegroup.lfs: hgext.lfs.checkrequireslfs
   $ mv .hg/store/lfs_ .hg/store/lfs
 
 Clear the cache to force a download
   $ rm -rf `hg config lfs.usercache`
   $ cd ../repo2
   $ hg update tip -v
   resolving manifests
   lfs: downloading 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b (12 bytes)
   lfs: adding 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b to the usercache
   lfs: processed: 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b
   getting a
   lfs: found 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b in the local lfs store
   1 files updated, 0 files merged, 0 files removed, 0 files unresolved
 
 When the server has some blobs already
 
   $ hg mv a b
   $ echo ANOTHER-LARGE-FILE > c
   $ echo ANOTHER-LARGE-FILE2 > d
   $ hg commit -m b-and-c -A b c d
   $ hg push ../repo1 -v | grep -v '^  '
   pushing to ../repo1
   searching for changes
   lfs: need to transfer 2 objects (39 bytes)
   lfs: uploading 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19 (20 bytes)
   lfs: processed: 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19
   lfs: uploading d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 (19 bytes)
   lfs: processed: d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998
   lfs: uploaded 2 files (39 bytes)
   1 changesets found
   uncompressed size of bundle content:
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 3 changes to 3 files
 
 Clear the cache to force a download
   $ rm -rf `hg config lfs.usercache`
   $ hg --repo ../repo1 update tip -v
   resolving manifests
   lfs: need to transfer 2 objects (39 bytes)
   lfs: downloading 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19 (20 bytes)
   lfs: adding 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19 to the usercache
   lfs: processed: 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19
   lfs: downloading d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 (19 bytes)
   lfs: adding d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 to the usercache
   lfs: processed: d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998
   getting b
   lfs: found 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b in the local lfs store
   getting c
   lfs: found d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 in the local lfs store
   getting d
   lfs: found 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19 in the local lfs store
   3 files updated, 0 files merged, 0 files removed, 0 files unresolved
 
 Test a corrupt file download, but clear the cache first to force a download.
 
   $ rm -rf `hg config lfs.usercache`
   $ cp $TESTTMP/lfs-content/d1/1e/1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 blob
   $ echo 'damage' > $TESTTMP/lfs-content/d1/1e/1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998
   $ rm ../repo1/.hg/store/lfs/objects/d1/1e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998
   $ rm ../repo1/*
 
   $ hg --repo ../repo1 update -C tip -v
   resolving manifests
   lfs: downloading d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 (19 bytes)
   abort: corrupt remote lfs object: d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998
   [255]
 
 The corrupted blob is not added to the usercache or local store
 
   $ test -f ../repo1/.hg/store/lfs/objects/d1/1e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998
   [1]
   $ test -f `hg config lfs.usercache`/d1/1e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998
   [1]
   $ cp blob $TESTTMP/lfs-content/d1/1e/1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998
 
 Test a corrupted file upload
 
   $ echo 'another lfs blob' > b
   $ hg ci -m 'another blob'
   $ echo 'damage' > .hg/store/lfs/objects/e6/59058e26b07b39d2a9c7145b3f99b41f797b6621c8076600e9cb7ee88291f0
   $ hg push -v ../repo1
   pushing to ../repo1
   searching for changes
   lfs: uploading e659058e26b07b39d2a9c7145b3f99b41f797b6621c8076600e9cb7ee88291f0 (17 bytes)
   abort: detected corrupt lfs object: e659058e26b07b39d2a9c7145b3f99b41f797b6621c8076600e9cb7ee88291f0
   (run hg verify)
   [255]
 
+Archive will prefetch blobs in a group
+
+  $ rm -rf .hg/store/lfs `hg config lfs.usercache`
+  $ hg archive -vr 1 ../archive
+  lfs: need to transfer 4 objects (63 bytes)
+  lfs: downloading 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b (12 bytes)
+  lfs: adding 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b to the usercache
+  lfs: processed: 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b
+  lfs: downloading 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b (12 bytes)
+  lfs: processed: 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b
+  lfs: downloading 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19 (20 bytes)
+  lfs: adding 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19 to the usercache
+  lfs: processed: 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19
+  lfs: downloading d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 (19 bytes)
+  lfs: adding d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 to the usercache
+  lfs: processed: d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998
+  lfs: found 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b in the local lfs store
+  lfs: found 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b in the local lfs store
+  lfs: found d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 in the local lfs store
+  lfs: found 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19 in the local lfs store
+  $ find ../archive | sort
+  ../archive
+  ../archive/.hg_archival.txt
+  ../archive/a
+  ../archive/b
+  ../archive/c
+  ../archive/d
+
 Revert will prefetch blobs in a group
 
   $ rm -rf .hg/store/lfs
   $ rm -rf `hg config lfs.usercache`
   $ rm *
   $ hg revert --all -r 1 -v
   adding a
   reverting b
   reverting c
   reverting d
   lfs: need to transfer 4 objects (63 bytes)
   lfs: downloading 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b (12 bytes)
   lfs: adding 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b to the usercache
   lfs: processed: 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b
   lfs: downloading 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b (12 bytes)
   lfs: processed: 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b
   lfs: downloading 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19 (20 bytes)
   lfs: adding 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19 to the usercache
   lfs: processed: 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19
   lfs: downloading d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 (19 bytes)
   lfs: adding d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 to the usercache
   lfs: processed: d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998
   lfs: found 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b in the local lfs store
   lfs: found d11e1a642b60813aee592094109b406089b8dff4cb157157f753418ec7857998 in the local lfs store
   lfs: found 37a65ab78d5ecda767e8622c248b5dbff1e68b1678ab0e730d5eb8601ec8ad19 in the local lfs store
   lfs: found 31cf46fbc4ecd458a0943c5b4881f1f5a6dd36c53d6167d5b69ac45149b38e5b in the local lfs store
 
 Check error message when the remote missed a blob:
 
   $ echo FFFFF > b
   $ hg commit -m b -A b
   $ echo FFFFF >> b
   $ hg commit -m b b
   $ rm -rf .hg/store/lfs
   $ rm -rf `hg config lfs.usercache`
   $ hg update -C '.^'
   abort: LFS server error. Remote object for "b" not found:(.*)! (re)
   [255]
 
 Check error message when object does not exist:
 
   $ cd $TESTTMP
   $ hg init test && cd test
   $ echo "[extensions]" >> .hg/hgrc
   $ echo "lfs=" >> .hg/hgrc
   $ echo "[lfs]" >> .hg/hgrc
   $ echo "threshold=1" >> .hg/hgrc
   $ echo a > a
   $ hg add a
   $ hg commit -m 'test'
   $ echo aaaaa > a
   $ hg commit -m 'largefile'
   $ hg debugdata .hg/store/data/a.i 1 # verify this is no the file content but includes "oid", the LFS "pointer".
   version https://git-lfs.github.com/spec/v1
   oid sha256:bdc26931acfb734b142a8d675f205becf27560dc461f501822de13274fe6fc8a
   size 6
   x-is-binary 0
   $ cd ..
   $ rm -rf `hg config lfs.usercache`
 
 (Restart the server in a different location so it no longer has the content)
 
   $ $PYTHON $RUNTESTDIR/killdaemons.py $DAEMON_PIDS
   $ rm $DAEMON_PIDS
   $ mkdir $TESTTMP/lfs-server2
   $ cd $TESTTMP/lfs-server2
 #if no-windows
   $ lfs-test-server &> lfs-server.log &
   $ echo $! >> $DAEMON_PIDS
 #else
   $ $PYTHON $TESTTMP/spawn.py >> $DAEMON_PIDS
 #endif
 
   $ cd $TESTTMP
   $ hg clone test test2
   updating to branch default
   abort: LFS server error. Remote object for "a" not found:(.*)! (re)
   [255]
 
   $ $PYTHON $RUNTESTDIR/killdaemons.py $DAEMON_PIDS