py3: convert the mode argument of os.fdopen to unicodes (1 of 2)

os.fdopen() does not accepts bytes as its second argument which represent the
mode in which the file is to be opened. This patch makes sure unicodes are
passed in py3 by using pycompat.sysstr().

mercurial/bundlerepo.py

@@ -1,557 +1,557 b''
 # bundlerepo.py - repository class for viewing uncompressed bundles
 #
 # Copyright 2006, 2007 Benoit Boissinot <[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.
 
 """Repository class for viewing uncompressed bundles.
 
 This provides a read-only repository interface to bundles as if they
 were part of the actual repository.
 """
 
 from __future__ import absolute_import
 
 import os
 import shutil
 import tempfile
 
 from .i18n import _
 from .node import nullid
 
 from . import (
     bundle2,
     changegroup,
     changelog,
     cmdutil,
     discovery,
     error,
     exchange,
     filelog,
     localrepo,
     manifest,
     mdiff,
     node as nodemod,
     pathutil,
     phases,
     pycompat,
     revlog,
     scmutil,
     util,
 )
 
 class bundlerevlog(revlog.revlog):
     def __init__(self, opener, indexfile, bundle, linkmapper):
         # How it works:
         # To retrieve a revision, we need to know the offset of the revision in
         # the bundle (an unbundle object). We store this offset in the index
         # (start). The base of the delta is stored in the base field.
         #
         # To differentiate a rev in the bundle from a rev in the revlog, we
         # check revision against repotiprev.
         opener = scmutil.readonlyvfs(opener)
         revlog.revlog.__init__(self, opener, indexfile)
         self.bundle = bundle
         n = len(self)
         self.repotiprev = n - 1
         chain = None
         self.bundlerevs = set() # used by 'bundle()' revset expression
         getchunk = lambda: bundle.deltachunk(chain)
         for chunkdata in iter(getchunk, {}):
             node = chunkdata['node']
             p1 = chunkdata['p1']
             p2 = chunkdata['p2']
             cs = chunkdata['cs']
             deltabase = chunkdata['deltabase']
             delta = chunkdata['delta']
 
             size = len(delta)
             start = bundle.tell() - size
 
             link = linkmapper(cs)
             if node in self.nodemap:
                 # this can happen if two branches make the same change
                 chain = node
                 self.bundlerevs.add(self.nodemap[node])
                 continue
 
             for p in (p1, p2):
                 if p not in self.nodemap:
                     raise error.LookupError(p, self.indexfile,
                                             _("unknown parent"))
 
             if deltabase not in self.nodemap:
                 raise LookupError(deltabase, self.indexfile,
                                   _('unknown delta base'))
 
             baserev = self.rev(deltabase)
             # start, size, full unc. size, base (unused), link, p1, p2, node
             e = (revlog.offset_type(start, 0), size, -1, baserev, link,
                  self.rev(p1), self.rev(p2), node)
             self.index.insert(-1, e)
             self.nodemap[node] = n
             self.bundlerevs.add(n)
             chain = node
             n += 1
 
     def _chunk(self, rev):
         # Warning: in case of bundle, the diff is against what we stored as
         # delta base, not against rev - 1
         # XXX: could use some caching
         if rev <= self.repotiprev:
             return revlog.revlog._chunk(self, rev)
         self.bundle.seek(self.start(rev))
         return self.bundle.read(self.length(rev))
 
     def revdiff(self, rev1, rev2):
         """return or calculate a delta between two revisions"""
         if rev1 > self.repotiprev and rev2 > self.repotiprev:
             # hot path for bundle
             revb = self.index[rev2][3]
             if revb == rev1:
                 return self._chunk(rev2)
         elif rev1 <= self.repotiprev and rev2 <= self.repotiprev:
             return revlog.revlog.revdiff(self, rev1, rev2)
 
         return mdiff.textdiff(self.revision(self.node(rev1)),
                               self.revision(self.node(rev2)))
 
     def revision(self, nodeorrev, raw=False):
         """return an uncompressed revision of a given node or revision
         number.
         """
         if isinstance(nodeorrev, int):
             rev = nodeorrev
             node = self.node(rev)
         else:
             node = nodeorrev
             rev = self.rev(node)
 
         if node == nullid:
             return ""
 
         text = None
         chain = []
         iterrev = rev
         # reconstruct the revision if it is from a changegroup
         while iterrev > self.repotiprev:
             if self._cache and self._cache[1] == iterrev:
                 text = self._cache[2]
                 break
             chain.append(iterrev)
             iterrev = self.index[iterrev][3]
         if text is None:
             text = self.baserevision(iterrev)
 
         while chain:
             delta = self._chunk(chain.pop())
             text = mdiff.patches(text, [delta])
 
         text, validatehash = self._processflags(text, self.flags(rev),
                                                 'read', raw=raw)
         if validatehash:
             self.checkhash(text, node, rev=rev)
         self._cache = (node, rev, text)
         return text
 
     def baserevision(self, nodeorrev):
         # Revlog subclasses may override 'revision' method to modify format of
         # content retrieved from revlog. To use bundlerevlog with such class one
         # needs to override 'baserevision' and make more specific call here.
         return revlog.revlog.revision(self, nodeorrev)
 
     def addrevision(self, text, transaction, link, p1=None, p2=None, d=None):
         raise NotImplementedError
     def addgroup(self, revs, linkmapper, transaction):
         raise NotImplementedError
     def strip(self, rev, minlink):
         raise NotImplementedError
     def checksize(self):
         raise NotImplementedError
 
 class bundlechangelog(bundlerevlog, changelog.changelog):
     def __init__(self, opener, bundle):
         changelog.changelog.__init__(self, opener)
         linkmapper = lambda x: x
         bundlerevlog.__init__(self, opener, self.indexfile, bundle,
                               linkmapper)
 
     def baserevision(self, nodeorrev):
         # Although changelog doesn't override 'revision' method, some extensions
         # may replace this class with another that does. Same story with
         # manifest and filelog classes.
 
         # This bypasses filtering on changelog.node() and rev() because we need
         # revision text of the bundle base even if it is hidden.
         oldfilter = self.filteredrevs
         try:
             self.filteredrevs = ()
             return changelog.changelog.revision(self, nodeorrev)
         finally:
             self.filteredrevs = oldfilter
 
 class bundlemanifest(bundlerevlog, manifest.manifestrevlog):
     def __init__(self, opener, bundle, linkmapper, dirlogstarts=None, dir=''):
         manifest.manifestrevlog.__init__(self, opener, dir=dir)
         bundlerevlog.__init__(self, opener, self.indexfile, bundle,
                               linkmapper)
         if dirlogstarts is None:
             dirlogstarts = {}
             if self.bundle.version == "03":
                 dirlogstarts = _getfilestarts(self.bundle)
         self._dirlogstarts = dirlogstarts
         self._linkmapper = linkmapper
 
     def baserevision(self, nodeorrev):
         node = nodeorrev
         if isinstance(node, int):
             node = self.node(node)
 
         if node in self.fulltextcache:
             result = self.fulltextcache[node].tostring()
         else:
             result = manifest.manifestrevlog.revision(self, nodeorrev)
         return result
 
     def dirlog(self, d):
         if d in self._dirlogstarts:
             self.bundle.seek(self._dirlogstarts[d])
             return bundlemanifest(
                 self.opener, self.bundle, self._linkmapper,
                 self._dirlogstarts, dir=d)
         return super(bundlemanifest, self).dirlog(d)
 
 class bundlefilelog(bundlerevlog, filelog.filelog):
     def __init__(self, opener, path, bundle, linkmapper):
         filelog.filelog.__init__(self, opener, path)
         bundlerevlog.__init__(self, opener, self.indexfile, bundle,
                               linkmapper)
 
     def baserevision(self, nodeorrev):
         return filelog.filelog.revision(self, nodeorrev)
 
 class bundlepeer(localrepo.localpeer):
     def canpush(self):
         return False
 
 class bundlephasecache(phases.phasecache):
     def __init__(self, *args, **kwargs):
         super(bundlephasecache, self).__init__(*args, **kwargs)
         if util.safehasattr(self, 'opener'):
             self.opener = scmutil.readonlyvfs(self.opener)
 
     def write(self):
         raise NotImplementedError
 
     def _write(self, fp):
         raise NotImplementedError
 
     def _updateroots(self, phase, newroots, tr):
         self.phaseroots[phase] = newroots
         self.invalidate()
         self.dirty = True
 
 def _getfilestarts(bundle):
     bundlefilespos = {}
     for chunkdata in iter(bundle.filelogheader, {}):
         fname = chunkdata['filename']
         bundlefilespos[fname] = bundle.tell()
         for chunk in iter(lambda: bundle.deltachunk(None), {}):
             pass
     return bundlefilespos
 
 class bundlerepository(localrepo.localrepository):
     def __init__(self, ui, path, bundlename):
         def _writetempbundle(read, suffix, header=''):
             """Write a temporary file to disk
 
             This is closure because we need to make sure this tracked by
             self.tempfile for cleanup purposes."""
             fdtemp, temp = self.vfs.mkstemp(prefix="hg-bundle-",
                                             suffix=".hg10un")
             self.tempfile = temp
 
-            with os.fdopen(fdtemp, 'wb') as fptemp:
+            with os.fdopen(fdtemp, pycompat.sysstr('wb')) as fptemp:
                 fptemp.write(header)
                 while True:
                     chunk = read(2**18)
                     if not chunk:
                         break
                     fptemp.write(chunk)
 
             return self.vfs.open(self.tempfile, mode="rb")
         self._tempparent = None
         try:
             localrepo.localrepository.__init__(self, ui, path)
         except error.RepoError:
             self._tempparent = tempfile.mkdtemp()
             localrepo.instance(ui, self._tempparent, 1)
             localrepo.localrepository.__init__(self, ui, self._tempparent)
         self.ui.setconfig('phases', 'publish', False, 'bundlerepo')
 
         if path:
             self._url = 'bundle:' + util.expandpath(path) + '+' + bundlename
         else:
             self._url = 'bundle:' + bundlename
 
         self.tempfile = None
         f = util.posixfile(bundlename, "rb")
         self.bundlefile = self.bundle = exchange.readbundle(ui, f, bundlename)
 
         if isinstance(self.bundle, bundle2.unbundle20):
             cgstream = None
             for part in self.bundle.iterparts():
                 if part.type == 'changegroup':
                     if cgstream is not None:
                         raise NotImplementedError("can't process "
                                                   "multiple changegroups")
                     cgstream = part
                     version = part.params.get('version', '01')
                     legalcgvers = changegroup.supportedincomingversions(self)
                     if version not in legalcgvers:
                         msg = _('Unsupported changegroup version: %s')
                         raise error.Abort(msg % version)
                     if self.bundle.compressed():
                         cgstream = _writetempbundle(part.read,
                                                     ".cg%sun" % version)
 
             if cgstream is None:
                 raise error.Abort(_('No changegroups found'))
             cgstream.seek(0)
 
             self.bundle = changegroup.getunbundler(version, cgstream, 'UN')
 
         elif self.bundle.compressed():
             f = _writetempbundle(self.bundle.read, '.hg10un', header='HG10UN')
             self.bundlefile = self.bundle = exchange.readbundle(ui, f,
                                                                 bundlename,
                                                                 self.vfs)
 
         # dict with the mapping 'filename' -> position in the bundle
         self.bundlefilespos = {}
 
         self.firstnewrev = self.changelog.repotiprev + 1
         phases.retractboundary(self, None, phases.draft,
                                [ctx.node() for ctx in self[self.firstnewrev:]])
 
     @localrepo.unfilteredpropertycache
     def _phasecache(self):
         return bundlephasecache(self, self._phasedefaults)
 
     @localrepo.unfilteredpropertycache
     def changelog(self):
         # consume the header if it exists
         self.bundle.changelogheader()
         c = bundlechangelog(self.svfs, self.bundle)
         self.manstart = self.bundle.tell()
         return c
 
     def _constructmanifest(self):
         self.bundle.seek(self.manstart)
         # consume the header if it exists
         self.bundle.manifestheader()
         linkmapper = self.unfiltered().changelog.rev
         m = bundlemanifest(self.svfs, self.bundle, linkmapper)
         self.filestart = self.bundle.tell()
         return m
 
     @localrepo.unfilteredpropertycache
     def manstart(self):
         self.changelog
         return self.manstart
 
     @localrepo.unfilteredpropertycache
     def filestart(self):
         self.manifestlog
         return self.filestart
 
     def url(self):
         return self._url
 
     def file(self, f):
         if not self.bundlefilespos:
             self.bundle.seek(self.filestart)
             self.bundlefilespos = _getfilestarts(self.bundle)
 
         if f in self.bundlefilespos:
             self.bundle.seek(self.bundlefilespos[f])
             linkmapper = self.unfiltered().changelog.rev
             return bundlefilelog(self.svfs, f, self.bundle, linkmapper)
         else:
             return filelog.filelog(self.svfs, f)
 
     def close(self):
         """Close assigned bundle file immediately."""
         self.bundlefile.close()
         if self.tempfile is not None:
             self.vfs.unlink(self.tempfile)
         if self._tempparent:
             shutil.rmtree(self._tempparent, True)
 
     def cancopy(self):
         return False
 
     def peer(self):
         return bundlepeer(self)
 
     def getcwd(self):
         return pycompat.getcwd() # always outside the repo
 
     # Check if parents exist in localrepo before setting
     def setparents(self, p1, p2=nullid):
         p1rev = self.changelog.rev(p1)
         p2rev = self.changelog.rev(p2)
         msg = _("setting parent to node %s that only exists in the bundle\n")
         if self.changelog.repotiprev < p1rev:
             self.ui.warn(msg % nodemod.hex(p1))
         if self.changelog.repotiprev < p2rev:
             self.ui.warn(msg % nodemod.hex(p2))
         return super(bundlerepository, self).setparents(p1, p2)
 
 def instance(ui, path, create):
     if create:
         raise error.Abort(_('cannot create new bundle repository'))
     # internal config: bundle.mainreporoot
     parentpath = ui.config("bundle", "mainreporoot", "")
     if not parentpath:
         # try to find the correct path to the working directory repo
         parentpath = cmdutil.findrepo(pycompat.getcwd())
         if parentpath is None:
             parentpath = ''
     if parentpath:
         # Try to make the full path relative so we get a nice, short URL.
         # In particular, we don't want temp dir names in test outputs.
         cwd = pycompat.getcwd()
         if parentpath == cwd:
             parentpath = ''
         else:
             cwd = pathutil.normasprefix(cwd)
             if parentpath.startswith(cwd):
                 parentpath = parentpath[len(cwd):]
     u = util.url(path)
     path = u.localpath()
     if u.scheme == 'bundle':
         s = path.split("+", 1)
         if len(s) == 1:
             repopath, bundlename = parentpath, s[0]
         else:
             repopath, bundlename = s
     else:
         repopath, bundlename = parentpath, path
     return bundlerepository(ui, repopath, bundlename)
 
 class bundletransactionmanager(object):
     def transaction(self):
         return None
 
     def close(self):
         raise NotImplementedError
 
     def release(self):
         raise NotImplementedError
 
 def getremotechanges(ui, repo, other, onlyheads=None, bundlename=None,
                      force=False):
     '''obtains a bundle of changes incoming from other
 
     "onlyheads" restricts the returned changes to those reachable from the
       specified heads.
     "bundlename", if given, stores the bundle to this file path permanently;
       otherwise it's stored to a temp file and gets deleted again when you call
       the returned "cleanupfn".
     "force" indicates whether to proceed on unrelated repos.
 
     Returns a tuple (local, csets, cleanupfn):
 
     "local" is a local repo from which to obtain the actual incoming
       changesets; it is a bundlerepo for the obtained bundle when the
       original "other" is remote.
     "csets" lists the incoming changeset node ids.
     "cleanupfn" must be called without arguments when you're done processing
       the changes; it closes both the original "other" and the one returned
       here.
     '''
     tmp = discovery.findcommonincoming(repo, other, heads=onlyheads,
                                        force=force)
     common, incoming, rheads = tmp
     if not incoming:
         try:
             if bundlename:
                 os.unlink(bundlename)
         except OSError:
             pass
         return repo, [], other.close
 
     commonset = set(common)
     rheads = [x for x in rheads if x not in commonset]
 
     bundle = None
     bundlerepo = None
     localrepo = other.local()
     if bundlename or not localrepo:
         # create a bundle (uncompressed if other repo is not local)
 
         # developer config: devel.legacy.exchange
         legexc = ui.configlist('devel', 'legacy.exchange')
         forcebundle1 = 'bundle2' not in legexc and 'bundle1' in legexc
         canbundle2 = (not forcebundle1
                       and other.capable('getbundle')
                       and other.capable('bundle2'))
         if canbundle2:
             kwargs = {}
             kwargs['common'] = common
             kwargs['heads'] = rheads
             kwargs['bundlecaps'] = exchange.caps20to10(repo)
             kwargs['cg'] = True
             b2 = other.getbundle('incoming', **kwargs)
             fname = bundle = changegroup.writechunks(ui, b2._forwardchunks(),
                                                      bundlename)
         else:
             if other.capable('getbundle'):
                 cg = other.getbundle('incoming', common=common, heads=rheads)
             elif onlyheads is None and not other.capable('changegroupsubset'):
                 # compat with older servers when pulling all remote heads
                 cg = other.changegroup(incoming, "incoming")
                 rheads = None
             else:
                 cg = other.changegroupsubset(incoming, rheads, 'incoming')
             if localrepo:
                 bundletype = "HG10BZ"
             else:
                 bundletype = "HG10UN"
             fname = bundle = bundle2.writebundle(ui, cg, bundlename,
                                                      bundletype)
         # keep written bundle?
         if bundlename:
             bundle = None
         if not localrepo:
             # use the created uncompressed bundlerepo
             localrepo = bundlerepo = bundlerepository(repo.baseui, repo.root,
                                                       fname)
             # this repo contains local and other now, so filter out local again
             common = repo.heads()
     if localrepo:
         # Part of common may be remotely filtered
         # So use an unfiltered version
         # The discovery process probably need cleanup to avoid that
         localrepo = localrepo.unfiltered()
 
     csets = localrepo.changelog.findmissing(common, rheads)
 
     if bundlerepo:
         reponodes = [ctx.node() for ctx in bundlerepo[bundlerepo.firstnewrev:]]
         remotephases = other.listkeys('phases')
 
         pullop = exchange.pulloperation(bundlerepo, other, heads=reponodes)
         pullop.trmanager = bundletransactionmanager()
         exchange._pullapplyphases(pullop, remotephases)
 
     def cleanup():
         if bundlerepo:
             bundlerepo.close()
         if bundle:
             os.unlink(bundle)
         other.close()
 
     return (localrepo, csets, cleanup)

mercurial/chgserver.py

@@ -1,588 +1,588 b''
 # chgserver.py - command server extension for cHg
 #
 # Copyright 2011 Yuya Nishihara <[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.
 
 """command server extension for cHg
 
 'S' channel (read/write)
     propagate ui.system() request to client
 
 'attachio' command
     attach client's stdio passed by sendmsg()
 
 'chdir' command
     change current directory
 
 'setenv' command
     replace os.environ completely
 
 'setumask' command
     set umask
 
 'validate' command
     reload the config and check if the server is up to date
 
 Config
 ------
 
 ::
 
   [chgserver]
   idletimeout = 3600 # seconds, after which an idle server will exit
   skiphash = False   # whether to skip config or env change checks
 """
 
 from __future__ import absolute_import
 
 import errno
 import hashlib
 import inspect
 import os
 import re
 import struct
 import time
 
 from .i18n import _
 
 from . import (
     commandserver,
     encoding,
     error,
     extensions,
     osutil,
     pycompat,
     util,
 )
 
 _log = commandserver.log
 
 def _hashlist(items):
     """return sha1 hexdigest for a list"""
     return hashlib.sha1(str(items)).hexdigest()
 
 # sensitive config sections affecting confighash
 _configsections = [
     'alias',  # affects global state commands.table
     'extdiff',  # uisetup will register new commands
     'extensions',
 ]
 
 # sensitive environment variables affecting confighash
 _envre = re.compile(r'''\A(?:
                     CHGHG
                     |HG(?:[A-Z].*)?
                     |LANG(?:UAGE)?
                     |LC_.*
                     |LD_.*
                     |PATH
                     |PYTHON.*
                     |TERM(?:INFO)?
                     |TZ
                     )\Z''', re.X)
 
 def _confighash(ui):
     """return a quick hash for detecting config/env changes
 
     confighash is the hash of sensitive config items and environment variables.
 
     for chgserver, it is designed that once confighash changes, the server is
     not qualified to serve its client and should redirect the client to a new
     server. different from mtimehash, confighash change will not mark the
     server outdated and exit since the user can have different configs at the
     same time.
     """
     sectionitems = []
     for section in _configsections:
         sectionitems.append(ui.configitems(section))
     sectionhash = _hashlist(sectionitems)
     envitems = [(k, v) for k, v in encoding.environ.iteritems()
                 if _envre.match(k)]
     envhash = _hashlist(sorted(envitems))
     return sectionhash[:6] + envhash[:6]
 
 def _getmtimepaths(ui):
     """get a list of paths that should be checked to detect change
 
     The list will include:
     - extensions (will not cover all files for complex extensions)
     - mercurial/__version__.py
     - python binary
     """
     modules = [m for n, m in extensions.extensions(ui)]
     try:
         from . import __version__
         modules.append(__version__)
     except ImportError:
         pass
     files = [pycompat.sysexecutable]
     for m in modules:
         try:
             files.append(inspect.getabsfile(m))
         except TypeError:
             pass
     return sorted(set(files))
 
 def _mtimehash(paths):
     """return a quick hash for detecting file changes
 
     mtimehash calls stat on given paths and calculate a hash based on size and
     mtime of each file. mtimehash does not read file content because reading is
     expensive. therefore it's not 100% reliable for detecting content changes.
     it's possible to return different hashes for same file contents.
     it's also possible to return a same hash for different file contents for
     some carefully crafted situation.
 
     for chgserver, it is designed that once mtimehash changes, the server is
     considered outdated immediately and should no longer provide service.
 
     mtimehash is not included in confighash because we only know the paths of
     extensions after importing them (there is imp.find_module but that faces
     race conditions). We need to calculate confighash without importing.
     """
     def trystat(path):
         try:
             st = os.stat(path)
             return (st.st_mtime, st.st_size)
         except OSError:
             # could be ENOENT, EPERM etc. not fatal in any case
             pass
     return _hashlist(map(trystat, paths))[:12]
 
 class hashstate(object):
     """a structure storing confighash, mtimehash, paths used for mtimehash"""
     def __init__(self, confighash, mtimehash, mtimepaths):
         self.confighash = confighash
         self.mtimehash = mtimehash
         self.mtimepaths = mtimepaths
 
     @staticmethod
     def fromui(ui, mtimepaths=None):
         if mtimepaths is None:
             mtimepaths = _getmtimepaths(ui)
         confighash = _confighash(ui)
         mtimehash = _mtimehash(mtimepaths)
         _log('confighash = %s mtimehash = %s\n' % (confighash, mtimehash))
         return hashstate(confighash, mtimehash, mtimepaths)
 
 def _newchgui(srcui, csystem, attachio):
     class chgui(srcui.__class__):
         def __init__(self, src=None):
             super(chgui, self).__init__(src)
             if src:
                 self._csystem = getattr(src, '_csystem', csystem)
             else:
                 self._csystem = csystem
 
         def system(self, cmd, environ=None, cwd=None, onerr=None,
                    errprefix=None):
             # fallback to the original system method if the output needs to be
             # captured (to self._buffers), or the output stream is not stdout
             # (e.g. stderr, cStringIO), because the chg client is not aware of
             # these situations and will behave differently (write to stdout).
             if (any(s[1] for s in self._bufferstates)
                 or not util.safehasattr(self.fout, 'fileno')
                 or self.fout.fileno() != util.stdout.fileno()):
                 return super(chgui, self).system(cmd, environ, cwd, onerr,
                                                  errprefix)
             self.flush()
             rc = self._csystem(cmd, util.shellenviron(environ), cwd)
             if rc and onerr:
                 errmsg = '%s %s' % (os.path.basename(cmd.split(None, 1)[0]),
                                     util.explainexit(rc)[0])
                 if errprefix:
                     errmsg = '%s: %s' % (errprefix, errmsg)
                 raise onerr(errmsg)
             return rc
 
         def _runpager(self, cmd):
             self._csystem(cmd, util.shellenviron(), type='pager',
                           cmdtable={'attachio': attachio})
 
     return chgui(srcui)
 
 def _loadnewui(srcui, args):
     from . import dispatch  # avoid cycle
 
     newui = srcui.__class__.load()
     for a in ['fin', 'fout', 'ferr', 'environ']:
         setattr(newui, a, getattr(srcui, a))
     if util.safehasattr(srcui, '_csystem'):
         newui._csystem = srcui._csystem
 
     # command line args
     args = args[:]
     dispatch._parseconfig(newui, dispatch._earlygetopt(['--config'], args))
 
     # stolen from tortoisehg.util.copydynamicconfig()
     for section, name, value in srcui.walkconfig():
         source = srcui.configsource(section, name)
         if ':' in source or source == '--config':
             # path:line or command line
             continue
         newui.setconfig(section, name, value, source)
 
     # load wd and repo config, copied from dispatch.py
     cwds = dispatch._earlygetopt(['--cwd'], args)
     cwd = cwds and os.path.realpath(cwds[-1]) or None
     rpath = dispatch._earlygetopt(["-R", "--repository", "--repo"], args)
     path, newlui = dispatch._getlocal(newui, rpath, wd=cwd)
 
     return (newui, newlui)
 
 class channeledsystem(object):
     """Propagate ui.system() request in the following format:
 
     payload length (unsigned int),
     type, '\0',
     cmd, '\0',
     cwd, '\0',
     envkey, '=', val, '\0',
     ...
     envkey, '=', val
 
     if type == 'system', waits for:
 
     exitcode length (unsigned int),
     exitcode (int)
 
     if type == 'pager', repetitively waits for a command name ending with '\n'
     and executes it defined by cmdtable, or exits the loop if the command name
     is empty.
     """
     def __init__(self, in_, out, channel):
         self.in_ = in_
         self.out = out
         self.channel = channel
 
     def __call__(self, cmd, environ, cwd=None, type='system', cmdtable=None):
         args = [type, util.quotecommand(cmd), os.path.abspath(cwd or '.')]
         args.extend('%s=%s' % (k, v) for k, v in environ.iteritems())
         data = '\0'.join(args)
         self.out.write(struct.pack('>cI', self.channel, len(data)))
         self.out.write(data)
         self.out.flush()
 
         if type == 'system':
             length = self.in_.read(4)
             length, = struct.unpack('>I', length)
             if length != 4:
                 raise error.Abort(_('invalid response'))
             rc, = struct.unpack('>i', self.in_.read(4))
             return rc
         elif type == 'pager':
             while True:
                 cmd = self.in_.readline()[:-1]
                 if not cmd:
                     break
                 if cmdtable and cmd in cmdtable:
                     _log('pager subcommand: %s' % cmd)
                     cmdtable[cmd]()
                 else:
                     raise error.Abort(_('unexpected command: %s') % cmd)
         else:
             raise error.ProgrammingError('invalid S channel type: %s' % type)
 
 _iochannels = [
     # server.ch, ui.fp, mode
-    ('cin', 'fin', 'rb'),
-    ('cout', 'fout', 'wb'),
-    ('cerr', 'ferr', 'wb'),
+    ('cin', 'fin', pycompat.sysstr('rb')),
+    ('cout', 'fout', pycompat.sysstr('wb')),
+    ('cerr', 'ferr', pycompat.sysstr('wb')),
 ]
 
 class chgcmdserver(commandserver.server):
     def __init__(self, ui, repo, fin, fout, sock, hashstate, baseaddress):
         super(chgcmdserver, self).__init__(
             _newchgui(ui, channeledsystem(fin, fout, 'S'), self.attachio),
             repo, fin, fout)
         self.clientsock = sock
         self._oldios = []  # original (self.ch, ui.fp, fd) before "attachio"
         self.hashstate = hashstate
         self.baseaddress = baseaddress
         if hashstate is not None:
             self.capabilities = self.capabilities.copy()
             self.capabilities['validate'] = chgcmdserver.validate
 
     def cleanup(self):
         super(chgcmdserver, self).cleanup()
         # dispatch._runcatch() does not flush outputs if exception is not
         # handled by dispatch._dispatch()
         self.ui.flush()
         self._restoreio()
 
     def attachio(self):
         """Attach to client's stdio passed via unix domain socket; all
         channels except cresult will no longer be used
         """
         # tell client to sendmsg() with 1-byte payload, which makes it
         # distinctive from "attachio\n" command consumed by client.read()
         self.clientsock.sendall(struct.pack('>cI', 'I', 1))
         clientfds = osutil.recvfds(self.clientsock.fileno())
         _log('received fds: %r\n' % clientfds)
 
         ui = self.ui
         ui.flush()
         first = self._saveio()
         for fd, (cn, fn, mode) in zip(clientfds, _iochannels):
             assert fd > 0
             fp = getattr(ui, fn)
             os.dup2(fd, fp.fileno())
             os.close(fd)
             if not first:
                 continue
             # reset buffering mode when client is first attached. as we want
             # to see output immediately on pager, the mode stays unchanged
             # when client re-attached. ferr is unchanged because it should
             # be unbuffered no matter if it is a tty or not.
             if fn == 'ferr':
                 newfp = fp
             else:
                 # make it line buffered explicitly because the default is
                 # decided on first write(), where fout could be a pager.
                 if fp.isatty():
                     bufsize = 1  # line buffered
                 else:
                     bufsize = -1  # system default
                 newfp = os.fdopen(fp.fileno(), mode, bufsize)
                 setattr(ui, fn, newfp)
             setattr(self, cn, newfp)
 
         self.cresult.write(struct.pack('>i', len(clientfds)))
 
     def _saveio(self):
         if self._oldios:
             return False
         ui = self.ui
         for cn, fn, _mode in _iochannels:
             ch = getattr(self, cn)
             fp = getattr(ui, fn)
             fd = os.dup(fp.fileno())
             self._oldios.append((ch, fp, fd))
         return True
 
     def _restoreio(self):
         ui = self.ui
         for (ch, fp, fd), (cn, fn, _mode) in zip(self._oldios, _iochannels):
             newfp = getattr(ui, fn)
             # close newfp while it's associated with client; otherwise it
             # would be closed when newfp is deleted
             if newfp is not fp:
                 newfp.close()
             # restore original fd: fp is open again
             os.dup2(fd, fp.fileno())
             os.close(fd)
             setattr(self, cn, ch)
             setattr(ui, fn, fp)
         del self._oldios[:]
 
     def validate(self):
         """Reload the config and check if the server is up to date
 
         Read a list of '\0' separated arguments.
         Write a non-empty list of '\0' separated instruction strings or '\0'
         if the list is empty.
         An instruction string could be either:
             - "unlink $path", the client should unlink the path to stop the
               outdated server.
             - "redirect $path", the client should attempt to connect to $path
               first. If it does not work, start a new server. It implies
               "reconnect".
             - "exit $n", the client should exit directly with code n.
               This may happen if we cannot parse the config.
             - "reconnect", the client should close the connection and
               reconnect.
         If neither "reconnect" nor "redirect" is included in the instruction
         list, the client can continue with this server after completing all
         the instructions.
         """
         from . import dispatch  # avoid cycle
 
         args = self._readlist()
         try:
             self.ui, lui = _loadnewui(self.ui, args)
         except error.ParseError as inst:
             dispatch._formatparse(self.ui.warn, inst)
             self.ui.flush()
             self.cresult.write('exit 255')
             return
         newhash = hashstate.fromui(lui, self.hashstate.mtimepaths)
         insts = []
         if newhash.mtimehash != self.hashstate.mtimehash:
             addr = _hashaddress(self.baseaddress, self.hashstate.confighash)
             insts.append('unlink %s' % addr)
             # mtimehash is empty if one or more extensions fail to load.
             # to be compatible with hg, still serve the client this time.
             if self.hashstate.mtimehash:
                 insts.append('reconnect')
         if newhash.confighash != self.hashstate.confighash:
             addr = _hashaddress(self.baseaddress, newhash.confighash)
             insts.append('redirect %s' % addr)
         _log('validate: %s\n' % insts)
         self.cresult.write('\0'.join(insts) or '\0')
 
     def chdir(self):
         """Change current directory
 
         Note that the behavior of --cwd option is bit different from this.
         It does not affect --config parameter.
         """
         path = self._readstr()
         if not path:
             return
         _log('chdir to %r\n' % path)
         os.chdir(path)
 
     def setumask(self):
         """Change umask"""
         mask = struct.unpack('>I', self._read(4))[0]
         _log('setumask %r\n' % mask)
         os.umask(mask)
 
     def runcommand(self):
         return super(chgcmdserver, self).runcommand()
 
     def setenv(self):
         """Clear and update os.environ
 
         Note that not all variables can make an effect on the running process.
         """
         l = self._readlist()
         try:
             newenv = dict(s.split('=', 1) for s in l)
         except ValueError:
             raise ValueError('unexpected value in setenv request')
         _log('setenv: %r\n' % sorted(newenv.keys()))
         encoding.environ.clear()
         encoding.environ.update(newenv)
 
     capabilities = commandserver.server.capabilities.copy()
     capabilities.update({'attachio': attachio,
                          'chdir': chdir,
                          'runcommand': runcommand,
                          'setenv': setenv,
                          'setumask': setumask})
 
     if util.safehasattr(osutil, 'setprocname'):
         def setprocname(self):
             """Change process title"""
             name = self._readstr()
             _log('setprocname: %r\n' % name)
             osutil.setprocname(name)
         capabilities['setprocname'] = setprocname
 
 def _tempaddress(address):
     return '%s.%d.tmp' % (address, os.getpid())
 
 def _hashaddress(address, hashstr):
     # if the basename of address contains '.', use only the left part. this
     # makes it possible for the client to pass 'server.tmp$PID' and follow by
     # an atomic rename to avoid locking when spawning new servers.
     dirname, basename = os.path.split(address)
     basename = basename.split('.', 1)[0]
     return '%s-%s' % (os.path.join(dirname, basename), hashstr)
 
 class chgunixservicehandler(object):
     """Set of operations for chg services"""
 
     pollinterval = 1  # [sec]
 
     def __init__(self, ui):
         self.ui = ui
         self._idletimeout = ui.configint('chgserver', 'idletimeout', 3600)
         self._lastactive = time.time()
 
     def bindsocket(self, sock, address):
         self._inithashstate(address)
         self._checkextensions()
         self._bind(sock)
         self._createsymlink()
 
     def _inithashstate(self, address):
         self._baseaddress = address
         if self.ui.configbool('chgserver', 'skiphash', False):
             self._hashstate = None
             self._realaddress = address
             return
         self._hashstate = hashstate.fromui(self.ui)
         self._realaddress = _hashaddress(address, self._hashstate.confighash)
 
     def _checkextensions(self):
         if not self._hashstate:
             return
         if extensions.notloaded():
             # one or more extensions failed to load. mtimehash becomes
             # meaningless because we do not know the paths of those extensions.
             # set mtimehash to an illegal hash value to invalidate the server.
             self._hashstate.mtimehash = ''
 
     def _bind(self, sock):
         # use a unique temp address so we can stat the file and do ownership
         # check later
         tempaddress = _tempaddress(self._realaddress)
         util.bindunixsocket(sock, tempaddress)
         self._socketstat = os.stat(tempaddress)
         # rename will replace the old socket file if exists atomically. the
         # old server will detect ownership change and exit.
         util.rename(tempaddress, self._realaddress)
 
     def _createsymlink(self):
         if self._baseaddress == self._realaddress:
             return
         tempaddress = _tempaddress(self._baseaddress)
         os.symlink(os.path.basename(self._realaddress), tempaddress)
         util.rename(tempaddress, self._baseaddress)
 
     def _issocketowner(self):
         try:
             stat = os.stat(self._realaddress)
             return (stat.st_ino == self._socketstat.st_ino and
                     stat.st_mtime == self._socketstat.st_mtime)
         except OSError:
             return False
 
     def unlinksocket(self, address):
         if not self._issocketowner():
             return
         # it is possible to have a race condition here that we may
         # remove another server's socket file. but that's okay
         # since that server will detect and exit automatically and
         # the client will start a new server on demand.
         try:
             os.unlink(self._realaddress)
         except OSError as exc:
             if exc.errno != errno.ENOENT:
                 raise
 
     def printbanner(self, address):
         # no "listening at" message should be printed to simulate hg behavior
         pass
 
     def shouldexit(self):
         if not self._issocketowner():
             self.ui.debug('%s is not owned, exiting.\n' % self._realaddress)
             return True
         if time.time() - self._lastactive > self._idletimeout:
             self.ui.debug('being idle too long. exiting.\n')
             return True
         return False
 
     def newconnection(self):
         self._lastactive = time.time()
 
     def createcmdserver(self, repo, conn, fin, fout):
         return chgcmdserver(self.ui, repo, fin, fout, conn,
                             self._hashstate, self._baseaddress)
 
 def chgunixservice(ui, repo, opts):
     # CHGINTERNALMARK is temporarily set by chg client to detect if chg will
     # start another chg. drop it to avoid possible side effects.
     if 'CHGINTERNALMARK' in encoding.environ:
         del encoding.environ['CHGINTERNALMARK']
 
     if repo:
         # one chgserver can serve multiple repos. drop repo information
         ui.setconfig('bundle', 'mainreporoot', '', 'repo')
     h = chgunixservicehandler(ui)
     return commandserver.unixforkingservice(ui, repo=None, opts=opts, handler=h)

mercurial/commandserver.py

@@ -1,551 +1,551 b''
 # commandserver.py - communicate with Mercurial's API over a pipe
 #
 #  Copyright Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import errno
 import gc
 import os
 import random
 import select
 import signal
 import socket
 import struct
 import traceback
 
 from .i18n import _
 from . import (
     encoding,
     error,
     pycompat,
     util,
 )
 
 logfile = None
 
 def log(*args):
     if not logfile:
         return
 
     for a in args:
         logfile.write(str(a))
 
     logfile.flush()
 
 class channeledoutput(object):
     """
     Write data to out in the following format:
 
     data length (unsigned int),
     data
     """
     def __init__(self, out, channel):
         self.out = out
         self.channel = channel
 
     @property
     def name(self):
         return '<%c-channel>' % self.channel
 
     def write(self, data):
         if not data:
             return
         # single write() to guarantee the same atomicity as the underlying file
         self.out.write(struct.pack('>cI', self.channel, len(data)) + data)
         self.out.flush()
 
     def __getattr__(self, attr):
         if attr in ('isatty', 'fileno', 'tell', 'seek'):
             raise AttributeError(attr)
         return getattr(self.out, attr)
 
 class channeledinput(object):
     """
     Read data from in_.
 
     Requests for input are written to out in the following format:
     channel identifier - 'I' for plain input, 'L' line based (1 byte)
     how many bytes to send at most (unsigned int),
 
     The client replies with:
     data length (unsigned int), 0 meaning EOF
     data
     """
 
     maxchunksize = 4 * 1024
 
     def __init__(self, in_, out, channel):
         self.in_ = in_
         self.out = out
         self.channel = channel
 
     @property
     def name(self):
         return '<%c-channel>' % self.channel
 
     def read(self, size=-1):
         if size < 0:
             # if we need to consume all the clients input, ask for 4k chunks
             # so the pipe doesn't fill up risking a deadlock
             size = self.maxchunksize
             s = self._read(size, self.channel)
             buf = s
             while s:
                 s = self._read(size, self.channel)
                 buf += s
 
             return buf
         else:
             return self._read(size, self.channel)
 
     def _read(self, size, channel):
         if not size:
             return ''
         assert size > 0
 
         # tell the client we need at most size bytes
         self.out.write(struct.pack('>cI', channel, size))
         self.out.flush()
 
         length = self.in_.read(4)
         length = struct.unpack('>I', length)[0]
         if not length:
             return ''
         else:
             return self.in_.read(length)
 
     def readline(self, size=-1):
         if size < 0:
             size = self.maxchunksize
             s = self._read(size, 'L')
             buf = s
             # keep asking for more until there's either no more or
             # we got a full line
             while s and s[-1] != '\n':
                 s = self._read(size, 'L')
                 buf += s
 
             return buf
         else:
             return self._read(size, 'L')
 
     def __iter__(self):
         return self
 
     def next(self):
         l = self.readline()
         if not l:
             raise StopIteration
         return l
 
     def __getattr__(self, attr):
         if attr in ('isatty', 'fileno', 'tell', 'seek'):
             raise AttributeError(attr)
         return getattr(self.in_, attr)
 
 class server(object):
     """
     Listens for commands on fin, runs them and writes the output on a channel
     based stream to fout.
     """
     def __init__(self, ui, repo, fin, fout):
         self.cwd = pycompat.getcwd()
 
         # developer config: cmdserver.log
         logpath = ui.config("cmdserver", "log", None)
         if logpath:
             global logfile
             if logpath == '-':
                 # write log on a special 'd' (debug) channel
                 logfile = channeledoutput(fout, 'd')
             else:
                 logfile = open(logpath, 'a')
 
         if repo:
             # the ui here is really the repo ui so take its baseui so we don't
             # end up with its local configuration
             self.ui = repo.baseui
             self.repo = repo
             self.repoui = repo.ui
         else:
             self.ui = ui
             self.repo = self.repoui = None
 
         self.cerr = channeledoutput(fout, 'e')
         self.cout = channeledoutput(fout, 'o')
         self.cin = channeledinput(fin, fout, 'I')
         self.cresult = channeledoutput(fout, 'r')
 
         self.client = fin
 
     def cleanup(self):
         """release and restore resources taken during server session"""
         pass
 
     def _read(self, size):
         if not size:
             return ''
 
         data = self.client.read(size)
 
         # is the other end closed?
         if not data:
             raise EOFError
 
         return data
 
     def _readstr(self):
         """read a string from the channel
 
         format:
         data length (uint32), data
         """
         length = struct.unpack('>I', self._read(4))[0]
         if not length:
             return ''
         return self._read(length)
 
     def _readlist(self):
         """read a list of NULL separated strings from the channel"""
         s = self._readstr()
         if s:
             return s.split('\0')
         else:
             return []
 
     def runcommand(self):
         """ reads a list of \0 terminated arguments, executes
         and writes the return code to the result channel """
         from . import dispatch  # avoid cycle
 
         args = self._readlist()
 
         # copy the uis so changes (e.g. --config or --verbose) don't
         # persist between requests
         copiedui = self.ui.copy()
         uis = [copiedui]
         if self.repo:
             self.repo.baseui = copiedui
             # clone ui without using ui.copy because this is protected
             repoui = self.repoui.__class__(self.repoui)
             repoui.copy = copiedui.copy # redo copy protection
             uis.append(repoui)
             self.repo.ui = self.repo.dirstate._ui = repoui
             self.repo.invalidateall()
 
         for ui in uis:
             ui.resetstate()
             # any kind of interaction must use server channels, but chg may
             # replace channels by fully functional tty files. so nontty is
             # enforced only if cin is a channel.
             if not util.safehasattr(self.cin, 'fileno'):
                 ui.setconfig('ui', 'nontty', 'true', 'commandserver')
 
         req = dispatch.request(args[:], copiedui, self.repo, self.cin,
                                self.cout, self.cerr)
 
         ret = (dispatch.dispatch(req) or 0) & 255 # might return None
 
         # restore old cwd
         if '--cwd' in args:
             os.chdir(self.cwd)
 
         self.cresult.write(struct.pack('>i', int(ret)))
 
     def getencoding(self):
         """ writes the current encoding to the result channel """
         self.cresult.write(encoding.encoding)
 
     def serveone(self):
         cmd = self.client.readline()[:-1]
         if cmd:
             handler = self.capabilities.get(cmd)
             if handler:
                 handler(self)
             else:
                 # clients are expected to check what commands are supported by
                 # looking at the servers capabilities
                 raise error.Abort(_('unknown command %s') % cmd)
 
         return cmd != ''
 
     capabilities = {'runcommand'  : runcommand,
                     'getencoding' : getencoding}
 
     def serve(self):
         hellomsg = 'capabilities: ' + ' '.join(sorted(self.capabilities))
         hellomsg += '\n'
         hellomsg += 'encoding: ' + encoding.encoding
         hellomsg += '\n'
         hellomsg += 'pid: %d' % util.getpid()
         if util.safehasattr(os, 'getpgid'):
             hellomsg += '\n'
             hellomsg += 'pgid: %d' % os.getpgid(0)
 
         # write the hello msg in -one- chunk
         self.cout.write(hellomsg)
 
         try:
             while self.serveone():
                 pass
         except EOFError:
             # we'll get here if the client disconnected while we were reading
             # its request
             return 1
 
         return 0
 
 def _protectio(ui):
     """ duplicates streams and redirect original to null if ui uses stdio """
     ui.flush()
     newfiles = []
     nullfd = os.open(os.devnull, os.O_RDWR)
-    for f, sysf, mode in [(ui.fin, util.stdin, 'rb'),
-                          (ui.fout, util.stdout, 'wb')]:
+    for f, sysf, mode in [(ui.fin, util.stdin, pycompat.sysstr('rb')),
+                          (ui.fout, util.stdout, pycompat.sysstr('wb'))]:
         if f is sysf:
             newfd = os.dup(f.fileno())
             os.dup2(nullfd, f.fileno())
             f = os.fdopen(newfd, mode)
         newfiles.append(f)
     os.close(nullfd)
     return tuple(newfiles)
 
 def _restoreio(ui, fin, fout):
     """ restores streams from duplicated ones """
     ui.flush()
     for f, uif in [(fin, ui.fin), (fout, ui.fout)]:
         if f is not uif:
             os.dup2(f.fileno(), uif.fileno())
             f.close()
 
 class pipeservice(object):
     def __init__(self, ui, repo, opts):
         self.ui = ui
         self.repo = repo
 
     def init(self):
         pass
 
     def run(self):
         ui = self.ui
         # redirect stdio to null device so that broken extensions or in-process
         # hooks will never cause corruption of channel protocol.
         fin, fout = _protectio(ui)
         try:
             sv = server(ui, self.repo, fin, fout)
             return sv.serve()
         finally:
             sv.cleanup()
             _restoreio(ui, fin, fout)
 
 def _initworkerprocess():
     # use a different process group from the master process, in order to:
     # 1. make the current process group no longer "orphaned" (because the
     #    parent of this process is in a different process group while
     #    remains in a same session)
     #    according to POSIX 2.2.2.52, orphaned process group will ignore
     #    terminal-generated stop signals like SIGTSTP (Ctrl+Z), which will
     #    cause trouble for things like ncurses.
     # 2. the client can use kill(-pgid, sig) to simulate terminal-generated
     #    SIGINT (Ctrl+C) and process-exit-generated SIGHUP. our child
     #    processes like ssh will be killed properly, without affecting
     #    unrelated processes.
     os.setpgid(0, 0)
     # change random state otherwise forked request handlers would have a
     # same state inherited from parent.
     random.seed()
 
 def _serverequest(ui, repo, conn, createcmdserver):
     fin = conn.makefile('rb')
     fout = conn.makefile('wb')
     sv = None
     try:
         sv = createcmdserver(repo, conn, fin, fout)
         try:
             sv.serve()
         # handle exceptions that may be raised by command server. most of
         # known exceptions are caught by dispatch.
         except error.Abort as inst:
             ui.warn(_('abort: %s\n') % inst)
         except IOError as inst:
             if inst.errno != errno.EPIPE:
                 raise
         except KeyboardInterrupt:
             pass
         finally:
             sv.cleanup()
     except: # re-raises
         # also write traceback to error channel. otherwise client cannot
         # see it because it is written to server's stderr by default.
         if sv:
             cerr = sv.cerr
         else:
             cerr = channeledoutput(fout, 'e')
         traceback.print_exc(file=cerr)
         raise
     finally:
         fin.close()
         try:
             fout.close()  # implicit flush() may cause another EPIPE
         except IOError as inst:
             if inst.errno != errno.EPIPE:
                 raise
 
 class unixservicehandler(object):
     """Set of pluggable operations for unix-mode services
 
     Almost all methods except for createcmdserver() are called in the main
     process. You can't pass mutable resource back from createcmdserver().
     """
 
     pollinterval = None
 
     def __init__(self, ui):
         self.ui = ui
 
     def bindsocket(self, sock, address):
         util.bindunixsocket(sock, address)
 
     def unlinksocket(self, address):
         os.unlink(address)
 
     def printbanner(self, address):
         self.ui.status(_('listening at %s\n') % address)
         self.ui.flush()  # avoid buffering of status message
 
     def shouldexit(self):
         """True if server should shut down; checked per pollinterval"""
         return False
 
     def newconnection(self):
         """Called when main process notices new connection"""
         pass
 
     def createcmdserver(self, repo, conn, fin, fout):
         """Create new command server instance; called in the process that
         serves for the current connection"""
         return server(self.ui, repo, fin, fout)
 
 class unixforkingservice(object):
     """
     Listens on unix domain socket and forks server per connection
     """
 
     def __init__(self, ui, repo, opts, handler=None):
         self.ui = ui
         self.repo = repo
         self.address = opts['address']
         if not util.safehasattr(socket, 'AF_UNIX'):
             raise error.Abort(_('unsupported platform'))
         if not self.address:
             raise error.Abort(_('no socket path specified with --address'))
         self._servicehandler = handler or unixservicehandler(ui)
         self._sock = None
         self._oldsigchldhandler = None
         self._workerpids = set()  # updated by signal handler; do not iterate
         self._socketunlinked = None
 
     def init(self):
         self._sock = socket.socket(socket.AF_UNIX)
         self._servicehandler.bindsocket(self._sock, self.address)
         self._sock.listen(socket.SOMAXCONN)
         o = signal.signal(signal.SIGCHLD, self._sigchldhandler)
         self._oldsigchldhandler = o
         self._servicehandler.printbanner(self.address)
         self._socketunlinked = False
 
     def _unlinksocket(self):
         if not self._socketunlinked:
             self._servicehandler.unlinksocket(self.address)
             self._socketunlinked = True
 
     def _cleanup(self):
         signal.signal(signal.SIGCHLD, self._oldsigchldhandler)
         self._sock.close()
         self._unlinksocket()
         # don't kill child processes as they have active clients, just wait
         self._reapworkers(0)
 
     def run(self):
         try:
             self._mainloop()
         finally:
             self._cleanup()
 
     def _mainloop(self):
         exiting = False
         h = self._servicehandler
         while True:
             if not exiting and h.shouldexit():
                 # clients can no longer connect() to the domain socket, so
                 # we stop queuing new requests.
                 # for requests that are queued (connect()-ed, but haven't been
                 # accept()-ed), handle them before exit. otherwise, clients
                 # waiting for recv() will receive ECONNRESET.
                 self._unlinksocket()
                 exiting = True
             try:
                 ready = select.select([self._sock], [], [], h.pollinterval)[0]
                 if not ready:
                     # only exit if we completed all queued requests
                     if exiting:
                         break
                     continue
                 conn, _addr = self._sock.accept()
             except (select.error, socket.error) as inst:
                 if inst.args[0] == errno.EINTR:
                     continue
                 raise
 
             pid = os.fork()
             if pid:
                 try:
                     self.ui.debug('forked worker process (pid=%d)\n' % pid)
                     self._workerpids.add(pid)
                     h.newconnection()
                 finally:
                     conn.close()  # release handle in parent process
             else:
                 try:
                     self._runworker(conn)
                     conn.close()
                     os._exit(0)
                 except:  # never return, hence no re-raises
                     try:
                         self.ui.traceback(force=True)
                     finally:
                         os._exit(255)
 
     def _sigchldhandler(self, signal, frame):
         self._reapworkers(os.WNOHANG)
 
     def _reapworkers(self, options):
         while self._workerpids:
             try:
                 pid, _status = os.waitpid(-1, options)
             except OSError as inst:
                 if inst.errno == errno.EINTR:
                     continue
                 if inst.errno != errno.ECHILD:
                     raise
                 # no child processes at all (reaped by other waitpid()?)
                 self._workerpids.clear()
                 return
             if pid == 0:
                 # no waitable child processes
                 return
             self.ui.debug('worker process exited (pid=%d)\n' % pid)
             self._workerpids.discard(pid)
 
     def _runworker(self, conn):
         signal.signal(signal.SIGCHLD, self._oldsigchldhandler)
         _initworkerprocess()
         h = self._servicehandler
         try:
             _serverequest(self.ui, self.repo, conn, h.createcmdserver)
         finally:
             gc.collect()  # trigger __del__ since worker process uses os._exit

mercurial/filemerge.py

@@ -1,718 +1,718 b''
 # filemerge.py - file-level merge handling for Mercurial
 #
 # Copyright 2006, 2007, 2008 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 filecmp
 import os
 import re
 import tempfile
 
 from .i18n import _
 from .node import nullid, short
 
 from . import (
     encoding,
     error,
     formatter,
     match,
     pycompat,
     scmutil,
     simplemerge,
     tagmerge,
     templatekw,
     templater,
     util,
 )
 
 def _toolstr(ui, tool, part, default=""):
     return ui.config("merge-tools", tool + "." + part, default)
 
 def _toolbool(ui, tool, part, default=False):
     return ui.configbool("merge-tools", tool + "." + part, default)
 
 def _toollist(ui, tool, part, default=[]):
     return ui.configlist("merge-tools", tool + "." + part, default)
 
 internals = {}
 # Merge tools to document.
 internalsdoc = {}
 
 # internal tool merge types
 nomerge = None
 mergeonly = 'mergeonly'  # just the full merge, no premerge
 fullmerge = 'fullmerge'  # both premerge and merge
 
 class absentfilectx(object):
     """Represents a file that's ostensibly in a context but is actually not
     present in it.
 
     This is here because it's very specific to the filemerge code for now --
     other code is likely going to break with the values this returns."""
     def __init__(self, ctx, f):
         self._ctx = ctx
         self._f = f
 
     def path(self):
         return self._f
 
     def size(self):
         return None
 
     def data(self):
         return None
 
     def filenode(self):
         return nullid
 
     _customcmp = True
     def cmp(self, fctx):
         """compare with other file context
 
         returns True if different from fctx.
         """
         return not (fctx.isabsent() and
                     fctx.ctx() == self.ctx() and
                     fctx.path() == self.path())
 
     def flags(self):
         return ''
 
     def changectx(self):
         return self._ctx
 
     def isbinary(self):
         return False
 
     def isabsent(self):
         return True
 
 def internaltool(name, mergetype, onfailure=None, precheck=None):
     '''return a decorator for populating internal merge tool table'''
     def decorator(func):
         fullname = ':' + name
         func.__doc__ = (pycompat.sysstr("``%s``\n" % fullname)
                         + func.__doc__.strip())
         internals[fullname] = func
         internals['internal:' + name] = func
         internalsdoc[fullname] = func
         func.mergetype = mergetype
         func.onfailure = onfailure
         func.precheck = precheck
         return func
     return decorator
 
 def _findtool(ui, tool):
     if tool in internals:
         return tool
     return findexternaltool(ui, tool)
 
 def findexternaltool(ui, tool):
     for kn in ("regkey", "regkeyalt"):
         k = _toolstr(ui, tool, kn)
         if not k:
             continue
         p = util.lookupreg(k, _toolstr(ui, tool, "regname"))
         if p:
             p = util.findexe(p + _toolstr(ui, tool, "regappend"))
             if p:
                 return p
     exe = _toolstr(ui, tool, "executable", tool)
     return util.findexe(util.expandpath(exe))
 
 def _picktool(repo, ui, path, binary, symlink, changedelete):
     def supportscd(tool):
         return tool in internals and internals[tool].mergetype == nomerge
 
     def check(tool, pat, symlink, binary, changedelete):
         tmsg = tool
         if pat:
             tmsg += " specified for " + pat
         if not _findtool(ui, tool):
             if pat: # explicitly requested tool deserves a warning
                 ui.warn(_("couldn't find merge tool %s\n") % tmsg)
             else: # configured but non-existing tools are more silent
                 ui.note(_("couldn't find merge tool %s\n") % tmsg)
         elif symlink and not _toolbool(ui, tool, "symlink"):
             ui.warn(_("tool %s can't handle symlinks\n") % tmsg)
         elif binary and not _toolbool(ui, tool, "binary"):
             ui.warn(_("tool %s can't handle binary\n") % tmsg)
         elif changedelete and not supportscd(tool):
             # the nomerge tools are the only tools that support change/delete
             # conflicts
             pass
         elif not util.gui() and _toolbool(ui, tool, "gui"):
             ui.warn(_("tool %s requires a GUI\n") % tmsg)
         else:
             return True
         return False
 
     # internal config: ui.forcemerge
     # forcemerge comes from command line arguments, highest priority
     force = ui.config('ui', 'forcemerge')
     if force:
         toolpath = _findtool(ui, force)
         if changedelete and not supportscd(toolpath):
             return ":prompt", None
         else:
             if toolpath:
                 return (force, util.shellquote(toolpath))
             else:
                 # mimic HGMERGE if given tool not found
                 return (force, force)
 
     # HGMERGE takes next precedence
     hgmerge = encoding.environ.get("HGMERGE")
     if hgmerge:
         if changedelete and not supportscd(hgmerge):
             return ":prompt", None
         else:
             return (hgmerge, hgmerge)
 
     # then patterns
     for pat, tool in ui.configitems("merge-patterns"):
         mf = match.match(repo.root, '', [pat])
         if mf(path) and check(tool, pat, symlink, False, changedelete):
             toolpath = _findtool(ui, tool)
             return (tool, util.shellquote(toolpath))
 
     # then merge tools
     tools = {}
     disabled = set()
     for k, v in ui.configitems("merge-tools"):
         t = k.split('.')[0]
         if t not in tools:
             tools[t] = int(_toolstr(ui, t, "priority", "0"))
         if _toolbool(ui, t, "disabled", False):
             disabled.add(t)
     names = tools.keys()
     tools = sorted([(-p, tool) for tool, p in tools.items()
                     if tool not in disabled])
     uimerge = ui.config("ui", "merge")
     if uimerge:
         # external tools defined in uimerge won't be able to handle
         # change/delete conflicts
         if uimerge not in names and not changedelete:
             return (uimerge, uimerge)
         tools.insert(0, (None, uimerge)) # highest priority
     tools.append((None, "hgmerge")) # the old default, if found
     for p, t in tools:
         if check(t, None, symlink, binary, changedelete):
             toolpath = _findtool(ui, t)
             return (t, util.shellquote(toolpath))
 
     # internal merge or prompt as last resort
     if symlink or binary or changedelete:
         return ":prompt", None
     return ":merge", None
 
 def _eoltype(data):
     "Guess the EOL type of a file"
     if '\0' in data: # binary
         return None
     if '\r\n' in data: # Windows
         return '\r\n'
     if '\r' in data: # Old Mac
         return '\r'
     if '\n' in data: # UNIX
         return '\n'
     return None # unknown
 
 def _matcheol(file, origfile):
     "Convert EOL markers in a file to match origfile"
     tostyle = _eoltype(util.readfile(origfile))
     if tostyle:
         data = util.readfile(file)
         style = _eoltype(data)
         if style:
             newdata = data.replace(style, tostyle)
             if newdata != data:
                 util.writefile(file, newdata)
 
 @internaltool('prompt', nomerge)
 def _iprompt(repo, mynode, orig, fcd, fco, fca, toolconf, labels=None):
     """Asks the user which of the local `p1()` or the other `p2()` version to
     keep as the merged version."""
     ui = repo.ui
     fd = fcd.path()
 
     prompts = partextras(labels)
     prompts['fd'] = fd
     try:
         if fco.isabsent():
             index = ui.promptchoice(
                 _("local%(l)s changed %(fd)s which other%(o)s deleted\n"
                   "use (c)hanged version, (d)elete, or leave (u)nresolved?"
                   "$$ &Changed $$ &Delete $$ &Unresolved") % prompts, 2)
             choice = ['local', 'other', 'unresolved'][index]
         elif fcd.isabsent():
             index = ui.promptchoice(
                 _("other%(o)s changed %(fd)s which local%(l)s deleted\n"
                   "use (c)hanged version, leave (d)eleted, or "
                   "leave (u)nresolved?"
                   "$$ &Changed $$ &Deleted $$ &Unresolved") % prompts, 2)
             choice = ['other', 'local', 'unresolved'][index]
         else:
             index = ui.promptchoice(
                 _("no tool found to merge %(fd)s\n"
                   "keep (l)ocal%(l)s, take (o)ther%(o)s, or leave (u)nresolved?"
                   "$$ &Local $$ &Other $$ &Unresolved") % prompts, 2)
             choice = ['local', 'other', 'unresolved'][index]
 
         if choice == 'other':
             return _iother(repo, mynode, orig, fcd, fco, fca, toolconf,
                            labels)
         elif choice == 'local':
             return _ilocal(repo, mynode, orig, fcd, fco, fca, toolconf,
                            labels)
         elif choice == 'unresolved':
             return _ifail(repo, mynode, orig, fcd, fco, fca, toolconf,
                           labels)
     except error.ResponseExpected:
         ui.write("\n")
         return _ifail(repo, mynode, orig, fcd, fco, fca, toolconf,
                       labels)
 
 @internaltool('local', nomerge)
 def _ilocal(repo, mynode, orig, fcd, fco, fca, toolconf, labels=None):
     """Uses the local `p1()` version of files as the merged version."""
     return 0, fcd.isabsent()
 
 @internaltool('other', nomerge)
 def _iother(repo, mynode, orig, fcd, fco, fca, toolconf, labels=None):
     """Uses the other `p2()` version of files as the merged version."""
     if fco.isabsent():
         # local changed, remote deleted -- 'deleted' picked
         repo.wvfs.unlinkpath(fcd.path())
         deleted = True
     else:
         repo.wwrite(fcd.path(), fco.data(), fco.flags())
         deleted = False
     return 0, deleted
 
 @internaltool('fail', nomerge)
 def _ifail(repo, mynode, orig, fcd, fco, fca, toolconf, labels=None):
     """
     Rather than attempting to merge files that were modified on both
     branches, it marks them as unresolved. The resolve command must be
     used to resolve these conflicts."""
     # for change/delete conflicts write out the changed version, then fail
     if fcd.isabsent():
         repo.wwrite(fcd.path(), fco.data(), fco.flags())
     return 1, False
 
 def _premerge(repo, fcd, fco, fca, toolconf, files, labels=None):
     tool, toolpath, binary, symlink = toolconf
     if symlink or fcd.isabsent() or fco.isabsent():
         return 1
     a, b, c, back = files
 
     ui = repo.ui
 
     validkeep = ['keep', 'keep-merge3']
 
     # do we attempt to simplemerge first?
     try:
         premerge = _toolbool(ui, tool, "premerge", not binary)
     except error.ConfigError:
         premerge = _toolstr(ui, tool, "premerge").lower()
         if premerge not in validkeep:
             _valid = ', '.join(["'" + v + "'" for v in validkeep])
             raise error.ConfigError(_("%s.premerge not valid "
                                       "('%s' is neither boolean nor %s)") %
                                     (tool, premerge, _valid))
 
     if premerge:
         if premerge == 'keep-merge3':
             if not labels:
                 labels = _defaultconflictlabels
             if len(labels) < 3:
                 labels.append('base')
         r = simplemerge.simplemerge(ui, a, b, c, quiet=True, label=labels)
         if not r:
             ui.debug(" premerge successful\n")
             return 0
         if premerge not in validkeep:
             util.copyfile(back, a) # restore from backup and try again
     return 1 # continue merging
 
 def _mergecheck(repo, mynode, orig, fcd, fco, fca, toolconf):
     tool, toolpath, binary, symlink = toolconf
     if symlink:
         repo.ui.warn(_('warning: internal %s cannot merge symlinks '
                        'for %s\n') % (tool, fcd.path()))
         return False
     if fcd.isabsent() or fco.isabsent():
         repo.ui.warn(_('warning: internal %s cannot merge change/delete '
                        'conflict for %s\n') % (tool, fcd.path()))
         return False
     return True
 
 def _merge(repo, mynode, orig, fcd, fco, fca, toolconf, files, labels, mode):
     """
     Uses the internal non-interactive simple merge algorithm for merging
     files. It will fail if there are any conflicts and leave markers in
     the partially merged file. Markers will have two sections, one for each side
     of merge, unless mode equals 'union' which suppresses the markers."""
     a, b, c, back = files
 
     ui = repo.ui
 
     r = simplemerge.simplemerge(ui, a, b, c, label=labels, mode=mode)
     return True, r, False
 
 @internaltool('union', fullmerge,
               _("warning: conflicts while merging %s! "
                 "(edit, then use 'hg resolve --mark')\n"),
               precheck=_mergecheck)
 def _iunion(repo, mynode, orig, fcd, fco, fca, toolconf, files, labels=None):
     """
     Uses the internal non-interactive simple merge algorithm for merging
     files. It will use both left and right sides for conflict regions.
     No markers are inserted."""
     return _merge(repo, mynode, orig, fcd, fco, fca, toolconf,
                   files, labels, 'union')
 
 @internaltool('merge', fullmerge,
               _("warning: conflicts while merging %s! "
                 "(edit, then use 'hg resolve --mark')\n"),
               precheck=_mergecheck)
 def _imerge(repo, mynode, orig, fcd, fco, fca, toolconf, files, labels=None):
     """
     Uses the internal non-interactive simple merge algorithm for merging
     files. It will fail if there are any conflicts and leave markers in
     the partially merged file. Markers will have two sections, one for each side
     of merge."""
     return _merge(repo, mynode, orig, fcd, fco, fca, toolconf,
                   files, labels, 'merge')
 
 @internaltool('merge3', fullmerge,
               _("warning: conflicts while merging %s! "
                 "(edit, then use 'hg resolve --mark')\n"),
               precheck=_mergecheck)
 def _imerge3(repo, mynode, orig, fcd, fco, fca, toolconf, files, labels=None):
     """
     Uses the internal non-interactive simple merge algorithm for merging
     files. It will fail if there are any conflicts and leave markers in
     the partially merged file. Marker will have three sections, one from each
     side of the merge and one for the base content."""
     if not labels:
         labels = _defaultconflictlabels
     if len(labels) < 3:
         labels.append('base')
     return _imerge(repo, mynode, orig, fcd, fco, fca, toolconf, files, labels)
 
 def _imergeauto(repo, mynode, orig, fcd, fco, fca, toolconf, files,
                 labels=None, localorother=None):
     """
     Generic driver for _imergelocal and _imergeother
     """
     assert localorother is not None
     tool, toolpath, binary, symlink = toolconf
     a, b, c, back = files
     r = simplemerge.simplemerge(repo.ui, a, b, c, label=labels,
                                 localorother=localorother)
     return True, r
 
 @internaltool('merge-local', mergeonly, precheck=_mergecheck)
 def _imergelocal(*args, **kwargs):
     """
     Like :merge, but resolve all conflicts non-interactively in favor
     of the local `p1()` changes."""
     success, status = _imergeauto(localorother='local', *args, **kwargs)
     return success, status, False
 
 @internaltool('merge-other', mergeonly, precheck=_mergecheck)
 def _imergeother(*args, **kwargs):
     """
     Like :merge, but resolve all conflicts non-interactively in favor
     of the other `p2()` changes."""
     success, status = _imergeauto(localorother='other', *args, **kwargs)
     return success, status, False
 
 @internaltool('tagmerge', mergeonly,
               _("automatic tag merging of %s failed! "
                 "(use 'hg resolve --tool :merge' or another merge "
                 "tool of your choice)\n"))
 def _itagmerge(repo, mynode, orig, fcd, fco, fca, toolconf, files, labels=None):
     """
     Uses the internal tag merge algorithm (experimental).
     """
     success, status = tagmerge.merge(repo, fcd, fco, fca)
     return success, status, False
 
 @internaltool('dump', fullmerge)
 def _idump(repo, mynode, orig, fcd, fco, fca, toolconf, files, labels=None):
     """
     Creates three versions of the files to merge, containing the
     contents of local, other and base. These files can then be used to
     perform a merge manually. If the file to be merged is named
     ``a.txt``, these files will accordingly be named ``a.txt.local``,
     ``a.txt.other`` and ``a.txt.base`` and they will be placed in the
     same directory as ``a.txt``."""
     a, b, c, back = files
 
     fd = fcd.path()
 
     util.copyfile(a, a + ".local")
     repo.wwrite(fd + ".other", fco.data(), fco.flags())
     repo.wwrite(fd + ".base", fca.data(), fca.flags())
     return False, 1, False
 
 def _xmerge(repo, mynode, orig, fcd, fco, fca, toolconf, files, labels=None):
     tool, toolpath, binary, symlink = toolconf
     if fcd.isabsent() or fco.isabsent():
         repo.ui.warn(_('warning: %s cannot merge change/delete conflict '
                        'for %s\n') % (tool, fcd.path()))
         return False, 1, None
     a, b, c, back = files
     out = ""
     env = {'HG_FILE': fcd.path(),
            'HG_MY_NODE': short(mynode),
            'HG_OTHER_NODE': str(fco.changectx()),
            'HG_BASE_NODE': str(fca.changectx()),
            'HG_MY_ISLINK': 'l' in fcd.flags(),
            'HG_OTHER_ISLINK': 'l' in fco.flags(),
            'HG_BASE_ISLINK': 'l' in fca.flags(),
            }
 
     ui = repo.ui
 
     args = _toolstr(ui, tool, "args", '$local $base $other')
     if "$output" in args:
         out, a = a, back # read input from backup, write to original
     replace = {'local': a, 'base': b, 'other': c, 'output': out}
     args = util.interpolate(r'\$', replace, args,
                             lambda s: util.shellquote(util.localpath(s)))
     cmd = toolpath + ' ' + args
     if _toolbool(ui, tool, "gui"):
         repo.ui.status(_('running merge tool %s for file %s\n') %
                        (tool, fcd.path()))
     repo.ui.debug('launching merge tool: %s\n' % cmd)
     r = ui.system(cmd, cwd=repo.root, environ=env)
     repo.ui.debug('merge tool returned: %s\n' % r)
     return True, r, False
 
 def _formatconflictmarker(repo, ctx, template, label, pad):
     """Applies the given template to the ctx, prefixed by the label.
 
     Pad is the minimum width of the label prefix, so that multiple markers
     can have aligned templated parts.
     """
     if ctx.node() is None:
         ctx = ctx.p1()
 
     props = templatekw.keywords.copy()
     props['templ'] = template
     props['ctx'] = ctx
     props['repo'] = repo
     templateresult = template('conflictmarker', **props)
 
     label = ('%s:' % label).ljust(pad + 1)
     mark = '%s %s' % (label, templater.stringify(templateresult))
 
     if mark:
         mark = mark.splitlines()[0] # split for safety
 
     # 8 for the prefix of conflict marker lines (e.g. '<<<<<<< ')
     return util.ellipsis(mark, 80 - 8)
 
 _defaultconflictmarker = ('{node|short} '
                           '{ifeq(tags, "tip", "", '
                            'ifeq(tags, "", "", "{tags} "))}'
                           '{if(bookmarks, "{bookmarks} ")}'
                           '{ifeq(branch, "default", "", "{branch} ")}'
                           '- {author|user}: {desc|firstline}')
 
 _defaultconflictlabels = ['local', 'other']
 
 def _formatlabels(repo, fcd, fco, fca, labels):
     """Formats the given labels using the conflict marker template.
 
     Returns a list of formatted labels.
     """
     cd = fcd.changectx()
     co = fco.changectx()
     ca = fca.changectx()
 
     ui = repo.ui
     template = ui.config('ui', 'mergemarkertemplate', _defaultconflictmarker)
     tmpl = formatter.maketemplater(ui, 'conflictmarker', template)
 
     pad = max(len(l) for l in labels)
 
     newlabels = [_formatconflictmarker(repo, cd, tmpl, labels[0], pad),
                  _formatconflictmarker(repo, co, tmpl, labels[1], pad)]
     if len(labels) > 2:
         newlabels.append(_formatconflictmarker(repo, ca, tmpl, labels[2], pad))
     return newlabels
 
 def partextras(labels):
     """Return a dictionary of extra labels for use in prompts to the user
 
     Intended use is in strings of the form "(l)ocal%(l)s".
     """
     if labels is None:
         return {
             "l": "",
             "o": "",
         }
 
     return {
         "l": " [%s]" % labels[0],
         "o": " [%s]" % labels[1],
     }
 
 def _filemerge(premerge, repo, mynode, orig, fcd, fco, fca, labels=None):
     """perform a 3-way merge in the working directory
 
     premerge = whether this is a premerge
     mynode = parent node before merge
     orig = original local filename before merge
     fco = other file context
     fca = ancestor file context
     fcd = local file context for current/destination file
 
     Returns whether the merge is complete, the return value of the merge, and
     a boolean indicating whether the file was deleted from disk."""
 
     def temp(prefix, ctx):
         fullbase, ext = os.path.splitext(ctx.path())
         pre = "%s~%s." % (os.path.basename(fullbase), prefix)
         (fd, name) = tempfile.mkstemp(prefix=pre, suffix=ext)
         data = repo.wwritedata(ctx.path(), ctx.data())
-        f = os.fdopen(fd, "wb")
+        f = os.fdopen(fd, pycompat.sysstr("wb"))
         f.write(data)
         f.close()
         return name
 
     if not fco.cmp(fcd): # files identical?
         return True, None, False
 
     ui = repo.ui
     fd = fcd.path()
     binary = fcd.isbinary() or fco.isbinary() or fca.isbinary()
     symlink = 'l' in fcd.flags() + fco.flags()
     changedelete = fcd.isabsent() or fco.isabsent()
     tool, toolpath = _picktool(repo, ui, fd, binary, symlink, changedelete)
     if tool in internals and tool.startswith('internal:'):
         # normalize to new-style names (':merge' etc)
         tool = tool[len('internal'):]
     ui.debug("picked tool '%s' for %s (binary %s symlink %s changedelete %s)\n"
              % (tool, fd, binary, symlink, changedelete))
 
     if tool in internals:
         func = internals[tool]
         mergetype = func.mergetype
         onfailure = func.onfailure
         precheck = func.precheck
     else:
         func = _xmerge
         mergetype = fullmerge
         onfailure = _("merging %s failed!\n")
         precheck = None
 
     toolconf = tool, toolpath, binary, symlink
 
     if mergetype == nomerge:
         r, deleted = func(repo, mynode, orig, fcd, fco, fca, toolconf, labels)
         return True, r, deleted
 
     if premerge:
         if orig != fco.path():
             ui.status(_("merging %s and %s to %s\n") % (orig, fco.path(), fd))
         else:
             ui.status(_("merging %s\n") % fd)
 
     ui.debug("my %s other %s ancestor %s\n" % (fcd, fco, fca))
 
     if precheck and not precheck(repo, mynode, orig, fcd, fco, fca,
                                  toolconf):
         if onfailure:
             ui.warn(onfailure % fd)
         return True, 1, False
 
     a = repo.wjoin(fd)
     b = temp("base", fca)
     c = temp("other", fco)
     if not fcd.isabsent():
         back = scmutil.origpath(ui, repo, a)
         if premerge:
             util.copyfile(a, back)
     else:
         back = None
     files = (a, b, c, back)
 
     r = 1
     try:
         markerstyle = ui.config('ui', 'mergemarkers', 'basic')
         if not labels:
             labels = _defaultconflictlabels
         if markerstyle != 'basic':
             labels = _formatlabels(repo, fcd, fco, fca, labels)
 
         if premerge and mergetype == fullmerge:
             r = _premerge(repo, fcd, fco, fca, toolconf, files, labels=labels)
             # complete if premerge successful (r is 0)
             return not r, r, False
 
         needcheck, r, deleted = func(repo, mynode, orig, fcd, fco, fca,
                                      toolconf, files, labels=labels)
 
         if needcheck:
             r = _check(r, ui, tool, fcd, files)
 
         if r:
             if onfailure:
                 ui.warn(onfailure % fd)
 
         return True, r, deleted
     finally:
         if not r and back is not None:
             util.unlink(back)
         util.unlink(b)
         util.unlink(c)
 
 def _check(r, ui, tool, fcd, files):
     fd = fcd.path()
     a, b, c, back = files
 
     if not r and (_toolbool(ui, tool, "checkconflicts") or
                   'conflicts' in _toollist(ui, tool, "check")):
         if re.search("^(<<<<<<< .*|=======|>>>>>>> .*)$", fcd.data(),
                      re.MULTILINE):
             r = 1
 
     checked = False
     if 'prompt' in _toollist(ui, tool, "check"):
         checked = True
         if ui.promptchoice(_("was merge of '%s' successful (yn)?"
                              "$$ &Yes $$ &No") % fd, 1):
             r = 1
 
     if not r and not checked and (_toolbool(ui, tool, "checkchanged") or
                                   'changed' in
                                   _toollist(ui, tool, "check")):
         if back is not None and filecmp.cmp(a, back):
             if ui.promptchoice(_(" output file %s appears unchanged\n"
                                  "was merge successful (yn)?"
                                  "$$ &Yes $$ &No") % fd, 1):
                 r = 1
 
     if back is not None and _toolbool(ui, tool, "fixeol"):
         _matcheol(a, back)
 
     return r
 
 def premerge(repo, mynode, orig, fcd, fco, fca, labels=None):
     return _filemerge(True, repo, mynode, orig, fcd, fco, fca, labels=labels)
 
 def filemerge(repo, mynode, orig, fcd, fco, fca, labels=None):
     return _filemerge(False, repo, mynode, orig, fcd, fco, fca, labels=labels)
 
 # tell hggettext to extract docstrings from these functions:
 i18nfunctions = internals.values()

mercurial/httppeer.py

@@ -1,382 +1,383 b''
 # httppeer.py - HTTP repository proxy classes for mercurial
 #
 # Copyright 2005, 2006 Matt Mackall <[e-mail]>
 # 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 errno
 import os
 import socket
 import struct
 import tempfile
 
 from .i18n import _
 from .node import nullid
 from . import (
     bundle2,
     error,
     httpconnection,
+    pycompat,
     statichttprepo,
     url,
     util,
     wireproto,
 )
 
 httplib = util.httplib
 urlerr = util.urlerr
 urlreq = util.urlreq
 
 # FUTURE: consider refactoring this API to use generators. This will
 # require a compression engine API to emit generators.
 def decompressresponse(response, engine):
     try:
         reader = engine.decompressorreader(response)
     except httplib.HTTPException:
         raise IOError(None, _('connection ended unexpectedly'))
 
     # We need to wrap reader.read() so HTTPException on subsequent
     # reads is also converted.
     # Ideally we'd use super() here. However, if ``reader`` isn't a new-style
     # class, this can raise:
     # TypeError: super() argument 1 must be type, not classobj
     origread = reader.read
     class readerproxy(reader.__class__):
         def read(self, *args, **kwargs):
             try:
                 return origread(*args, **kwargs)
             except httplib.HTTPException:
                 raise IOError(None, _('connection ended unexpectedly'))
 
     reader.__class__ = readerproxy
     return reader
 
 def encodevalueinheaders(value, header, limit):
     """Encode a string value into multiple HTTP headers.
 
     ``value`` will be encoded into 1 or more HTTP headers with the names
     ``header-<N>`` where ``<N>`` is an integer starting at 1. Each header
     name + value will be at most ``limit`` bytes long.
 
     Returns an iterable of 2-tuples consisting of header names and values.
     """
     fmt = header + '-%s'
     valuelen = limit - len(fmt % '000') - len(': \r\n')
     result = []
 
     n = 0
     for i in xrange(0, len(value), valuelen):
         n += 1
         result.append((fmt % str(n), value[i:i + valuelen]))
 
     return result
 
 class httppeer(wireproto.wirepeer):
     def __init__(self, ui, path):
         self.path = path
         self.caps = None
         self.handler = None
         self.urlopener = None
         self.requestbuilder = None
         u = util.url(path)
         if u.query or u.fragment:
             raise error.Abort(_('unsupported URL component: "%s"') %
                              (u.query or u.fragment))
 
         # urllib cannot handle URLs with embedded user or passwd
         self._url, authinfo = u.authinfo()
 
         self.ui = ui
         self.ui.debug('using %s\n' % self._url)
 
         self.urlopener = url.opener(ui, authinfo)
         self.requestbuilder = urlreq.request
 
     def __del__(self):
         urlopener = getattr(self, 'urlopener', None)
         if urlopener:
             for h in urlopener.handlers:
                 h.close()
                 getattr(h, "close_all", lambda : None)()
 
     def url(self):
         return self.path
 
     # look up capabilities only when needed
 
     def _fetchcaps(self):
         self.caps = set(self._call('capabilities').split())
 
     def _capabilities(self):
         if self.caps is None:
             try:
                 self._fetchcaps()
             except error.RepoError:
                 self.caps = set()
             self.ui.debug('capabilities: %s\n' %
                           (' '.join(self.caps or ['none'])))
         return self.caps
 
     def lock(self):
         raise error.Abort(_('operation not supported over http'))
 
     def _callstream(self, cmd, _compressible=False, **args):
         if cmd == 'pushkey':
             args['data'] = ''
         data = args.pop('data', None)
         headers = args.pop('headers', {})
 
         self.ui.debug("sending %s command\n" % cmd)
         q = [('cmd', cmd)]
         headersize = 0
         varyheaders = []
         # Important: don't use self.capable() here or else you end up
         # with infinite recursion when trying to look up capabilities
         # for the first time.
         postargsok = self.caps is not None and 'httppostargs' in self.caps
         # TODO: support for httppostargs when data is a file-like
         # object rather than a basestring
         canmungedata = not data or isinstance(data, basestring)
         if postargsok and canmungedata:
             strargs = urlreq.urlencode(sorted(args.items()))
             if strargs:
                 if not data:
                     data = strargs
                 elif isinstance(data, basestring):
                     data = strargs + data
                 headers['X-HgArgs-Post'] = len(strargs)
         else:
             if len(args) > 0:
                 httpheader = self.capable('httpheader')
                 if httpheader:
                     headersize = int(httpheader.split(',', 1)[0])
             if headersize > 0:
                 # The headers can typically carry more data than the URL.
                 encargs = urlreq.urlencode(sorted(args.items()))
                 for header, value in encodevalueinheaders(encargs, 'X-HgArg',
                                                           headersize):
                     headers[header] = value
                     varyheaders.append(header)
             else:
                 q += sorted(args.items())
         qs = '?%s' % urlreq.urlencode(q)
         cu = "%s%s" % (self._url, qs)
         size = 0
         if util.safehasattr(data, 'length'):
             size = data.length
         elif data is not None:
             size = len(data)
         if size and self.ui.configbool('ui', 'usehttp2', False):
             headers['Expect'] = '100-Continue'
             headers['X-HgHttp2'] = '1'
         if data is not None and 'Content-Type' not in headers:
             headers['Content-Type'] = 'application/mercurial-0.1'
 
         # Tell the server we accept application/mercurial-0.2 and multiple
         # compression formats if the server is capable of emitting those
         # payloads.
         protoparams = []
 
         mediatypes = set()
         if self.caps is not None:
             mt = self.capable('httpmediatype')
             if mt:
                 protoparams.append('0.1')
                 mediatypes = set(mt.split(','))
 
         if '0.2tx' in mediatypes:
             protoparams.append('0.2')
 
         if '0.2tx' in mediatypes and self.capable('compression'):
             # We /could/ compare supported compression formats and prune
             # non-mutually supported or error if nothing is mutually supported.
             # For now, send the full list to the server and have it error.
             comps = [e.wireprotosupport().name for e in
                      util.compengines.supportedwireengines(util.CLIENTROLE)]
             protoparams.append('comp=%s' % ','.join(comps))
 
         if protoparams:
             protoheaders = encodevalueinheaders(' '.join(protoparams),
                                                 'X-HgProto',
                                                 headersize or 1024)
             for header, value in protoheaders:
                 headers[header] = value
                 varyheaders.append(header)
 
         headers['Vary'] = ','.join(varyheaders)
         req = self.requestbuilder(cu, data, headers)
 
         if data is not None:
             self.ui.debug("sending %s bytes\n" % size)
             req.add_unredirected_header('Content-Length', '%d' % size)
         try:
             resp = self.urlopener.open(req)
         except urlerr.httperror as inst:
             if inst.code == 401:
                 raise error.Abort(_('authorization failed'))
             raise
         except httplib.HTTPException as inst:
             self.ui.debug('http error while sending %s command\n' % cmd)
             self.ui.traceback()
             raise IOError(None, inst)
         # record the url we got redirected to
         resp_url = resp.geturl()
         if resp_url.endswith(qs):
             resp_url = resp_url[:-len(qs)]
         if self._url.rstrip('/') != resp_url.rstrip('/'):
             if not self.ui.quiet:
                 self.ui.warn(_('real URL is %s\n') % resp_url)
         self._url = resp_url
         try:
             proto = resp.getheader('content-type')
         except AttributeError:
             proto = resp.headers.get('content-type', '')
 
         safeurl = util.hidepassword(self._url)
         if proto.startswith('application/hg-error'):
             raise error.OutOfBandError(resp.read())
         # accept old "text/plain" and "application/hg-changegroup" for now
         if not (proto.startswith('application/mercurial-') or
                 (proto.startswith('text/plain')
                  and not resp.headers.get('content-length')) or
                 proto.startswith('application/hg-changegroup')):
             self.ui.debug("requested URL: '%s'\n" % util.hidepassword(cu))
             raise error.RepoError(
                 _("'%s' does not appear to be an hg repository:\n"
                   "---%%<--- (%s)\n%s\n---%%<---\n")
                 % (safeurl, proto or 'no content-type', resp.read(1024)))
 
         if proto.startswith('application/mercurial-'):
             try:
                 version = proto.split('-', 1)[1]
                 version_info = tuple([int(n) for n in version.split('.')])
             except ValueError:
                 raise error.RepoError(_("'%s' sent a broken Content-Type "
                                         "header (%s)") % (safeurl, proto))
 
             if version_info == (0, 1):
                 if _compressible:
                     return decompressresponse(resp, util.compengines['zlib'])
                 return resp
             elif version_info == (0, 2):
                 # application/mercurial-0.2 always identifies the compression
                 # engine in the payload header.
                 elen = struct.unpack('B', resp.read(1))[0]
                 ename = resp.read(elen)
                 engine = util.compengines.forwiretype(ename)
                 return decompressresponse(resp, engine)
             else:
                 raise error.RepoError(_("'%s' uses newer protocol %s") %
                                       (safeurl, version))
 
         if _compressible:
             return decompressresponse(resp, util.compengines['zlib'])
 
         return resp
 
     def _call(self, cmd, **args):
         fp = self._callstream(cmd, **args)
         try:
             return fp.read()
         finally:
             # if using keepalive, allow connection to be reused
             fp.close()
 
     def _callpush(self, cmd, cg, **args):
         # have to stream bundle to a temp file because we do not have
         # http 1.1 chunked transfer.
 
         types = self.capable('unbundle')
         try:
             types = types.split(',')
         except AttributeError:
             # servers older than d1b16a746db6 will send 'unbundle' as a
             # boolean capability. They only support headerless/uncompressed
             # bundles.
             types = [""]
         for x in types:
             if x in bundle2.bundletypes:
                 type = x
                 break
 
         tempname = bundle2.writebundle(self.ui, cg, None, type)
         fp = httpconnection.httpsendfile(self.ui, tempname, "rb")
         headers = {'Content-Type': 'application/mercurial-0.1'}
 
         try:
             r = self._call(cmd, data=fp, headers=headers, **args)
             vals = r.split('\n', 1)
             if len(vals) < 2:
                 raise error.ResponseError(_("unexpected response:"), r)
             return vals
         except socket.error as err:
             if err.args[0] in (errno.ECONNRESET, errno.EPIPE):
                 raise error.Abort(_('push failed: %s') % err.args[1])
             raise error.Abort(err.args[1])
         finally:
             fp.close()
             os.unlink(tempname)
 
     def _calltwowaystream(self, cmd, fp, **args):
         fh = None
         fp_ = None
         filename = None
         try:
             # dump bundle to disk
             fd, filename = tempfile.mkstemp(prefix="hg-bundle-", suffix=".hg")
-            fh = os.fdopen(fd, "wb")
+            fh = os.fdopen(fd, pycompat.sysstr("wb"))
             d = fp.read(4096)
             while d:
                 fh.write(d)
                 d = fp.read(4096)
             fh.close()
             # start http push
             fp_ = httpconnection.httpsendfile(self.ui, filename, "rb")
             headers = {'Content-Type': 'application/mercurial-0.1'}
             return self._callstream(cmd, data=fp_, headers=headers, **args)
         finally:
             if fp_ is not None:
                 fp_.close()
             if fh is not None:
                 fh.close()
                 os.unlink(filename)
 
     def _callcompressable(self, cmd, **args):
         return self._callstream(cmd, _compressible=True, **args)
 
     def _abort(self, exception):
         raise exception
 
 class httpspeer(httppeer):
     def __init__(self, ui, path):
         if not url.has_https:
             raise error.Abort(_('Python support for SSL and HTTPS '
                                'is not installed'))
         httppeer.__init__(self, ui, path)
 
 def instance(ui, path, create):
     if create:
         raise error.Abort(_('cannot create new http repository'))
     try:
         if path.startswith('https:'):
             inst = httpspeer(ui, path)
         else:
             inst = httppeer(ui, path)
         try:
             # Try to do useful work when checking compatibility.
             # Usually saves a roundtrip since we want the caps anyway.
             inst._fetchcaps()
         except error.RepoError:
             # No luck, try older compatibility check.
             inst.between([(nullid, nullid)])
         return inst
     except error.RepoError as httpexception:
         try:
             r = statichttprepo.instance(ui, "static-" + path, create)
             ui.note(_('(falling back to static-http)\n'))
             return r
         except error.RepoError:
             raise httpexception # use the original http RepoError instead

mercurial/patch.py

@@ -1,2653 +1,2654 b''
 # patch.py - patch file parsing routines
 #
 # Copyright 2006 Brendan Cully <[e-mail]>
 # Copyright 2007 Chris Mason <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import collections
 import copy
 import email
 import errno
 import hashlib
 import os
 import posixpath
 import re
 import shutil
 import tempfile
 import zlib
 
 from .i18n import _
 from .node import (
     hex,
     short,
 )
 from . import (
     base85,
     copies,
     diffhelpers,
     encoding,
     error,
     mail,
     mdiff,
     pathutil,
+    pycompat,
     scmutil,
     similar,
     util,
 )
 stringio = util.stringio
 
 gitre = re.compile('diff --git a/(.*) b/(.*)')
 tabsplitter = re.compile(r'(\t+|[^\t]+)')
 
 class PatchError(Exception):
     pass
 
 
 # public functions
 
 def split(stream):
     '''return an iterator of individual patches from a stream'''
     def isheader(line, inheader):
         if inheader and line[0] in (' ', '\t'):
             # continuation
             return True
         if line[0] in (' ', '-', '+'):
             # diff line - don't check for header pattern in there
             return False
         l = line.split(': ', 1)
         return len(l) == 2 and ' ' not in l[0]
 
     def chunk(lines):
         return stringio(''.join(lines))
 
     def hgsplit(stream, cur):
         inheader = True
 
         for line in stream:
             if not line.strip():
                 inheader = False
             if not inheader and line.startswith('# HG changeset patch'):
                 yield chunk(cur)
                 cur = []
                 inheader = True
 
             cur.append(line)
 
         if cur:
             yield chunk(cur)
 
     def mboxsplit(stream, cur):
         for line in stream:
             if line.startswith('From '):
                 for c in split(chunk(cur[1:])):
                     yield c
                 cur = []
 
             cur.append(line)
 
         if cur:
             for c in split(chunk(cur[1:])):
                 yield c
 
     def mimesplit(stream, cur):
         def msgfp(m):
             fp = stringio()
             g = email.Generator.Generator(fp, mangle_from_=False)
             g.flatten(m)
             fp.seek(0)
             return fp
 
         for line in stream:
             cur.append(line)
         c = chunk(cur)
 
         m = email.Parser.Parser().parse(c)
         if not m.is_multipart():
             yield msgfp(m)
         else:
             ok_types = ('text/plain', 'text/x-diff', 'text/x-patch')
             for part in m.walk():
                 ct = part.get_content_type()
                 if ct not in ok_types:
                     continue
                 yield msgfp(part)
 
     def headersplit(stream, cur):
         inheader = False
 
         for line in stream:
             if not inheader and isheader(line, inheader):
                 yield chunk(cur)
                 cur = []
                 inheader = True
             if inheader and not isheader(line, inheader):
                 inheader = False
 
             cur.append(line)
 
         if cur:
             yield chunk(cur)
 
     def remainder(cur):
         yield chunk(cur)
 
     class fiter(object):
         def __init__(self, fp):
             self.fp = fp
 
         def __iter__(self):
             return self
 
         def next(self):
             l = self.fp.readline()
             if not l:
                 raise StopIteration
             return l
 
     inheader = False
     cur = []
 
     mimeheaders = ['content-type']
 
     if not util.safehasattr(stream, 'next'):
         # http responses, for example, have readline but not next
         stream = fiter(stream)
 
     for line in stream:
         cur.append(line)
         if line.startswith('# HG changeset patch'):
             return hgsplit(stream, cur)
         elif line.startswith('From '):
             return mboxsplit(stream, cur)
         elif isheader(line, inheader):
             inheader = True
             if line.split(':', 1)[0].lower() in mimeheaders:
                 # let email parser handle this
                 return mimesplit(stream, cur)
         elif line.startswith('--- ') and inheader:
             # No evil headers seen by diff start, split by hand
             return headersplit(stream, cur)
         # Not enough info, keep reading
 
     # if we are here, we have a very plain patch
     return remainder(cur)
 
 ## Some facility for extensible patch parsing:
 # list of pairs ("header to match", "data key")
 patchheadermap = [('Date', 'date'),
                   ('Branch', 'branch'),
                   ('Node ID', 'nodeid'),
                  ]
 
 def extract(ui, fileobj):
     '''extract patch from data read from fileobj.
 
     patch can be a normal patch or contained in an email message.
 
     return a dictionary. Standard keys are:
       - filename,
       - message,
       - user,
       - date,
       - branch,
       - node,
       - p1,
       - p2.
     Any item can be missing from the dictionary. If filename is missing,
     fileobj did not contain a patch. Caller must unlink filename when done.'''
 
     # attempt to detect the start of a patch
     # (this heuristic is borrowed from quilt)
     diffre = re.compile(r'^(?:Index:[ \t]|diff[ \t]|RCS file: |'
                         r'retrieving revision [0-9]+(\.[0-9]+)*$|'
                         r'---[ \t].*?^\+\+\+[ \t]|'
                         r'\*\*\*[ \t].*?^---[ \t])', re.MULTILINE|re.DOTALL)
 
     data = {}
     fd, tmpname = tempfile.mkstemp(prefix='hg-patch-')
-    tmpfp = os.fdopen(fd, 'w')
+    tmpfp = os.fdopen(fd, pycompat.sysstr('w'))
     try:
         msg = email.Parser.Parser().parse(fileobj)
 
         subject = msg['Subject'] and mail.headdecode(msg['Subject'])
         data['user'] = msg['From'] and mail.headdecode(msg['From'])
         if not subject and not data['user']:
             # Not an email, restore parsed headers if any
             subject = '\n'.join(': '.join(h) for h in msg.items()) + '\n'
 
         # should try to parse msg['Date']
         parents = []
 
         if subject:
             if subject.startswith('[PATCH'):
                 pend = subject.find(']')
                 if pend >= 0:
                     subject = subject[pend + 1:].lstrip()
             subject = re.sub(r'\n[ \t]+', ' ', subject)
             ui.debug('Subject: %s\n' % subject)
         if data['user']:
             ui.debug('From: %s\n' % data['user'])
         diffs_seen = 0
         ok_types = ('text/plain', 'text/x-diff', 'text/x-patch')
         message = ''
         for part in msg.walk():
             content_type = part.get_content_type()
             ui.debug('Content-Type: %s\n' % content_type)
             if content_type not in ok_types:
                 continue
             payload = part.get_payload(decode=True)
             m = diffre.search(payload)
             if m:
                 hgpatch = False
                 hgpatchheader = False
                 ignoretext = False
 
                 ui.debug('found patch at byte %d\n' % m.start(0))
                 diffs_seen += 1
                 cfp = stringio()
                 for line in payload[:m.start(0)].splitlines():
                     if line.startswith('# HG changeset patch') and not hgpatch:
                         ui.debug('patch generated by hg export\n')
                         hgpatch = True
                         hgpatchheader = True
                         # drop earlier commit message content
                         cfp.seek(0)
                         cfp.truncate()
                         subject = None
                     elif hgpatchheader:
                         if line.startswith('# User '):
                             data['user'] = line[7:]
                             ui.debug('From: %s\n' % data['user'])
                         elif line.startswith("# Parent "):
                             parents.append(line[9:].lstrip())
                         elif line.startswith("# "):
                             for header, key in patchheadermap:
                                 prefix = '# %s ' % header
                                 if line.startswith(prefix):
                                     data[key] = line[len(prefix):]
                         else:
                             hgpatchheader = False
                     elif line == '---':
                         ignoretext = True
                     if not hgpatchheader and not ignoretext:
                         cfp.write(line)
                         cfp.write('\n')
                 message = cfp.getvalue()
                 if tmpfp:
                     tmpfp.write(payload)
                     if not payload.endswith('\n'):
                         tmpfp.write('\n')
             elif not diffs_seen and message and content_type == 'text/plain':
                 message += '\n' + payload
     except: # re-raises
         tmpfp.close()
         os.unlink(tmpname)
         raise
 
     if subject and not message.startswith(subject):
         message = '%s\n%s' % (subject, message)
     data['message'] = message
     tmpfp.close()
     if parents:
         data['p1'] = parents.pop(0)
         if parents:
             data['p2'] = parents.pop(0)
 
     if diffs_seen:
         data['filename'] = tmpname
     else:
         os.unlink(tmpname)
     return data
 
 class patchmeta(object):
     """Patched file metadata
 
     'op' is the performed operation within ADD, DELETE, RENAME, MODIFY
     or COPY.  'path' is patched file path. 'oldpath' is set to the
     origin file when 'op' is either COPY or RENAME, None otherwise. If
     file mode is changed, 'mode' is a tuple (islink, isexec) where
     'islink' is True if the file is a symlink and 'isexec' is True if
     the file is executable. Otherwise, 'mode' is None.
     """
     def __init__(self, path):
         self.path = path
         self.oldpath = None
         self.mode = None
         self.op = 'MODIFY'
         self.binary = False
 
     def setmode(self, mode):
         islink = mode & 0o20000
         isexec = mode & 0o100
         self.mode = (islink, isexec)
 
     def copy(self):
         other = patchmeta(self.path)
         other.oldpath = self.oldpath
         other.mode = self.mode
         other.op = self.op
         other.binary = self.binary
         return other
 
     def _ispatchinga(self, afile):
         if afile == '/dev/null':
             return self.op == 'ADD'
         return afile == 'a/' + (self.oldpath or self.path)
 
     def _ispatchingb(self, bfile):
         if bfile == '/dev/null':
             return self.op == 'DELETE'
         return bfile == 'b/' + self.path
 
     def ispatching(self, afile, bfile):
         return self._ispatchinga(afile) and self._ispatchingb(bfile)
 
     def __repr__(self):
         return "<patchmeta %s %r>" % (self.op, self.path)
 
 def readgitpatch(lr):
     """extract git-style metadata about patches from <patchname>"""
 
     # Filter patch for git information
     gp = None
     gitpatches = []
     for line in lr:
         line = line.rstrip(' \r\n')
         if line.startswith('diff --git a/'):
             m = gitre.match(line)
             if m:
                 if gp:
                     gitpatches.append(gp)
                 dst = m.group(2)
                 gp = patchmeta(dst)
         elif gp:
             if line.startswith('--- '):
                 gitpatches.append(gp)
                 gp = None
                 continue
             if line.startswith('rename from '):
                 gp.op = 'RENAME'
                 gp.oldpath = line[12:]
             elif line.startswith('rename to '):
                 gp.path = line[10:]
             elif line.startswith('copy from '):
                 gp.op = 'COPY'
                 gp.oldpath = line[10:]
             elif line.startswith('copy to '):
                 gp.path = line[8:]
             elif line.startswith('deleted file'):
                 gp.op = 'DELETE'
             elif line.startswith('new file mode '):
                 gp.op = 'ADD'
                 gp.setmode(int(line[-6:], 8))
             elif line.startswith('new mode '):
                 gp.setmode(int(line[-6:], 8))
             elif line.startswith('GIT binary patch'):
                 gp.binary = True
     if gp:
         gitpatches.append(gp)
 
     return gitpatches
 
 class linereader(object):
     # simple class to allow pushing lines back into the input stream
     def __init__(self, fp):
         self.fp = fp
         self.buf = []
 
     def push(self, line):
         if line is not None:
             self.buf.append(line)
 
     def readline(self):
         if self.buf:
             l = self.buf[0]
             del self.buf[0]
             return l
         return self.fp.readline()
 
     def __iter__(self):
         return iter(self.readline, '')
 
 class abstractbackend(object):
     def __init__(self, ui):
         self.ui = ui
 
     def getfile(self, fname):
         """Return target file data and flags as a (data, (islink,
         isexec)) tuple. Data is None if file is missing/deleted.
         """
         raise NotImplementedError
 
     def setfile(self, fname, data, mode, copysource):
         """Write data to target file fname and set its mode. mode is a
         (islink, isexec) tuple. If data is None, the file content should
         be left unchanged. If the file is modified after being copied,
         copysource is set to the original file name.
         """
         raise NotImplementedError
 
     def unlink(self, fname):
         """Unlink target file."""
         raise NotImplementedError
 
     def writerej(self, fname, failed, total, lines):
         """Write rejected lines for fname. total is the number of hunks
         which failed to apply and total the total number of hunks for this
         files.
         """
         pass
 
     def exists(self, fname):
         raise NotImplementedError
 
 class fsbackend(abstractbackend):
     def __init__(self, ui, basedir):
         super(fsbackend, self).__init__(ui)
         self.opener = scmutil.opener(basedir)
 
     def _join(self, f):
         return os.path.join(self.opener.base, f)
 
     def getfile(self, fname):
         if self.opener.islink(fname):
             return (self.opener.readlink(fname), (True, False))
 
         isexec = False
         try:
             isexec = self.opener.lstat(fname).st_mode & 0o100 != 0
         except OSError as e:
             if e.errno != errno.ENOENT:
                 raise
         try:
             return (self.opener.read(fname), (False, isexec))
         except IOError as e:
             if e.errno != errno.ENOENT:
                 raise
             return None, None
 
     def setfile(self, fname, data, mode, copysource):
         islink, isexec = mode
         if data is None:
             self.opener.setflags(fname, islink, isexec)
             return
         if islink:
             self.opener.symlink(data, fname)
         else:
             self.opener.write(fname, data)
             if isexec:
                 self.opener.setflags(fname, False, True)
 
     def unlink(self, fname):
         self.opener.unlinkpath(fname, ignoremissing=True)
 
     def writerej(self, fname, failed, total, lines):
         fname = fname + ".rej"
         self.ui.warn(
             _("%d out of %d hunks FAILED -- saving rejects to file %s\n") %
             (failed, total, fname))
         fp = self.opener(fname, 'w')
         fp.writelines(lines)
         fp.close()
 
     def exists(self, fname):
         return self.opener.lexists(fname)
 
 class workingbackend(fsbackend):
     def __init__(self, ui, repo, similarity):
         super(workingbackend, self).__init__(ui, repo.root)
         self.repo = repo
         self.similarity = similarity
         self.removed = set()
         self.changed = set()
         self.copied = []
 
     def _checkknown(self, fname):
         if self.repo.dirstate[fname] == '?' and self.exists(fname):
             raise PatchError(_('cannot patch %s: file is not tracked') % fname)
 
     def setfile(self, fname, data, mode, copysource):
         self._checkknown(fname)
         super(workingbackend, self).setfile(fname, data, mode, copysource)
         if copysource is not None:
             self.copied.append((copysource, fname))
         self.changed.add(fname)
 
     def unlink(self, fname):
         self._checkknown(fname)
         super(workingbackend, self).unlink(fname)
         self.removed.add(fname)
         self.changed.add(fname)
 
     def close(self):
         wctx = self.repo[None]
         changed = set(self.changed)
         for src, dst in self.copied:
             scmutil.dirstatecopy(self.ui, self.repo, wctx, src, dst)
         if self.removed:
             wctx.forget(sorted(self.removed))
             for f in self.removed:
                 if f not in self.repo.dirstate:
                     # File was deleted and no longer belongs to the
                     # dirstate, it was probably marked added then
                     # deleted, and should not be considered by
                     # marktouched().
                     changed.discard(f)
         if changed:
             scmutil.marktouched(self.repo, changed, self.similarity)
         return sorted(self.changed)
 
 class filestore(object):
     def __init__(self, maxsize=None):
         self.opener = None
         self.files = {}
         self.created = 0
         self.maxsize = maxsize
         if self.maxsize is None:
             self.maxsize = 4*(2**20)
         self.size = 0
         self.data = {}
 
     def setfile(self, fname, data, mode, copied=None):
         if self.maxsize < 0 or (len(data) + self.size) <= self.maxsize:
             self.data[fname] = (data, mode, copied)
             self.size += len(data)
         else:
             if self.opener is None:
                 root = tempfile.mkdtemp(prefix='hg-patch-')
                 self.opener = scmutil.opener(root)
             # Avoid filename issues with these simple names
             fn = str(self.created)
             self.opener.write(fn, data)
             self.created += 1
             self.files[fname] = (fn, mode, copied)
 
     def getfile(self, fname):
         if fname in self.data:
             return self.data[fname]
         if not self.opener or fname not in self.files:
             return None, None, None
         fn, mode, copied = self.files[fname]
         return self.opener.read(fn), mode, copied
 
     def close(self):
         if self.opener:
             shutil.rmtree(self.opener.base)
 
 class repobackend(abstractbackend):
     def __init__(self, ui, repo, ctx, store):
         super(repobackend, self).__init__(ui)
         self.repo = repo
         self.ctx = ctx
         self.store = store
         self.changed = set()
         self.removed = set()
         self.copied = {}
 
     def _checkknown(self, fname):
         if fname not in self.ctx:
             raise PatchError(_('cannot patch %s: file is not tracked') % fname)
 
     def getfile(self, fname):
         try:
             fctx = self.ctx[fname]
         except error.LookupError:
             return None, None
         flags = fctx.flags()
         return fctx.data(), ('l' in flags, 'x' in flags)
 
     def setfile(self, fname, data, mode, copysource):
         if copysource:
             self._checkknown(copysource)
         if data is None:
             data = self.ctx[fname].data()
         self.store.setfile(fname, data, mode, copysource)
         self.changed.add(fname)
         if copysource:
             self.copied[fname] = copysource
 
     def unlink(self, fname):
         self._checkknown(fname)
         self.removed.add(fname)
 
     def exists(self, fname):
         return fname in self.ctx
 
     def close(self):
         return self.changed | self.removed
 
 # @@ -start,len +start,len @@ or @@ -start +start @@ if len is 1
 unidesc = re.compile('@@ -(\d+)(?:,(\d+))? \+(\d+)(?:,(\d+))? @@')
 contextdesc = re.compile('(?:---|\*\*\*) (\d+)(?:,(\d+))? (?:---|\*\*\*)')
 eolmodes = ['strict', 'crlf', 'lf', 'auto']
 
 class patchfile(object):
     def __init__(self, ui, gp, backend, store, eolmode='strict'):
         self.fname = gp.path
         self.eolmode = eolmode
         self.eol = None
         self.backend = backend
         self.ui = ui
         self.lines = []
         self.exists = False
         self.missing = True
         self.mode = gp.mode
         self.copysource = gp.oldpath
         self.create = gp.op in ('ADD', 'COPY', 'RENAME')
         self.remove = gp.op == 'DELETE'
         if self.copysource is None:
             data, mode = backend.getfile(self.fname)
         else:
             data, mode = store.getfile(self.copysource)[:2]
         if data is not None:
             self.exists = self.copysource is None or backend.exists(self.fname)
             self.missing = False
             if data:
                 self.lines = mdiff.splitnewlines(data)
             if self.mode is None:
                 self.mode = mode
             if self.lines:
                 # Normalize line endings
                 if self.lines[0].endswith('\r\n'):
                     self.eol = '\r\n'
                 elif self.lines[0].endswith('\n'):
                     self.eol = '\n'
                 if eolmode != 'strict':
                     nlines = []
                     for l in self.lines:
                         if l.endswith('\r\n'):
                             l = l[:-2] + '\n'
                         nlines.append(l)
                     self.lines = nlines
         else:
             if self.create:
                 self.missing = False
             if self.mode is None:
                 self.mode = (False, False)
         if self.missing:
             self.ui.warn(_("unable to find '%s' for patching\n") % self.fname)
             self.ui.warn(_("(use '--prefix' to apply patch relative to the "
                            "current directory)\n"))
 
         self.hash = {}
         self.dirty = 0
         self.offset = 0
         self.skew = 0
         self.rej = []
         self.fileprinted = False
         self.printfile(False)
         self.hunks = 0
 
     def writelines(self, fname, lines, mode):
         if self.eolmode == 'auto':
             eol = self.eol
         elif self.eolmode == 'crlf':
             eol = '\r\n'
         else:
             eol = '\n'
 
         if self.eolmode != 'strict' and eol and eol != '\n':
             rawlines = []
             for l in lines:
                 if l and l[-1] == '\n':
                     l = l[:-1] + eol
                 rawlines.append(l)
             lines = rawlines
 
         self.backend.setfile(fname, ''.join(lines), mode, self.copysource)
 
     def printfile(self, warn):
         if self.fileprinted:
             return
         if warn or self.ui.verbose:
             self.fileprinted = True
         s = _("patching file %s\n") % self.fname
         if warn:
             self.ui.warn(s)
         else:
             self.ui.note(s)
 
 
     def findlines(self, l, linenum):
         # looks through the hash and finds candidate lines.  The
         # result is a list of line numbers sorted based on distance
         # from linenum
 
         cand = self.hash.get(l, [])
         if len(cand) > 1:
             # resort our list of potentials forward then back.
             cand.sort(key=lambda x: abs(x - linenum))
         return cand
 
     def write_rej(self):
         # our rejects are a little different from patch(1).  This always
         # creates rejects in the same form as the original patch.  A file
         # header is inserted so that you can run the reject through patch again
         # without having to type the filename.
         if not self.rej:
             return
         base = os.path.basename(self.fname)
         lines = ["--- %s\n+++ %s\n" % (base, base)]
         for x in self.rej:
             for l in x.hunk:
                 lines.append(l)
                 if l[-1] != '\n':
                     lines.append("\n\ No newline at end of file\n")
         self.backend.writerej(self.fname, len(self.rej), self.hunks, lines)
 
     def apply(self, h):
         if not h.complete():
             raise PatchError(_("bad hunk #%d %s (%d %d %d %d)") %
                             (h.number, h.desc, len(h.a), h.lena, len(h.b),
                             h.lenb))
 
         self.hunks += 1
 
         if self.missing:
             self.rej.append(h)
             return -1
 
         if self.exists and self.create:
             if self.copysource:
                 self.ui.warn(_("cannot create %s: destination already "
                                "exists\n") % self.fname)
             else:
                 self.ui.warn(_("file %s already exists\n") % self.fname)
             self.rej.append(h)
             return -1
 
         if isinstance(h, binhunk):
             if self.remove:
                 self.backend.unlink(self.fname)
             else:
                 l = h.new(self.lines)
                 self.lines[:] = l
                 self.offset += len(l)
                 self.dirty = True
             return 0
 
         horig = h
         if (self.eolmode in ('crlf', 'lf')
             or self.eolmode == 'auto' and self.eol):
             # If new eols are going to be normalized, then normalize
             # hunk data before patching. Otherwise, preserve input
             # line-endings.
             h = h.getnormalized()
 
         # fast case first, no offsets, no fuzz
         old, oldstart, new, newstart = h.fuzzit(0, False)
         oldstart += self.offset
         orig_start = oldstart
         # if there's skew we want to emit the "(offset %d lines)" even
         # when the hunk cleanly applies at start + skew, so skip the
         # fast case code
         if (self.skew == 0 and
             diffhelpers.testhunk(old, self.lines, oldstart) == 0):
             if self.remove:
                 self.backend.unlink(self.fname)
             else:
                 self.lines[oldstart:oldstart + len(old)] = new
                 self.offset += len(new) - len(old)
                 self.dirty = True
             return 0
 
         # ok, we couldn't match the hunk. Lets look for offsets and fuzz it
         self.hash = {}
         for x, s in enumerate(self.lines):
             self.hash.setdefault(s, []).append(x)
 
         for fuzzlen in xrange(self.ui.configint("patch", "fuzz", 2) + 1):
             for toponly in [True, False]:
                 old, oldstart, new, newstart = h.fuzzit(fuzzlen, toponly)
                 oldstart = oldstart + self.offset + self.skew
                 oldstart = min(oldstart, len(self.lines))
                 if old:
                     cand = self.findlines(old[0][1:], oldstart)
                 else:
                     # Only adding lines with no or fuzzed context, just
                     # take the skew in account
                     cand = [oldstart]
 
                 for l in cand:
                     if not old or diffhelpers.testhunk(old, self.lines, l) == 0:
                         self.lines[l : l + len(old)] = new
                         self.offset += len(new) - len(old)
                         self.skew = l - orig_start
                         self.dirty = True
                         offset = l - orig_start - fuzzlen
                         if fuzzlen:
                             msg = _("Hunk #%d succeeded at %d "
                                     "with fuzz %d "
                                     "(offset %d lines).\n")
                             self.printfile(True)
                             self.ui.warn(msg %
                                 (h.number, l + 1, fuzzlen, offset))
                         else:
                             msg = _("Hunk #%d succeeded at %d "
                                     "(offset %d lines).\n")
                             self.ui.note(msg % (h.number, l + 1, offset))
                         return fuzzlen
         self.printfile(True)
         self.ui.warn(_("Hunk #%d FAILED at %d\n") % (h.number, orig_start))
         self.rej.append(horig)
         return -1
 
     def close(self):
         if self.dirty:
             self.writelines(self.fname, self.lines, self.mode)
         self.write_rej()
         return len(self.rej)
 
 class header(object):
     """patch header
     """
     diffgit_re = re.compile('diff --git a/(.*) b/(.*)$')
     diff_re = re.compile('diff -r .* (.*)$')
     allhunks_re = re.compile('(?:index|deleted file) ')
     pretty_re = re.compile('(?:new file|deleted file) ')
     special_re = re.compile('(?:index|deleted|copy|rename) ')
     newfile_re = re.compile('(?:new file)')
 
     def __init__(self, header):
         self.header = header
         self.hunks = []
 
     def binary(self):
         return any(h.startswith('index ') for h in self.header)
 
     def pretty(self, fp):
         for h in self.header:
             if h.startswith('index '):
                 fp.write(_('this modifies a binary file (all or nothing)\n'))
                 break
             if self.pretty_re.match(h):
                 fp.write(h)
                 if self.binary():
                     fp.write(_('this is a binary file\n'))
                 break
             if h.startswith('---'):
                 fp.write(_('%d hunks, %d lines changed\n') %
                          (len(self.hunks),
                           sum([max(h.added, h.removed) for h in self.hunks])))
                 break
             fp.write(h)
 
     def write(self, fp):
         fp.write(''.join(self.header))
 
     def allhunks(self):
         return any(self.allhunks_re.match(h) for h in self.header)
 
     def files(self):
         match = self.diffgit_re.match(self.header[0])
         if match:
             fromfile, tofile = match.groups()
             if fromfile == tofile:
                 return [fromfile]
             return [fromfile, tofile]
         else:
             return self.diff_re.match(self.header[0]).groups()
 
     def filename(self):
         return self.files()[-1]
 
     def __repr__(self):
         return '<header %s>' % (' '.join(map(repr, self.files())))
 
     def isnewfile(self):
         return any(self.newfile_re.match(h) for h in self.header)
 
     def special(self):
         # Special files are shown only at the header level and not at the hunk
         # level for example a file that has been deleted is a special file.
         # The user cannot change the content of the operation, in the case of
         # the deleted file he has to take the deletion or not take it, he
         # cannot take some of it.
         # Newly added files are special if they are empty, they are not special
         # if they have some content as we want to be able to change it
         nocontent = len(self.header) == 2
         emptynewfile = self.isnewfile() and nocontent
         return emptynewfile or \
                 any(self.special_re.match(h) for h in self.header)
 
 class recordhunk(object):
     """patch hunk
 
     XXX shouldn't we merge this with the other hunk class?
     """
     maxcontext = 3
 
     def __init__(self, header, fromline, toline, proc, before, hunk, after):
         def trimcontext(number, lines):
             delta = len(lines) - self.maxcontext
             if False and delta > 0:
                 return number + delta, lines[:self.maxcontext]
             return number, lines
 
         self.header = header
         self.fromline, self.before = trimcontext(fromline, before)
         self.toline, self.after = trimcontext(toline, after)
         self.proc = proc
         self.hunk = hunk
         self.added, self.removed = self.countchanges(self.hunk)
 
     def __eq__(self, v):
         if not isinstance(v, recordhunk):
             return False
 
         return ((v.hunk == self.hunk) and
                 (v.proc == self.proc) and
                 (self.fromline == v.fromline) and
                 (self.header.files() == v.header.files()))
 
     def __hash__(self):
         return hash((tuple(self.hunk),
             tuple(self.header.files()),
             self.fromline,
             self.proc))
 
     def countchanges(self, hunk):
         """hunk -> (n+,n-)"""
         add = len([h for h in hunk if h[0] == '+'])
         rem = len([h for h in hunk if h[0] == '-'])
         return add, rem
 
     def write(self, fp):
         delta = len(self.before) + len(self.after)
         if self.after and self.after[-1] == '\\ No newline at end of file\n':
             delta -= 1
         fromlen = delta + self.removed
         tolen = delta + self.added
         fp.write('@@ -%d,%d +%d,%d @@%s\n' %
                  (self.fromline, fromlen, self.toline, tolen,
                   self.proc and (' ' + self.proc)))
         fp.write(''.join(self.before + self.hunk + self.after))
 
     pretty = write
 
     def filename(self):
         return self.header.filename()
 
     def __repr__(self):
         return '<hunk %r@%d>' % (self.filename(), self.fromline)
 
 def filterpatch(ui, headers, operation=None):
     """Interactively filter patch chunks into applied-only chunks"""
     if operation is None:
         operation = 'record'
     messages = {
         'multiple': {
             'discard': _("discard change %d/%d to '%s'?"),
             'record': _("record change %d/%d to '%s'?"),
             'revert': _("revert change %d/%d to '%s'?"),
         }[operation],
         'single': {
             'discard': _("discard this change to '%s'?"),
             'record': _("record this change to '%s'?"),
             'revert': _("revert this change to '%s'?"),
         }[operation],
     }
 
     def prompt(skipfile, skipall, query, chunk):
         """prompt query, and process base inputs
 
         - y/n for the rest of file
         - y/n for the rest
         - ? (help)
         - q (quit)
 
         Return True/False and possibly updated skipfile and skipall.
         """
         newpatches = None
         if skipall is not None:
             return skipall, skipfile, skipall, newpatches
         if skipfile is not None:
             return skipfile, skipfile, skipall, newpatches
         while True:
             resps = _('[Ynesfdaq?]'
                       '$$ &Yes, record this change'
                       '$$ &No, skip this change'
                       '$$ &Edit this change manually'
                       '$$ &Skip remaining changes to this file'
                       '$$ Record remaining changes to this &file'
                       '$$ &Done, skip remaining changes and files'
                       '$$ Record &all changes to all remaining files'
                       '$$ &Quit, recording no changes'
                       '$$ &? (display help)')
             r = ui.promptchoice("%s %s" % (query, resps))
             ui.write("\n")
             if r == 8: # ?
                 for c, t in ui.extractchoices(resps)[1]:
                     ui.write('%s - %s\n' % (c, encoding.lower(t)))
                 continue
             elif r == 0: # yes
                 ret = True
             elif r == 1: # no
                 ret = False
             elif r == 2: # Edit patch
                 if chunk is None:
                     ui.write(_('cannot edit patch for whole file'))
                     ui.write("\n")
                     continue
                 if chunk.header.binary():
                     ui.write(_('cannot edit patch for binary file'))
                     ui.write("\n")
                     continue
                 # Patch comment based on the Git one (based on comment at end of
                 # https://mercurial-scm.org/wiki/RecordExtension)
                 phelp = '---' + _("""
 To remove '-' lines, make them ' ' lines (context).
 To remove '+' lines, delete them.
 Lines starting with # will be removed from the patch.
 
 If the patch applies cleanly, the edited hunk will immediately be
 added to the record list. If it does not apply cleanly, a rejects
 file will be generated: you can use that when you try again. If
 all lines of the hunk are removed, then the edit is aborted and
 the hunk is left unchanged.
 """)
                 (patchfd, patchfn) = tempfile.mkstemp(prefix="hg-editor-",
                         suffix=".diff", text=True)
                 ncpatchfp = None
                 try:
                     # Write the initial patch
-                    f = os.fdopen(patchfd, "w")
+                    f = os.fdopen(patchfd, pycompat.sysstr("w"))
                     chunk.header.write(f)
                     chunk.write(f)
                     f.write('\n'.join(['# ' + i for i in phelp.splitlines()]))
                     f.close()
                     # Start the editor and wait for it to complete
                     editor = ui.geteditor()
                     ret = ui.system("%s \"%s\"" % (editor, patchfn),
                                     environ={'HGUSER': ui.username()})
                     if ret != 0:
                         ui.warn(_("editor exited with exit code %d\n") % ret)
                         continue
                     # Remove comment lines
                     patchfp = open(patchfn)
                     ncpatchfp = stringio()
                     for line in util.iterfile(patchfp):
                         if not line.startswith('#'):
                             ncpatchfp.write(line)
                     patchfp.close()
                     ncpatchfp.seek(0)
                     newpatches = parsepatch(ncpatchfp)
                 finally:
                     os.unlink(patchfn)
                     del ncpatchfp
                 # Signal that the chunk shouldn't be applied as-is, but
                 # provide the new patch to be used instead.
                 ret = False
             elif r == 3: # Skip
                 ret = skipfile = False
             elif r == 4: # file (Record remaining)
                 ret = skipfile = True
             elif r == 5: # done, skip remaining
                 ret = skipall = False
             elif r == 6: # all
                 ret = skipall = True
             elif r == 7: # quit
                 raise error.Abort(_('user quit'))
             return ret, skipfile, skipall, newpatches
 
     seen = set()
     applied = {}        # 'filename' -> [] of chunks
     skipfile, skipall = None, None
     pos, total = 1, sum(len(h.hunks) for h in headers)
     for h in headers:
         pos += len(h.hunks)
         skipfile = None
         fixoffset = 0
         hdr = ''.join(h.header)
         if hdr in seen:
             continue
         seen.add(hdr)
         if skipall is None:
             h.pretty(ui)
         msg = (_('examine changes to %s?') %
                _(' and ').join("'%s'" % f for f in h.files()))
         r, skipfile, skipall, np = prompt(skipfile, skipall, msg, None)
         if not r:
             continue
         applied[h.filename()] = [h]
         if h.allhunks():
             applied[h.filename()] += h.hunks
             continue
         for i, chunk in enumerate(h.hunks):
             if skipfile is None and skipall is None:
                 chunk.pretty(ui)
             if total == 1:
                 msg = messages['single'] % chunk.filename()
             else:
                 idx = pos - len(h.hunks) + i
                 msg = messages['multiple'] % (idx, total, chunk.filename())
             r, skipfile, skipall, newpatches = prompt(skipfile,
                     skipall, msg, chunk)
             if r:
                 if fixoffset:
                     chunk = copy.copy(chunk)
                     chunk.toline += fixoffset
                 applied[chunk.filename()].append(chunk)
             elif newpatches is not None:
                 for newpatch in newpatches:
                     for newhunk in newpatch.hunks:
                         if fixoffset:
                             newhunk.toline += fixoffset
                         applied[newhunk.filename()].append(newhunk)
             else:
                 fixoffset += chunk.removed - chunk.added
     return (sum([h for h in applied.itervalues()
                if h[0].special() or len(h) > 1], []), {})
 class hunk(object):
     def __init__(self, desc, num, lr, context):
         self.number = num
         self.desc = desc
         self.hunk = [desc]
         self.a = []
         self.b = []
         self.starta = self.lena = None
         self.startb = self.lenb = None
         if lr is not None:
             if context:
                 self.read_context_hunk(lr)
             else:
                 self.read_unified_hunk(lr)
 
     def getnormalized(self):
         """Return a copy with line endings normalized to LF."""
 
         def normalize(lines):
             nlines = []
             for line in lines:
                 if line.endswith('\r\n'):
                     line = line[:-2] + '\n'
                 nlines.append(line)
             return nlines
 
         # Dummy object, it is rebuilt manually
         nh = hunk(self.desc, self.number, None, None)
         nh.number = self.number
         nh.desc = self.desc
         nh.hunk = self.hunk
         nh.a = normalize(self.a)
         nh.b = normalize(self.b)
         nh.starta = self.starta
         nh.startb = self.startb
         nh.lena = self.lena
         nh.lenb = self.lenb
         return nh
 
     def read_unified_hunk(self, lr):
         m = unidesc.match(self.desc)
         if not m:
             raise PatchError(_("bad hunk #%d") % self.number)
         self.starta, self.lena, self.startb, self.lenb = m.groups()
         if self.lena is None:
             self.lena = 1
         else:
             self.lena = int(self.lena)
         if self.lenb is None:
             self.lenb = 1
         else:
             self.lenb = int(self.lenb)
         self.starta = int(self.starta)
         self.startb = int(self.startb)
         diffhelpers.addlines(lr, self.hunk, self.lena, self.lenb, self.a,
                              self.b)
         # if we hit eof before finishing out the hunk, the last line will
         # be zero length.  Lets try to fix it up.
         while len(self.hunk[-1]) == 0:
             del self.hunk[-1]
             del self.a[-1]
             del self.b[-1]
             self.lena -= 1
             self.lenb -= 1
         self._fixnewline(lr)
 
     def read_context_hunk(self, lr):
         self.desc = lr.readline()
         m = contextdesc.match(self.desc)
         if not m:
             raise PatchError(_("bad hunk #%d") % self.number)
         self.starta, aend = m.groups()
         self.starta = int(self.starta)
         if aend is None:
             aend = self.starta
         self.lena = int(aend) - self.starta
         if self.starta:
             self.lena += 1
         for x in xrange(self.lena):
             l = lr.readline()
             if l.startswith('---'):
                 # lines addition, old block is empty
                 lr.push(l)
                 break
             s = l[2:]
             if l.startswith('- ') or l.startswith('! '):
                 u = '-' + s
             elif l.startswith('  '):
                 u = ' ' + s
             else:
                 raise PatchError(_("bad hunk #%d old text line %d") %
                                  (self.number, x))
             self.a.append(u)
             self.hunk.append(u)
 
         l = lr.readline()
         if l.startswith('\ '):
             s = self.a[-1][:-1]
             self.a[-1] = s
             self.hunk[-1] = s
             l = lr.readline()
         m = contextdesc.match(l)
         if not m:
             raise PatchError(_("bad hunk #%d") % self.number)
         self.startb, bend = m.groups()
         self.startb = int(self.startb)
         if bend is None:
             bend = self.startb
         self.lenb = int(bend) - self.startb
         if self.startb:
             self.lenb += 1
         hunki = 1
         for x in xrange(self.lenb):
             l = lr.readline()
             if l.startswith('\ '):
                 # XXX: the only way to hit this is with an invalid line range.
                 # The no-eol marker is not counted in the line range, but I
                 # guess there are diff(1) out there which behave differently.
                 s = self.b[-1][:-1]
                 self.b[-1] = s
                 self.hunk[hunki - 1] = s
                 continue
             if not l:
                 # line deletions, new block is empty and we hit EOF
                 lr.push(l)
                 break
             s = l[2:]
             if l.startswith('+ ') or l.startswith('! '):
                 u = '+' + s
             elif l.startswith('  '):
                 u = ' ' + s
             elif len(self.b) == 0:
                 # line deletions, new block is empty
                 lr.push(l)
                 break
             else:
                 raise PatchError(_("bad hunk #%d old text line %d") %
                                  (self.number, x))
             self.b.append(s)
             while True:
                 if hunki >= len(self.hunk):
                     h = ""
                 else:
                     h = self.hunk[hunki]
                 hunki += 1
                 if h == u:
                     break
                 elif h.startswith('-'):
                     continue
                 else:
                     self.hunk.insert(hunki - 1, u)
                     break
 
         if not self.a:
             # this happens when lines were only added to the hunk
             for x in self.hunk:
                 if x.startswith('-') or x.startswith(' '):
                     self.a.append(x)
         if not self.b:
             # this happens when lines were only deleted from the hunk
             for x in self.hunk:
                 if x.startswith('+') or x.startswith(' '):
                     self.b.append(x[1:])
         # @@ -start,len +start,len @@
         self.desc = "@@ -%d,%d +%d,%d @@\n" % (self.starta, self.lena,
                                              self.startb, self.lenb)
         self.hunk[0] = self.desc
         self._fixnewline(lr)
 
     def _fixnewline(self, lr):
         l = lr.readline()
         if l.startswith('\ '):
             diffhelpers.fix_newline(self.hunk, self.a, self.b)
         else:
             lr.push(l)
 
     def complete(self):
         return len(self.a) == self.lena and len(self.b) == self.lenb
 
     def _fuzzit(self, old, new, fuzz, toponly):
         # this removes context lines from the top and bottom of list 'l'.  It
         # checks the hunk to make sure only context lines are removed, and then
         # returns a new shortened list of lines.
         fuzz = min(fuzz, len(old))
         if fuzz:
             top = 0
             bot = 0
             hlen = len(self.hunk)
             for x in xrange(hlen - 1):
                 # the hunk starts with the @@ line, so use x+1
                 if self.hunk[x + 1][0] == ' ':
                     top += 1
                 else:
                     break
             if not toponly:
                 for x in xrange(hlen - 1):
                     if self.hunk[hlen - bot - 1][0] == ' ':
                         bot += 1
                     else:
                         break
 
             bot = min(fuzz, bot)
             top = min(fuzz, top)
             return old[top:len(old) - bot], new[top:len(new) - bot], top
         return old, new, 0
 
     def fuzzit(self, fuzz, toponly):
         old, new, top = self._fuzzit(self.a, self.b, fuzz, toponly)
         oldstart = self.starta + top
         newstart = self.startb + top
         # zero length hunk ranges already have their start decremented
         if self.lena and oldstart > 0:
             oldstart -= 1
         if self.lenb and newstart > 0:
             newstart -= 1
         return old, oldstart, new, newstart
 
 class binhunk(object):
     'A binary patch file.'
     def __init__(self, lr, fname):
         self.text = None
         self.delta = False
         self.hunk = ['GIT binary patch\n']
         self._fname = fname
         self._read(lr)
 
     def complete(self):
         return self.text is not None
 
     def new(self, lines):
         if self.delta:
             return [applybindelta(self.text, ''.join(lines))]
         return [self.text]
 
     def _read(self, lr):
         def getline(lr, hunk):
             l = lr.readline()
             hunk.append(l)
             return l.rstrip('\r\n')
 
         size = 0
         while True:
             line = getline(lr, self.hunk)
             if not line:
                 raise PatchError(_('could not extract "%s" binary data')
                                  % self._fname)
             if line.startswith('literal '):
                 size = int(line[8:].rstrip())
                 break
             if line.startswith('delta '):
                 size = int(line[6:].rstrip())
                 self.delta = True
                 break
         dec = []
         line = getline(lr, self.hunk)
         while len(line) > 1:
             l = line[0]
             if l <= 'Z' and l >= 'A':
                 l = ord(l) - ord('A') + 1
             else:
                 l = ord(l) - ord('a') + 27
             try:
                 dec.append(base85.b85decode(line[1:])[:l])
             except ValueError as e:
                 raise PatchError(_('could not decode "%s" binary patch: %s')
                                  % (self._fname, str(e)))
             line = getline(lr, self.hunk)
         text = zlib.decompress(''.join(dec))
         if len(text) != size:
             raise PatchError(_('"%s" length is %d bytes, should be %d')
                              % (self._fname, len(text), size))
         self.text = text
 
 def parsefilename(str):
     # --- filename \t|space stuff
     s = str[4:].rstrip('\r\n')
     i = s.find('\t')
     if i < 0:
         i = s.find(' ')
         if i < 0:
             return s
     return s[:i]
 
 def reversehunks(hunks):
     '''reverse the signs in the hunks given as argument
 
     This function operates on hunks coming out of patch.filterpatch, that is
     a list of the form: [header1, hunk1, hunk2, header2...]. Example usage:
 
     >>> rawpatch = """diff --git a/folder1/g b/folder1/g
     ... --- a/folder1/g
     ... +++ b/folder1/g
     ... @@ -1,7 +1,7 @@
     ... +firstline
     ...  c
     ...  1
     ...  2
     ... + 3
     ... -4
     ...  5
     ...  d
     ... +lastline"""
     >>> hunks = parsepatch(rawpatch)
     >>> hunkscomingfromfilterpatch = []
     >>> for h in hunks:
     ...     hunkscomingfromfilterpatch.append(h)
     ...     hunkscomingfromfilterpatch.extend(h.hunks)
 
     >>> reversedhunks = reversehunks(hunkscomingfromfilterpatch)
     >>> from . import util
     >>> fp = util.stringio()
     >>> for c in reversedhunks:
     ...      c.write(fp)
     >>> fp.seek(0)
     >>> reversedpatch = fp.read()
     >>> print reversedpatch
     diff --git a/folder1/g b/folder1/g
     --- a/folder1/g
     +++ b/folder1/g
     @@ -1,4 +1,3 @@
     -firstline
      c
      1
      2
     @@ -1,6 +2,6 @@
      c
      1
      2
     - 3
     +4
      5
      d
     @@ -5,3 +6,2 @@
      5
      d
     -lastline
 
     '''
 
     from . import crecord as crecordmod
     newhunks = []
     for c in hunks:
         if isinstance(c, crecordmod.uihunk):
             # curses hunks encapsulate the record hunk in _hunk
             c = c._hunk
         if isinstance(c, recordhunk):
             for j, line in enumerate(c.hunk):
                 if line.startswith("-"):
                     c.hunk[j] = "+" + c.hunk[j][1:]
                 elif line.startswith("+"):
                     c.hunk[j] = "-" + c.hunk[j][1:]
             c.added, c.removed = c.removed, c.added
         newhunks.append(c)
     return newhunks
 
 def parsepatch(originalchunks):
     """patch -> [] of headers -> [] of hunks """
     class parser(object):
         """patch parsing state machine"""
         def __init__(self):
             self.fromline = 0
             self.toline = 0
             self.proc = ''
             self.header = None
             self.context = []
             self.before = []
             self.hunk = []
             self.headers = []
 
         def addrange(self, limits):
             fromstart, fromend, tostart, toend, proc = limits
             self.fromline = int(fromstart)
             self.toline = int(tostart)
             self.proc = proc
 
         def addcontext(self, context):
             if self.hunk:
                 h = recordhunk(self.header, self.fromline, self.toline,
                         self.proc, self.before, self.hunk, context)
                 self.header.hunks.append(h)
                 self.fromline += len(self.before) + h.removed
                 self.toline += len(self.before) + h.added
                 self.before = []
                 self.hunk = []
             self.context = context
 
         def addhunk(self, hunk):
             if self.context:
                 self.before = self.context
                 self.context = []
             self.hunk = hunk
 
         def newfile(self, hdr):
             self.addcontext([])
             h = header(hdr)
             self.headers.append(h)
             self.header = h
 
         def addother(self, line):
             pass # 'other' lines are ignored
 
         def finished(self):
             self.addcontext([])
             return self.headers
 
         transitions = {
             'file': {'context': addcontext,
                      'file': newfile,
                      'hunk': addhunk,
                      'range': addrange},
             'context': {'file': newfile,
                         'hunk': addhunk,
                         'range': addrange,
                         'other': addother},
             'hunk': {'context': addcontext,
                      'file': newfile,
                      'range': addrange},
             'range': {'context': addcontext,
                       'hunk': addhunk},
             'other': {'other': addother},
             }
 
     p = parser()
     fp = stringio()
     fp.write(''.join(originalchunks))
     fp.seek(0)
 
     state = 'context'
     for newstate, data in scanpatch(fp):
         try:
             p.transitions[state][newstate](p, data)
         except KeyError:
             raise PatchError('unhandled transition: %s -> %s' %
                                    (state, newstate))
         state = newstate
     del fp
     return p.finished()
 
 def pathtransform(path, strip, prefix):
     '''turn a path from a patch into a path suitable for the repository
 
     prefix, if not empty, is expected to be normalized with a / at the end.
 
     Returns (stripped components, path in repository).
 
     >>> pathtransform('a/b/c', 0, '')
     ('', 'a/b/c')
     >>> pathtransform('   a/b/c   ', 0, '')
     ('', '   a/b/c')
     >>> pathtransform('   a/b/c   ', 2, '')
     ('a/b/', 'c')
     >>> pathtransform('a/b/c', 0, 'd/e/')
     ('', 'd/e/a/b/c')
     >>> pathtransform('   a//b/c   ', 2, 'd/e/')
     ('a//b/', 'd/e/c')
     >>> pathtransform('a/b/c', 3, '')
     Traceback (most recent call last):
     PatchError: unable to strip away 1 of 3 dirs from a/b/c
     '''
     pathlen = len(path)
     i = 0
     if strip == 0:
         return '', prefix + path.rstrip()
     count = strip
     while count > 0:
         i = path.find('/', i)
         if i == -1:
             raise PatchError(_("unable to strip away %d of %d dirs from %s") %
                              (count, strip, path))
         i += 1
         # consume '//' in the path
         while i < pathlen - 1 and path[i] == '/':
             i += 1
         count -= 1
     return path[:i].lstrip(), prefix + path[i:].rstrip()
 
 def makepatchmeta(backend, afile_orig, bfile_orig, hunk, strip, prefix):
     nulla = afile_orig == "/dev/null"
     nullb = bfile_orig == "/dev/null"
     create = nulla and hunk.starta == 0 and hunk.lena == 0
     remove = nullb and hunk.startb == 0 and hunk.lenb == 0
     abase, afile = pathtransform(afile_orig, strip, prefix)
     gooda = not nulla and backend.exists(afile)
     bbase, bfile = pathtransform(bfile_orig, strip, prefix)
     if afile == bfile:
         goodb = gooda
     else:
         goodb = not nullb and backend.exists(bfile)
     missing = not goodb and not gooda and not create
 
     # some diff programs apparently produce patches where the afile is
     # not /dev/null, but afile starts with bfile
     abasedir = afile[:afile.rfind('/') + 1]
     bbasedir = bfile[:bfile.rfind('/') + 1]
     if (missing and abasedir == bbasedir and afile.startswith(bfile)
         and hunk.starta == 0 and hunk.lena == 0):
         create = True
         missing = False
 
     # If afile is "a/b/foo" and bfile is "a/b/foo.orig" we assume the
     # diff is between a file and its backup. In this case, the original
     # file should be patched (see original mpatch code).
     isbackup = (abase == bbase and bfile.startswith(afile))
     fname = None
     if not missing:
         if gooda and goodb:
             if isbackup:
                 fname = afile
             else:
                 fname = bfile
         elif gooda:
             fname = afile
 
     if not fname:
         if not nullb:
             if isbackup:
                 fname = afile
             else:
                 fname = bfile
         elif not nulla:
             fname = afile
         else:
             raise PatchError(_("undefined source and destination files"))
 
     gp = patchmeta(fname)
     if create:
         gp.op = 'ADD'
     elif remove:
         gp.op = 'DELETE'
     return gp
 
 def scanpatch(fp):
     """like patch.iterhunks, but yield different events
 
     - ('file',    [header_lines + fromfile + tofile])
     - ('context', [context_lines])
     - ('hunk',    [hunk_lines])
     - ('range',   (-start,len, +start,len, proc))
     """
     lines_re = re.compile(r'@@ -(\d+),(\d+) \+(\d+),(\d+) @@\s*(.*)')
     lr = linereader(fp)
 
     def scanwhile(first, p):
         """scan lr while predicate holds"""
         lines = [first]
         for line in iter(lr.readline, ''):
             if p(line):
                 lines.append(line)
             else:
                 lr.push(line)
                 break
         return lines
 
     for line in iter(lr.readline, ''):
         if line.startswith('diff --git a/') or line.startswith('diff -r '):
             def notheader(line):
                 s = line.split(None, 1)
                 return not s or s[0] not in ('---', 'diff')
             header = scanwhile(line, notheader)
             fromfile = lr.readline()
             if fromfile.startswith('---'):
                 tofile = lr.readline()
                 header += [fromfile, tofile]
             else:
                 lr.push(fromfile)
             yield 'file', header
         elif line[0] == ' ':
             yield 'context', scanwhile(line, lambda l: l[0] in ' \\')
         elif line[0] in '-+':
             yield 'hunk', scanwhile(line, lambda l: l[0] in '-+\\')
         else:
             m = lines_re.match(line)
             if m:
                 yield 'range', m.groups()
             else:
                 yield 'other', line
 
 def scangitpatch(lr, firstline):
     """
     Git patches can emit:
     - rename a to b
     - change b
     - copy a to c
     - change c
 
     We cannot apply this sequence as-is, the renamed 'a' could not be
     found for it would have been renamed already. And we cannot copy
     from 'b' instead because 'b' would have been changed already. So
     we scan the git patch for copy and rename commands so we can
     perform the copies ahead of time.
     """
     pos = 0
     try:
         pos = lr.fp.tell()
         fp = lr.fp
     except IOError:
         fp = stringio(lr.fp.read())
     gitlr = linereader(fp)
     gitlr.push(firstline)
     gitpatches = readgitpatch(gitlr)
     fp.seek(pos)
     return gitpatches
 
 def iterhunks(fp):
     """Read a patch and yield the following events:
     - ("file", afile, bfile, firsthunk): select a new target file.
     - ("hunk", hunk): a new hunk is ready to be applied, follows a
     "file" event.
     - ("git", gitchanges): current diff is in git format, gitchanges
     maps filenames to gitpatch records. Unique event.
     """
     afile = ""
     bfile = ""
     state = None
     hunknum = 0
     emitfile = newfile = False
     gitpatches = None
 
     # our states
     BFILE = 1
     context = None
     lr = linereader(fp)
 
     for x in iter(lr.readline, ''):
         if state == BFILE and (
             (not context and x[0] == '@')
             or (context is not False and x.startswith('***************'))
             or x.startswith('GIT binary patch')):
             gp = None
             if (gitpatches and
                 gitpatches[-1].ispatching(afile, bfile)):
                 gp = gitpatches.pop()
             if x.startswith('GIT binary patch'):
                 h = binhunk(lr, gp.path)
             else:
                 if context is None and x.startswith('***************'):
                     context = True
                 h = hunk(x, hunknum + 1, lr, context)
             hunknum += 1
             if emitfile:
                 emitfile = False
                 yield 'file', (afile, bfile, h, gp and gp.copy() or None)
             yield 'hunk', h
         elif x.startswith('diff --git a/'):
             m = gitre.match(x.rstrip(' \r\n'))
             if not m:
                 continue
             if gitpatches is None:
                 # scan whole input for git metadata
                 gitpatches = scangitpatch(lr, x)
                 yield 'git', [g.copy() for g in gitpatches
                               if g.op in ('COPY', 'RENAME')]
                 gitpatches.reverse()
             afile = 'a/' + m.group(1)
             bfile = 'b/' + m.group(2)
             while gitpatches and not gitpatches[-1].ispatching(afile, bfile):
                 gp = gitpatches.pop()
                 yield 'file', ('a/' + gp.path, 'b/' + gp.path, None, gp.copy())
             if not gitpatches:
                 raise PatchError(_('failed to synchronize metadata for "%s"')
                                  % afile[2:])
             gp = gitpatches[-1]
             newfile = True
         elif x.startswith('---'):
             # check for a unified diff
             l2 = lr.readline()
             if not l2.startswith('+++'):
                 lr.push(l2)
                 continue
             newfile = True
             context = False
             afile = parsefilename(x)
             bfile = parsefilename(l2)
         elif x.startswith('***'):
             # check for a context diff
             l2 = lr.readline()
             if not l2.startswith('---'):
                 lr.push(l2)
                 continue
             l3 = lr.readline()
             lr.push(l3)
             if not l3.startswith("***************"):
                 lr.push(l2)
                 continue
             newfile = True
             context = True
             afile = parsefilename(x)
             bfile = parsefilename(l2)
 
         if newfile:
             newfile = False
             emitfile = True
             state = BFILE
             hunknum = 0
 
     while gitpatches:
         gp = gitpatches.pop()
         yield 'file', ('a/' + gp.path, 'b/' + gp.path, None, gp.copy())
 
 def applybindelta(binchunk, data):
     """Apply a binary delta hunk
     The algorithm used is the algorithm from git's patch-delta.c
     """
     def deltahead(binchunk):
         i = 0
         for c in binchunk:
             i += 1
             if not (ord(c) & 0x80):
                 return i
         return i
     out = ""
     s = deltahead(binchunk)
     binchunk = binchunk[s:]
     s = deltahead(binchunk)
     binchunk = binchunk[s:]
     i = 0
     while i < len(binchunk):
         cmd = ord(binchunk[i])
         i += 1
         if (cmd & 0x80):
             offset = 0
             size = 0
             if (cmd & 0x01):
                 offset = ord(binchunk[i])
                 i += 1
             if (cmd & 0x02):
                 offset |= ord(binchunk[i]) << 8
                 i += 1
             if (cmd & 0x04):
                 offset |= ord(binchunk[i]) << 16
                 i += 1
             if (cmd & 0x08):
                 offset |= ord(binchunk[i]) << 24
                 i += 1
             if (cmd & 0x10):
                 size = ord(binchunk[i])
                 i += 1
             if (cmd & 0x20):
                 size |= ord(binchunk[i]) << 8
                 i += 1
             if (cmd & 0x40):
                 size |= ord(binchunk[i]) << 16
                 i += 1
             if size == 0:
                 size = 0x10000
             offset_end = offset + size
             out += data[offset:offset_end]
         elif cmd != 0:
             offset_end = i + cmd
             out += binchunk[i:offset_end]
             i += cmd
         else:
             raise PatchError(_('unexpected delta opcode 0'))
     return out
 
 def applydiff(ui, fp, backend, store, strip=1, prefix='', eolmode='strict'):
     """Reads a patch from fp and tries to apply it.
 
     Returns 0 for a clean patch, -1 if any rejects were found and 1 if
     there was any fuzz.
 
     If 'eolmode' is 'strict', the patch content and patched file are
     read in binary mode. Otherwise, line endings are ignored when
     patching then normalized according to 'eolmode'.
     """
     return _applydiff(ui, fp, patchfile, backend, store, strip=strip,
                       prefix=prefix, eolmode=eolmode)
 
 def _applydiff(ui, fp, patcher, backend, store, strip=1, prefix='',
                eolmode='strict'):
 
     if prefix:
         prefix = pathutil.canonpath(backend.repo.root, backend.repo.getcwd(),
                                     prefix)
         if prefix != '':
             prefix += '/'
     def pstrip(p):
         return pathtransform(p, strip - 1, prefix)[1]
 
     rejects = 0
     err = 0
     current_file = None
 
     for state, values in iterhunks(fp):
         if state == 'hunk':
             if not current_file:
                 continue
             ret = current_file.apply(values)
             if ret > 0:
                 err = 1
         elif state == 'file':
             if current_file:
                 rejects += current_file.close()
                 current_file = None
             afile, bfile, first_hunk, gp = values
             if gp:
                 gp.path = pstrip(gp.path)
                 if gp.oldpath:
                     gp.oldpath = pstrip(gp.oldpath)
             else:
                 gp = makepatchmeta(backend, afile, bfile, first_hunk, strip,
                                    prefix)
             if gp.op == 'RENAME':
                 backend.unlink(gp.oldpath)
             if not first_hunk:
                 if gp.op == 'DELETE':
                     backend.unlink(gp.path)
                     continue
                 data, mode = None, None
                 if gp.op in ('RENAME', 'COPY'):
                     data, mode = store.getfile(gp.oldpath)[:2]
                     if data is None:
                         # This means that the old path does not exist
                         raise PatchError(_("source file '%s' does not exist")
                                            % gp.oldpath)
                 if gp.mode:
                     mode = gp.mode
                     if gp.op == 'ADD':
                         # Added files without content have no hunk and
                         # must be created
                         data = ''
                 if data or mode:
                     if (gp.op in ('ADD', 'RENAME', 'COPY')
                         and backend.exists(gp.path)):
                         raise PatchError(_("cannot create %s: destination "
                                            "already exists") % gp.path)
                     backend.setfile(gp.path, data, mode, gp.oldpath)
                 continue
             try:
                 current_file = patcher(ui, gp, backend, store,
                                        eolmode=eolmode)
             except PatchError as inst:
                 ui.warn(str(inst) + '\n')
                 current_file = None
                 rejects += 1
                 continue
         elif state == 'git':
             for gp in values:
                 path = pstrip(gp.oldpath)
                 data, mode = backend.getfile(path)
                 if data is None:
                     # The error ignored here will trigger a getfile()
                     # error in a place more appropriate for error
                     # handling, and will not interrupt the patching
                     # process.
                     pass
                 else:
                     store.setfile(path, data, mode)
         else:
             raise error.Abort(_('unsupported parser state: %s') % state)
 
     if current_file:
         rejects += current_file.close()
 
     if rejects:
         return -1
     return err
 
 def _externalpatch(ui, repo, patcher, patchname, strip, files,
                    similarity):
     """use <patcher> to apply <patchname> to the working directory.
     returns whether patch was applied with fuzz factor."""
 
     fuzz = False
     args = []
     cwd = repo.root
     if cwd:
         args.append('-d %s' % util.shellquote(cwd))
     fp = util.popen('%s %s -p%d < %s' % (patcher, ' '.join(args), strip,
                                        util.shellquote(patchname)))
     try:
         for line in util.iterfile(fp):
             line = line.rstrip()
             ui.note(line + '\n')
             if line.startswith('patching file '):
                 pf = util.parsepatchoutput(line)
                 printed_file = False
                 files.add(pf)
             elif line.find('with fuzz') >= 0:
                 fuzz = True
                 if not printed_file:
                     ui.warn(pf + '\n')
                     printed_file = True
                 ui.warn(line + '\n')
             elif line.find('saving rejects to file') >= 0:
                 ui.warn(line + '\n')
             elif line.find('FAILED') >= 0:
                 if not printed_file:
                     ui.warn(pf + '\n')
                     printed_file = True
                 ui.warn(line + '\n')
     finally:
         if files:
             scmutil.marktouched(repo, files, similarity)
     code = fp.close()
     if code:
         raise PatchError(_("patch command failed: %s") %
                          util.explainexit(code)[0])
     return fuzz
 
 def patchbackend(ui, backend, patchobj, strip, prefix, files=None,
                  eolmode='strict'):
     if files is None:
         files = set()
     if eolmode is None:
         eolmode = ui.config('patch', 'eol', 'strict')
     if eolmode.lower() not in eolmodes:
         raise error.Abort(_('unsupported line endings type: %s') % eolmode)
     eolmode = eolmode.lower()
 
     store = filestore()
     try:
         fp = open(patchobj, 'rb')
     except TypeError:
         fp = patchobj
     try:
         ret = applydiff(ui, fp, backend, store, strip=strip, prefix=prefix,
                         eolmode=eolmode)
     finally:
         if fp != patchobj:
             fp.close()
         files.update(backend.close())
         store.close()
     if ret < 0:
         raise PatchError(_('patch failed to apply'))
     return ret > 0
 
 def internalpatch(ui, repo, patchobj, strip, prefix='', files=None,
                   eolmode='strict', similarity=0):
     """use builtin patch to apply <patchobj> to the working directory.
     returns whether patch was applied with fuzz factor."""
     backend = workingbackend(ui, repo, similarity)
     return patchbackend(ui, backend, patchobj, strip, prefix, files, eolmode)
 
 def patchrepo(ui, repo, ctx, store, patchobj, strip, prefix, files=None,
               eolmode='strict'):
     backend = repobackend(ui, repo, ctx, store)
     return patchbackend(ui, backend, patchobj, strip, prefix, files, eolmode)
 
 def patch(ui, repo, patchname, strip=1, prefix='', files=None, eolmode='strict',
           similarity=0):
     """Apply <patchname> to the working directory.
 
     'eolmode' specifies how end of lines should be handled. It can be:
     - 'strict': inputs are read in binary mode, EOLs are preserved
     - 'crlf': EOLs are ignored when patching and reset to CRLF
     - 'lf': EOLs are ignored when patching and reset to LF
     - None: get it from user settings, default to 'strict'
     'eolmode' is ignored when using an external patcher program.
 
     Returns whether patch was applied with fuzz factor.
     """
     patcher = ui.config('ui', 'patch')
     if files is None:
         files = set()
     if patcher:
         return _externalpatch(ui, repo, patcher, patchname, strip,
                               files, similarity)
     return internalpatch(ui, repo, patchname, strip, prefix, files, eolmode,
                          similarity)
 
 def changedfiles(ui, repo, patchpath, strip=1):
     backend = fsbackend(ui, repo.root)
     with open(patchpath, 'rb') as fp:
         changed = set()
         for state, values in iterhunks(fp):
             if state == 'file':
                 afile, bfile, first_hunk, gp = values
                 if gp:
                     gp.path = pathtransform(gp.path, strip - 1, '')[1]
                     if gp.oldpath:
                         gp.oldpath = pathtransform(gp.oldpath, strip - 1, '')[1]
                 else:
                     gp = makepatchmeta(backend, afile, bfile, first_hunk, strip,
                                        '')
                 changed.add(gp.path)
                 if gp.op == 'RENAME':
                     changed.add(gp.oldpath)
             elif state not in ('hunk', 'git'):
                 raise error.Abort(_('unsupported parser state: %s') % state)
         return changed
 
 class GitDiffRequired(Exception):
     pass
 
 def diffallopts(ui, opts=None, untrusted=False, section='diff'):
     '''return diffopts with all features supported and parsed'''
     return difffeatureopts(ui, opts=opts, untrusted=untrusted, section=section,
                            git=True, whitespace=True, formatchanging=True)
 
 diffopts = diffallopts
 
 def difffeatureopts(ui, opts=None, untrusted=False, section='diff', git=False,
                     whitespace=False, formatchanging=False):
     '''return diffopts with only opted-in features parsed
 
     Features:
     - git: git-style diffs
     - whitespace: whitespace options like ignoreblanklines and ignorews
     - formatchanging: options that will likely break or cause correctness issues
       with most diff parsers
     '''
     def get(key, name=None, getter=ui.configbool, forceplain=None):
         if opts:
             v = opts.get(key)
             # diffopts flags are either None-default (which is passed
             # through unchanged, so we can identify unset values), or
             # some other falsey default (eg --unified, which defaults
             # to an empty string). We only want to override the config
             # entries from hgrc with command line values if they
             # appear to have been set, which is any truthy value,
             # True, or False.
             if v or isinstance(v, bool):
                 return v
         if forceplain is not None and ui.plain():
             return forceplain
         return getter(section, name or key, None, untrusted=untrusted)
 
     # core options, expected to be understood by every diff parser
     buildopts = {
         'nodates': get('nodates'),
         'showfunc': get('show_function', 'showfunc'),
         'context': get('unified', getter=ui.config),
     }
 
     if git:
         buildopts['git'] = get('git')
 
         # since this is in the experimental section, we need to call
         # ui.configbool directory
         buildopts['showsimilarity'] = ui.configbool('experimental',
                                                     'extendedheader.similarity')
 
         # need to inspect the ui object instead of using get() since we want to
         # test for an int
         hconf = ui.config('experimental', 'extendedheader.index')
         if hconf is not None:
             hlen = None
             try:
                 # the hash config could be an integer (for length of hash) or a
                 # word (e.g. short, full, none)
                 hlen = int(hconf)
                 if hlen < 0 or hlen > 40:
                     msg = _("invalid length for extendedheader.index: '%d'\n")
                     ui.warn(msg % hlen)
             except ValueError:
                 # default value
                 if hconf == 'short' or hconf == '':
                     hlen = 12
                 elif hconf == 'full':
                     hlen = 40
                 elif hconf != 'none':
                     msg = _("invalid value for extendedheader.index: '%s'\n")
                     ui.warn(msg % hconf)
             finally:
                 buildopts['index'] = hlen
 
     if whitespace:
         buildopts['ignorews'] = get('ignore_all_space', 'ignorews')
         buildopts['ignorewsamount'] = get('ignore_space_change',
                                           'ignorewsamount')
         buildopts['ignoreblanklines'] = get('ignore_blank_lines',
                                             'ignoreblanklines')
     if formatchanging:
         buildopts['text'] = opts and opts.get('text')
         buildopts['nobinary'] = get('nobinary', forceplain=False)
         buildopts['noprefix'] = get('noprefix', forceplain=False)
 
     return mdiff.diffopts(**buildopts)
 
 def diff(repo, node1=None, node2=None, match=None, changes=None, opts=None,
          losedatafn=None, prefix='', relroot='', copy=None):
     '''yields diff of changes to files between two nodes, or node and
     working directory.
 
     if node1 is None, use first dirstate parent instead.
     if node2 is None, compare node1 with working directory.
 
     losedatafn(**kwarg) is a callable run when opts.upgrade=True and
     every time some change cannot be represented with the current
     patch format. Return False to upgrade to git patch format, True to
     accept the loss or raise an exception to abort the diff. It is
     called with the name of current file being diffed as 'fn'. If set
     to None, patches will always be upgraded to git format when
     necessary.
 
     prefix is a filename prefix that is prepended to all filenames on
     display (used for subrepos).
 
     relroot, if not empty, must be normalized with a trailing /. Any match
     patterns that fall outside it will be ignored.
 
     copy, if not empty, should contain mappings {dst@y: src@x} of copy
     information.'''
 
     if opts is None:
         opts = mdiff.defaultopts
 
     if not node1 and not node2:
         node1 = repo.dirstate.p1()
 
     def lrugetfilectx():
         cache = {}
         order = collections.deque()
         def getfilectx(f, ctx):
             fctx = ctx.filectx(f, filelog=cache.get(f))
             if f not in cache:
                 if len(cache) > 20:
                     del cache[order.popleft()]
                 cache[f] = fctx.filelog()
             else:
                 order.remove(f)
             order.append(f)
             return fctx
         return getfilectx
     getfilectx = lrugetfilectx()
 
     ctx1 = repo[node1]
     ctx2 = repo[node2]
 
     relfiltered = False
     if relroot != '' and match.always():
         # as a special case, create a new matcher with just the relroot
         pats = [relroot]
         match = scmutil.match(ctx2, pats, default='path')
         relfiltered = True
 
     if not changes:
         changes = repo.status(ctx1, ctx2, match=match)
     modified, added, removed = changes[:3]
 
     if not modified and not added and not removed:
         return []
 
     if repo.ui.debugflag:
         hexfunc = hex
     else:
         hexfunc = short
     revs = [hexfunc(node) for node in [ctx1.node(), ctx2.node()] if node]
 
     if copy is None:
         copy = {}
         if opts.git or opts.upgrade:
             copy = copies.pathcopies(ctx1, ctx2, match=match)
 
     if relroot is not None:
         if not relfiltered:
             # XXX this would ideally be done in the matcher, but that is
             # generally meant to 'or' patterns, not 'and' them. In this case we
             # need to 'and' all the patterns from the matcher with relroot.
             def filterrel(l):
                 return [f for f in l if f.startswith(relroot)]
             modified = filterrel(modified)
             added = filterrel(added)
             removed = filterrel(removed)
             relfiltered = True
         # filter out copies where either side isn't inside the relative root
         copy = dict(((dst, src) for (dst, src) in copy.iteritems()
                      if dst.startswith(relroot)
                      and src.startswith(relroot)))
 
     modifiedset = set(modified)
     addedset = set(added)
     removedset = set(removed)
     for f in modified:
         if f not in ctx1:
             # Fix up added, since merged-in additions appear as
             # modifications during merges
             modifiedset.remove(f)
             addedset.add(f)
     for f in removed:
         if f not in ctx1:
             # Merged-in additions that are then removed are reported as removed.
             # They are not in ctx1, so We don't want to show them in the diff.
             removedset.remove(f)
     modified = sorted(modifiedset)
     added = sorted(addedset)
     removed = sorted(removedset)
     for dst, src in copy.items():
         if src not in ctx1:
             # Files merged in during a merge and then copied/renamed are
             # reported as copies. We want to show them in the diff as additions.
             del copy[dst]
 
     def difffn(opts, losedata):
         return trydiff(repo, revs, ctx1, ctx2, modified, added, removed,
                        copy, getfilectx, opts, losedata, prefix, relroot)
     if opts.upgrade and not opts.git:
         try:
             def losedata(fn):
                 if not losedatafn or not losedatafn(fn=fn):
                     raise GitDiffRequired
             # Buffer the whole output until we are sure it can be generated
             return list(difffn(opts.copy(git=False), losedata))
         except GitDiffRequired:
             return difffn(opts.copy(git=True), None)
     else:
         return difffn(opts, None)
 
 def difflabel(func, *args, **kw):
     '''yields 2-tuples of (output, label) based on the output of func()'''
     headprefixes = [('diff', 'diff.diffline'),
                     ('copy', 'diff.extended'),
                     ('rename', 'diff.extended'),
                     ('old', 'diff.extended'),
                     ('new', 'diff.extended'),
                     ('deleted', 'diff.extended'),
                     ('index', 'diff.extended'),
                     ('similarity', 'diff.extended'),
                     ('---', 'diff.file_a'),
                     ('+++', 'diff.file_b')]
     textprefixes = [('@', 'diff.hunk'),
                     ('-', 'diff.deleted'),
                     ('+', 'diff.inserted')]
     head = False
     for chunk in func(*args, **kw):
         lines = chunk.split('\n')
         for i, line in enumerate(lines):
             if i != 0:
                 yield ('\n', '')
             if head:
                 if line.startswith('@'):
                     head = False
             else:
                 if line and line[0] not in ' +-@\\':
                     head = True
             stripline = line
             diffline = False
             if not head and line and line[0] in '+-':
                 # highlight tabs and trailing whitespace, but only in
                 # changed lines
                 stripline = line.rstrip()
                 diffline = True
 
             prefixes = textprefixes
             if head:
                 prefixes = headprefixes
             for prefix, label in prefixes:
                 if stripline.startswith(prefix):
                     if diffline:
                         for token in tabsplitter.findall(stripline):
                             if '\t' == token[0]:
                                 yield (token, 'diff.tab')
                             else:
                                 yield (token, label)
                     else:
                         yield (stripline, label)
                     break
             else:
                 yield (line, '')
             if line != stripline:
                 yield (line[len(stripline):], 'diff.trailingwhitespace')
 
 def diffui(*args, **kw):
     '''like diff(), but yields 2-tuples of (output, label) for ui.write()'''
     return difflabel(diff, *args, **kw)
 
 def _filepairs(modified, added, removed, copy, opts):
     '''generates tuples (f1, f2, copyop), where f1 is the name of the file
     before and f2 is the the name after. For added files, f1 will be None,
     and for removed files, f2 will be None. copyop may be set to None, 'copy'
     or 'rename' (the latter two only if opts.git is set).'''
     gone = set()
 
     copyto = dict([(v, k) for k, v in copy.items()])
 
     addedset, removedset = set(added), set(removed)
 
     for f in sorted(modified + added + removed):
         copyop = None
         f1, f2 = f, f
         if f in addedset:
             f1 = None
             if f in copy:
                 if opts.git:
                     f1 = copy[f]
                     if f1 in removedset and f1 not in gone:
                         copyop = 'rename'
                         gone.add(f1)
                     else:
                         copyop = 'copy'
         elif f in removedset:
             f2 = None
             if opts.git:
                 # have we already reported a copy above?
                 if (f in copyto and copyto[f] in addedset
                     and copy[copyto[f]] == f):
                     continue
         yield f1, f2, copyop
 
 def trydiff(repo, revs, ctx1, ctx2, modified, added, removed,
             copy, getfilectx, opts, losedatafn, prefix, relroot):
     '''given input data, generate a diff and yield it in blocks
 
     If generating a diff would lose data like flags or binary data and
     losedatafn is not None, it will be called.
 
     relroot is removed and prefix is added to every path in the diff output.
 
     If relroot is not empty, this function expects every path in modified,
     added, removed and copy to start with it.'''
 
     def gitindex(text):
         if not text:
             text = ""
         l = len(text)
         s = hashlib.sha1('blob %d\0' % l)
         s.update(text)
         return s.hexdigest()
 
     if opts.noprefix:
         aprefix = bprefix = ''
     else:
         aprefix = 'a/'
         bprefix = 'b/'
 
     def diffline(f, revs):
         revinfo = ' '.join(["-r %s" % rev for rev in revs])
         return 'diff %s %s' % (revinfo, f)
 
     date1 = util.datestr(ctx1.date())
     date2 = util.datestr(ctx2.date())
 
     gitmode = {'l': '120000', 'x': '100755', '': '100644'}
 
     if relroot != '' and (repo.ui.configbool('devel', 'all')
                           or repo.ui.configbool('devel', 'check-relroot')):
         for f in modified + added + removed + copy.keys() + copy.values():
             if f is not None and not f.startswith(relroot):
                 raise AssertionError(
                     "file %s doesn't start with relroot %s" % (f, relroot))
 
     for f1, f2, copyop in _filepairs(modified, added, removed, copy, opts):
         content1 = None
         content2 = None
         flag1 = None
         flag2 = None
         if f1:
             content1 = getfilectx(f1, ctx1).data()
             if opts.git or losedatafn:
                 flag1 = ctx1.flags(f1)
         if f2:
             content2 = getfilectx(f2, ctx2).data()
             if opts.git or losedatafn:
                 flag2 = ctx2.flags(f2)
         binary = False
         if opts.git or losedatafn:
             binary = util.binary(content1) or util.binary(content2)
 
         if losedatafn and not opts.git:
             if (binary or
                 # copy/rename
                 f2 in copy or
                 # empty file creation
                 (not f1 and not content2) or
                 # empty file deletion
                 (not content1 and not f2) or
                 # create with flags
                 (not f1 and flag2) or
                 # change flags
                 (f1 and f2 and flag1 != flag2)):
                 losedatafn(f2 or f1)
 
         path1 = f1 or f2
         path2 = f2 or f1
         path1 = posixpath.join(prefix, path1[len(relroot):])
         path2 = posixpath.join(prefix, path2[len(relroot):])
         header = []
         if opts.git:
             header.append('diff --git %s%s %s%s' %
                           (aprefix, path1, bprefix, path2))
             if not f1: # added
                 header.append('new file mode %s' % gitmode[flag2])
             elif not f2: # removed
                 header.append('deleted file mode %s' % gitmode[flag1])
             else:  # modified/copied/renamed
                 mode1, mode2 = gitmode[flag1], gitmode[flag2]
                 if mode1 != mode2:
                     header.append('old mode %s' % mode1)
                     header.append('new mode %s' % mode2)
                 if copyop is not None:
                     if opts.showsimilarity:
                         sim = similar.score(ctx1[path1], ctx2[path2]) * 100
                         header.append('similarity index %d%%' % sim)
                     header.append('%s from %s' % (copyop, path1))
                     header.append('%s to %s' % (copyop, path2))
         elif revs and not repo.ui.quiet:
             header.append(diffline(path1, revs))
 
         if binary and opts.git and not opts.nobinary:
             text = mdiff.b85diff(content1, content2)
             if text:
                 header.append('index %s..%s' %
                               (gitindex(content1), gitindex(content2)))
         else:
             if opts.git and opts.index > 0:
                 flag = flag1
                 if flag is None:
                     flag = flag2
                 header.append('index %s..%s %s' %
                               (gitindex(content1)[0:opts.index],
                                gitindex(content2)[0:opts.index],
                                gitmode[flag]))
 
             text = mdiff.unidiff(content1, date1,
                                  content2, date2,
                                  path1, path2, opts=opts)
         if header and (text or len(header) > 1):
             yield '\n'.join(header) + '\n'
         if text:
             yield text
 
 def diffstatsum(stats):
     maxfile, maxtotal, addtotal, removetotal, binary = 0, 0, 0, 0, False
     for f, a, r, b in stats:
         maxfile = max(maxfile, encoding.colwidth(f))
         maxtotal = max(maxtotal, a + r)
         addtotal += a
         removetotal += r
         binary = binary or b
 
     return maxfile, maxtotal, addtotal, removetotal, binary
 
 def diffstatdata(lines):
     diffre = re.compile('^diff .*-r [a-z0-9]+\s(.*)$')
 
     results = []
     filename, adds, removes, isbinary = None, 0, 0, False
 
     def addresult():
         if filename:
             results.append((filename, adds, removes, isbinary))
 
     for line in lines:
         if line.startswith('diff'):
             addresult()
             # set numbers to 0 anyway when starting new file
             adds, removes, isbinary = 0, 0, False
             if line.startswith('diff --git a/'):
                 filename = gitre.search(line).group(2)
             elif line.startswith('diff -r'):
                 # format: "diff -r ... -r ... filename"
                 filename = diffre.search(line).group(1)
         elif line.startswith('+') and not line.startswith('+++ '):
             adds += 1
         elif line.startswith('-') and not line.startswith('--- '):
             removes += 1
         elif (line.startswith('GIT binary patch') or
               line.startswith('Binary file')):
             isbinary = True
     addresult()
     return results
 
 def diffstat(lines, width=80):
     output = []
     stats = diffstatdata(lines)
     maxname, maxtotal, totaladds, totalremoves, hasbinary = diffstatsum(stats)
 
     countwidth = len(str(maxtotal))
     if hasbinary and countwidth < 3:
         countwidth = 3
     graphwidth = width - countwidth - maxname - 6
     if graphwidth < 10:
         graphwidth = 10
 
     def scale(i):
         if maxtotal <= graphwidth:
             return i
         # If diffstat runs out of room it doesn't print anything,
         # which isn't very useful, so always print at least one + or -
         # if there were at least some changes.
         return max(i * graphwidth // maxtotal, int(bool(i)))
 
     for filename, adds, removes, isbinary in stats:
         if isbinary:
             count = 'Bin'
         else:
             count = adds + removes
         pluses = '+' * scale(adds)
         minuses = '-' * scale(removes)
         output.append(' %s%s |  %*s %s%s\n' %
                       (filename, ' ' * (maxname - encoding.colwidth(filename)),
                        countwidth, count, pluses, minuses))
 
     if stats:
         output.append(_(' %d files changed, %d insertions(+), '
                         '%d deletions(-)\n')
                       % (len(stats), totaladds, totalremoves))
 
     return ''.join(output)
 
 def diffstatui(*args, **kw):
     '''like diffstat(), but yields 2-tuples of (output, label) for
     ui.write()
     '''
 
     for line in diffstat(*args, **kw).splitlines():
         if line and line[-1] in '+-':
             name, graph = line.rsplit(' ', 1)
             yield (name + ' ', '')
             m = re.search(r'\++', graph)
             if m:
                 yield (m.group(0), 'diffstat.inserted')
             m = re.search(r'-+', graph)
             if m:
                 yield (m.group(0), 'diffstat.deleted')
         else:
             yield (line, '')
         yield ('\n', '')

mercurial/wireproto.py

@@ -1,1050 +1,1051 b''
 # wireproto.py - generic wire protocol support functions
 #
 # Copyright 2005-2010 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 hashlib
 import itertools
 import os
 import tempfile
 
 from .i18n import _
 from .node import (
     bin,
     hex,
 )
 
 from . import (
     bundle2,
     changegroup as changegroupmod,
     encoding,
     error,
     exchange,
     peer,
     pushkey as pushkeymod,
+    pycompat,
     streamclone,
     util,
 )
 
 urlerr = util.urlerr
 urlreq = util.urlreq
 
 bundle2requiredmain = _('incompatible Mercurial client; bundle2 required')
 bundle2requiredhint = _('see https://www.mercurial-scm.org/wiki/'
                         'IncompatibleClient')
 bundle2required = '%s\n(%s)\n' % (bundle2requiredmain, bundle2requiredhint)
 
 class abstractserverproto(object):
     """abstract class that summarizes the protocol API
 
     Used as reference and documentation.
     """
 
     def getargs(self, args):
         """return the value for arguments in <args>
 
         returns a list of values (same order as <args>)"""
         raise NotImplementedError()
 
     def getfile(self, fp):
         """write the whole content of a file into a file like object
 
         The file is in the form::
 
             (<chunk-size>\n<chunk>)+0\n
 
         chunk size is the ascii version of the int.
         """
         raise NotImplementedError()
 
     def redirect(self):
         """may setup interception for stdout and stderr
 
         See also the `restore` method."""
         raise NotImplementedError()
 
     # If the `redirect` function does install interception, the `restore`
     # function MUST be defined. If interception is not used, this function
     # MUST NOT be defined.
     #
     # left commented here on purpose
     #
     #def restore(self):
     #    """reinstall previous stdout and stderr and return intercepted stdout
     #    """
     #    raise NotImplementedError()
 
 class remotebatch(peer.batcher):
     '''batches the queued calls; uses as few roundtrips as possible'''
     def __init__(self, remote):
         '''remote must support _submitbatch(encbatch) and
         _submitone(op, encargs)'''
         peer.batcher.__init__(self)
         self.remote = remote
     def submit(self):
         req, rsp = [], []
         for name, args, opts, resref in self.calls:
             mtd = getattr(self.remote, name)
             batchablefn = getattr(mtd, 'batchable', None)
             if batchablefn is not None:
                 batchable = batchablefn(mtd.im_self, *args, **opts)
                 encargsorres, encresref = next(batchable)
                 if encresref:
                     req.append((name, encargsorres,))
                     rsp.append((batchable, encresref, resref,))
                 else:
                     resref.set(encargsorres)
             else:
                 if req:
                     self._submitreq(req, rsp)
                     req, rsp = [], []
                 resref.set(mtd(*args, **opts))
         if req:
             self._submitreq(req, rsp)
     def _submitreq(self, req, rsp):
         encresults = self.remote._submitbatch(req)
         for encres, r in zip(encresults, rsp):
             batchable, encresref, resref = r
             encresref.set(encres)
             resref.set(next(batchable))
 
 class remoteiterbatcher(peer.iterbatcher):
     def __init__(self, remote):
         super(remoteiterbatcher, self).__init__()
         self._remote = remote
 
     def __getattr__(self, name):
         if not getattr(self._remote, name, False):
             raise AttributeError(
                 'Attempted to iterbatch non-batchable call to %r' % name)
         return super(remoteiterbatcher, self).__getattr__(name)
 
     def submit(self):
         """Break the batch request into many patch calls and pipeline them.
 
         This is mostly valuable over http where request sizes can be
         limited, but can be used in other places as well.
         """
         req, rsp = [], []
         for name, args, opts, resref in self.calls:
             mtd = getattr(self._remote, name)
             batchable = mtd.batchable(mtd.im_self, *args, **opts)
             encargsorres, encresref = next(batchable)
             assert encresref
             req.append((name, encargsorres))
             rsp.append((batchable, encresref))
         if req:
             self._resultiter = self._remote._submitbatch(req)
         self._rsp = rsp
 
     def results(self):
         for (batchable, encresref), encres in itertools.izip(
                 self._rsp, self._resultiter):
             encresref.set(encres)
             yield next(batchable)
 
 # Forward a couple of names from peer to make wireproto interactions
 # slightly more sensible.
 batchable = peer.batchable
 future = peer.future
 
 # list of nodes encoding / decoding
 
 def decodelist(l, sep=' '):
     if l:
         return map(bin, l.split(sep))
     return []
 
 def encodelist(l, sep=' '):
     try:
         return sep.join(map(hex, l))
     except TypeError:
         raise
 
 # batched call argument encoding
 
 def escapearg(plain):
     return (plain
             .replace(':', ':c')
             .replace(',', ':o')
             .replace(';', ':s')
             .replace('=', ':e'))
 
 def unescapearg(escaped):
     return (escaped
             .replace(':e', '=')
             .replace(':s', ';')
             .replace(':o', ',')
             .replace(':c', ':'))
 
 def encodebatchcmds(req):
     """Return a ``cmds`` argument value for the ``batch`` command."""
     cmds = []
     for op, argsdict in req:
         # Old servers didn't properly unescape argument names. So prevent
         # the sending of argument names that may not be decoded properly by
         # servers.
         assert all(escapearg(k) == k for k in argsdict)
 
         args = ','.join('%s=%s' % (escapearg(k), escapearg(v))
                         for k, v in argsdict.iteritems())
         cmds.append('%s %s' % (op, args))
 
     return ';'.join(cmds)
 
 # mapping of options accepted by getbundle and their types
 #
 # Meant to be extended by extensions. It is extensions responsibility to ensure
 # such options are properly processed in exchange.getbundle.
 #
 # supported types are:
 #
 # :nodes: list of binary nodes
 # :csv:   list of comma-separated values
 # :scsv:  list of comma-separated values return as set
 # :plain: string with no transformation needed.
 gboptsmap = {'heads':  'nodes',
              'common': 'nodes',
              'obsmarkers': 'boolean',
              'bundlecaps': 'scsv',
              'listkeys': 'csv',
              'cg': 'boolean',
              'cbattempted': 'boolean'}
 
 # client side
 
 class wirepeer(peer.peerrepository):
     """Client-side interface for communicating with a peer repository.
 
     Methods commonly call wire protocol commands of the same name.
 
     See also httppeer.py and sshpeer.py for protocol-specific
     implementations of this interface.
     """
     def batch(self):
         if self.capable('batch'):
             return remotebatch(self)
         else:
             return peer.localbatch(self)
     def _submitbatch(self, req):
         """run batch request <req> on the server
 
         Returns an iterator of the raw responses from the server.
         """
         rsp = self._callstream("batch", cmds=encodebatchcmds(req))
         chunk = rsp.read(1024)
         work = [chunk]
         while chunk:
             while ';' not in chunk and chunk:
                 chunk = rsp.read(1024)
                 work.append(chunk)
             merged = ''.join(work)
             while ';' in merged:
                 one, merged = merged.split(';', 1)
                 yield unescapearg(one)
             chunk = rsp.read(1024)
             work = [merged, chunk]
         yield unescapearg(''.join(work))
 
     def _submitone(self, op, args):
         return self._call(op, **args)
 
     def iterbatch(self):
         return remoteiterbatcher(self)
 
     @batchable
     def lookup(self, key):
         self.requirecap('lookup', _('look up remote revision'))
         f = future()
         yield {'key': encoding.fromlocal(key)}, f
         d = f.value
         success, data = d[:-1].split(" ", 1)
         if int(success):
             yield bin(data)
         self._abort(error.RepoError(data))
 
     @batchable
     def heads(self):
         f = future()
         yield {}, f
         d = f.value
         try:
             yield decodelist(d[:-1])
         except ValueError:
             self._abort(error.ResponseError(_("unexpected response:"), d))
 
     @batchable
     def known(self, nodes):
         f = future()
         yield {'nodes': encodelist(nodes)}, f
         d = f.value
         try:
             yield [bool(int(b)) for b in d]
         except ValueError:
             self._abort(error.ResponseError(_("unexpected response:"), d))
 
     @batchable
     def branchmap(self):
         f = future()
         yield {}, f
         d = f.value
         try:
             branchmap = {}
             for branchpart in d.splitlines():
                 branchname, branchheads = branchpart.split(' ', 1)
                 branchname = encoding.tolocal(urlreq.unquote(branchname))
                 branchheads = decodelist(branchheads)
                 branchmap[branchname] = branchheads
             yield branchmap
         except TypeError:
             self._abort(error.ResponseError(_("unexpected response:"), d))
 
     def branches(self, nodes):
         n = encodelist(nodes)
         d = self._call("branches", nodes=n)
         try:
             br = [tuple(decodelist(b)) for b in d.splitlines()]
             return br
         except ValueError:
             self._abort(error.ResponseError(_("unexpected response:"), d))
 
     def between(self, pairs):
         batch = 8 # avoid giant requests
         r = []
         for i in xrange(0, len(pairs), batch):
             n = " ".join([encodelist(p, '-') for p in pairs[i:i + batch]])
             d = self._call("between", pairs=n)
             try:
                 r.extend(l and decodelist(l) or [] for l in d.splitlines())
             except ValueError:
                 self._abort(error.ResponseError(_("unexpected response:"), d))
         return r
 
     @batchable
     def pushkey(self, namespace, key, old, new):
         if not self.capable('pushkey'):
             yield False, None
         f = future()
         self.ui.debug('preparing pushkey for "%s:%s"\n' % (namespace, key))
         yield {'namespace': encoding.fromlocal(namespace),
                'key': encoding.fromlocal(key),
                'old': encoding.fromlocal(old),
                'new': encoding.fromlocal(new)}, f
         d = f.value
         d, output = d.split('\n', 1)
         try:
             d = bool(int(d))
         except ValueError:
             raise error.ResponseError(
                 _('push failed (unexpected response):'), d)
         for l in output.splitlines(True):
             self.ui.status(_('remote: '), l)
         yield d
 
     @batchable
     def listkeys(self, namespace):
         if not self.capable('pushkey'):
             yield {}, None
         f = future()
         self.ui.debug('preparing listkeys for "%s"\n' % namespace)
         yield {'namespace': encoding.fromlocal(namespace)}, f
         d = f.value
         self.ui.debug('received listkey for "%s": %i bytes\n'
                       % (namespace, len(d)))
         yield pushkeymod.decodekeys(d)
 
     def stream_out(self):
         return self._callstream('stream_out')
 
     def changegroup(self, nodes, kind):
         n = encodelist(nodes)
         f = self._callcompressable("changegroup", roots=n)
         return changegroupmod.cg1unpacker(f, 'UN')
 
     def changegroupsubset(self, bases, heads, kind):
         self.requirecap('changegroupsubset', _('look up remote changes'))
         bases = encodelist(bases)
         heads = encodelist(heads)
         f = self._callcompressable("changegroupsubset",
                                    bases=bases, heads=heads)
         return changegroupmod.cg1unpacker(f, 'UN')
 
     def getbundle(self, source, **kwargs):
         self.requirecap('getbundle', _('look up remote changes'))
         opts = {}
         bundlecaps = kwargs.get('bundlecaps')
         if bundlecaps is not None:
             kwargs['bundlecaps'] = sorted(bundlecaps)
         else:
             bundlecaps = () # kwargs could have it to None
         for key, value in kwargs.iteritems():
             if value is None:
                 continue
             keytype = gboptsmap.get(key)
             if keytype is None:
                 assert False, 'unexpected'
             elif keytype == 'nodes':
                 value = encodelist(value)
             elif keytype in ('csv', 'scsv'):
                 value = ','.join(value)
             elif keytype == 'boolean':
                 value = '%i' % bool(value)
             elif keytype != 'plain':
                 raise KeyError('unknown getbundle option type %s'
                                % keytype)
             opts[key] = value
         f = self._callcompressable("getbundle", **opts)
         if any((cap.startswith('HG2') for cap in bundlecaps)):
             return bundle2.getunbundler(self.ui, f)
         else:
             return changegroupmod.cg1unpacker(f, 'UN')
 
     def unbundle(self, cg, heads, url):
         '''Send cg (a readable file-like object representing the
         changegroup to push, typically a chunkbuffer object) to the
         remote server as a bundle.
 
         When pushing a bundle10 stream, return an integer indicating the
         result of the push (see localrepository.addchangegroup()).
 
         When pushing a bundle20 stream, return a bundle20 stream.
 
         `url` is the url the client thinks it's pushing to, which is
         visible to hooks.
         '''
 
         if heads != ['force'] and self.capable('unbundlehash'):
             heads = encodelist(['hashed',
                                 hashlib.sha1(''.join(sorted(heads))).digest()])
         else:
             heads = encodelist(heads)
 
         if util.safehasattr(cg, 'deltaheader'):
             # this a bundle10, do the old style call sequence
             ret, output = self._callpush("unbundle", cg, heads=heads)
             if ret == "":
                 raise error.ResponseError(
                     _('push failed:'), output)
             try:
                 ret = int(ret)
             except ValueError:
                 raise error.ResponseError(
                     _('push failed (unexpected response):'), ret)
 
             for l in output.splitlines(True):
                 self.ui.status(_('remote: '), l)
         else:
             # bundle2 push. Send a stream, fetch a stream.
             stream = self._calltwowaystream('unbundle', cg, heads=heads)
             ret = bundle2.getunbundler(self.ui, stream)
         return ret
 
     def debugwireargs(self, one, two, three=None, four=None, five=None):
         # don't pass optional arguments left at their default value
         opts = {}
         if three is not None:
             opts['three'] = three
         if four is not None:
             opts['four'] = four
         return self._call('debugwireargs', one=one, two=two, **opts)
 
     def _call(self, cmd, **args):
         """execute <cmd> on the server
 
         The command is expected to return a simple string.
 
         returns the server reply as a string."""
         raise NotImplementedError()
 
     def _callstream(self, cmd, **args):
         """execute <cmd> on the server
 
         The command is expected to return a stream. Note that if the
         command doesn't return a stream, _callstream behaves
         differently for ssh and http peers.
 
         returns the server reply as a file like object.
         """
         raise NotImplementedError()
 
     def _callcompressable(self, cmd, **args):
         """execute <cmd> on the server
 
         The command is expected to return a stream.
 
         The stream may have been compressed in some implementations. This
         function takes care of the decompression. This is the only difference
         with _callstream.
 
         returns the server reply as a file like object.
         """
         raise NotImplementedError()
 
     def _callpush(self, cmd, fp, **args):
         """execute a <cmd> on server
 
         The command is expected to be related to a push. Push has a special
         return method.
 
         returns the server reply as a (ret, output) tuple. ret is either
         empty (error) or a stringified int.
         """
         raise NotImplementedError()
 
     def _calltwowaystream(self, cmd, fp, **args):
         """execute <cmd> on server
 
         The command will send a stream to the server and get a stream in reply.
         """
         raise NotImplementedError()
 
     def _abort(self, exception):
         """clearly abort the wire protocol connection and raise the exception
         """
         raise NotImplementedError()
 
 # server side
 
 # wire protocol command can either return a string or one of these classes.
 class streamres(object):
     """wireproto reply: binary stream
 
     The call was successful and the result is a stream.
 
     Accepts either a generator or an object with a ``read(size)`` method.
 
     ``v1compressible`` indicates whether this data can be compressed to
     "version 1" clients (technically: HTTP peers using
     application/mercurial-0.1 media type). This flag should NOT be used on
     new commands because new clients should support a more modern compression
     mechanism.
     """
     def __init__(self, gen=None, reader=None, v1compressible=False):
         self.gen = gen
         self.reader = reader
         self.v1compressible = v1compressible
 
 class pushres(object):
     """wireproto reply: success with simple integer return
 
     The call was successful and returned an integer contained in `self.res`.
     """
     def __init__(self, res):
         self.res = res
 
 class pusherr(object):
     """wireproto reply: failure
 
     The call failed. The `self.res` attribute contains the error message.
     """
     def __init__(self, res):
         self.res = res
 
 class ooberror(object):
     """wireproto reply: failure of a batch of operation
 
     Something failed during a batch call. The error message is stored in
     `self.message`.
     """
     def __init__(self, message):
         self.message = message
 
 def getdispatchrepo(repo, proto, command):
     """Obtain the repo used for processing wire protocol commands.
 
     The intent of this function is to serve as a monkeypatch point for
     extensions that need commands to operate on different repo views under
     specialized circumstances.
     """
     return repo.filtered('served')
 
 def dispatch(repo, proto, command):
     repo = getdispatchrepo(repo, proto, command)
     func, spec = commands[command]
     args = proto.getargs(spec)
     return func(repo, proto, *args)
 
 def options(cmd, keys, others):
     opts = {}
     for k in keys:
         if k in others:
             opts[k] = others[k]
             del others[k]
     if others:
         util.stderr.write("warning: %s ignored unexpected arguments %s\n"
                           % (cmd, ",".join(others)))
     return opts
 
 def bundle1allowed(repo, action):
     """Whether a bundle1 operation is allowed from the server.
 
     Priority is:
 
     1. server.bundle1gd.<action> (if generaldelta active)
     2. server.bundle1.<action>
     3. server.bundle1gd (if generaldelta active)
     4. server.bundle1
     """
     ui = repo.ui
     gd = 'generaldelta' in repo.requirements
 
     if gd:
         v = ui.configbool('server', 'bundle1gd.%s' % action, None)
         if v is not None:
             return v
 
     v = ui.configbool('server', 'bundle1.%s' % action, None)
     if v is not None:
         return v
 
     if gd:
         v = ui.configbool('server', 'bundle1gd', None)
         if v is not None:
             return v
 
     return ui.configbool('server', 'bundle1', True)
 
 def supportedcompengines(ui, proto, role):
     """Obtain the list of supported compression engines for a request."""
     assert role in (util.CLIENTROLE, util.SERVERROLE)
 
     compengines = util.compengines.supportedwireengines(role)
 
     # Allow config to override default list and ordering.
     if role == util.SERVERROLE:
         configengines = ui.configlist('server', 'compressionengines')
         config = 'server.compressionengines'
     else:
         # This is currently implemented mainly to facilitate testing. In most
         # cases, the server should be in charge of choosing a compression engine
         # because a server has the most to lose from a sub-optimal choice. (e.g.
         # CPU DoS due to an expensive engine or a network DoS due to poor
         # compression ratio).
         configengines = ui.configlist('experimental',
                                       'clientcompressionengines')
         config = 'experimental.clientcompressionengines'
 
     # No explicit config. Filter out the ones that aren't supposed to be
     # advertised and return default ordering.
     if not configengines:
         attr = 'serverpriority' if role == util.SERVERROLE else 'clientpriority'
         return [e for e in compengines
                 if getattr(e.wireprotosupport(), attr) > 0]
 
     # If compression engines are listed in the config, assume there is a good
     # reason for it (like server operators wanting to achieve specific
     # performance characteristics). So fail fast if the config references
     # unusable compression engines.
     validnames = set(e.name() for e in compengines)
     invalidnames = set(e for e in configengines if e not in validnames)
     if invalidnames:
         raise error.Abort(_('invalid compression engine defined in %s: %s') %
                           (config, ', '.join(sorted(invalidnames))))
 
     compengines = [e for e in compengines if e.name() in configengines]
     compengines = sorted(compengines,
                          key=lambda e: configengines.index(e.name()))
 
     if not compengines:
         raise error.Abort(_('%s config option does not specify any known '
                             'compression engines') % config,
                           hint=_('usable compression engines: %s') %
                           ', '.sorted(validnames))
 
     return compengines
 
 # list of commands
 commands = {}
 
 def wireprotocommand(name, args=''):
     """decorator for wire protocol command"""
     def register(func):
         commands[name] = (func, args)
         return func
     return register
 
 @wireprotocommand('batch', 'cmds *')
 def batch(repo, proto, cmds, others):
     repo = repo.filtered("served")
     res = []
     for pair in cmds.split(';'):
         op, args = pair.split(' ', 1)
         vals = {}
         for a in args.split(','):
             if a:
                 n, v = a.split('=')
                 vals[unescapearg(n)] = unescapearg(v)
         func, spec = commands[op]
         if spec:
             keys = spec.split()
             data = {}
             for k in keys:
                 if k == '*':
                     star = {}
                     for key in vals.keys():
                         if key not in keys:
                             star[key] = vals[key]
                     data['*'] = star
                 else:
                     data[k] = vals[k]
             result = func(repo, proto, *[data[k] for k in keys])
         else:
             result = func(repo, proto)
         if isinstance(result, ooberror):
             return result
         res.append(escapearg(result))
     return ';'.join(res)
 
 @wireprotocommand('between', 'pairs')
 def between(repo, proto, pairs):
     pairs = [decodelist(p, '-') for p in pairs.split(" ")]
     r = []
     for b in repo.between(pairs):
         r.append(encodelist(b) + "\n")
     return "".join(r)
 
 @wireprotocommand('branchmap')
 def branchmap(repo, proto):
     branchmap = repo.branchmap()
     heads = []
     for branch, nodes in branchmap.iteritems():
         branchname = urlreq.quote(encoding.fromlocal(branch))
         branchnodes = encodelist(nodes)
         heads.append('%s %s' % (branchname, branchnodes))
     return '\n'.join(heads)
 
 @wireprotocommand('branches', 'nodes')
 def branches(repo, proto, nodes):
     nodes = decodelist(nodes)
     r = []
     for b in repo.branches(nodes):
         r.append(encodelist(b) + "\n")
     return "".join(r)
 
 @wireprotocommand('clonebundles', '')
 def clonebundles(repo, proto):
     """Server command for returning info for available bundles to seed clones.
 
     Clients will parse this response and determine what bundle to fetch.
 
     Extensions may wrap this command to filter or dynamically emit data
     depending on the request. e.g. you could advertise URLs for the closest
     data center given the client's IP address.
     """
     return repo.opener.tryread('clonebundles.manifest')
 
 wireprotocaps = ['lookup', 'changegroupsubset', 'branchmap', 'pushkey',
                  'known', 'getbundle', 'unbundlehash', 'batch']
 
 def _capabilities(repo, proto):
     """return a list of capabilities for a repo
 
     This function exists to allow extensions to easily wrap capabilities
     computation
 
     - returns a lists: easy to alter
     - change done here will be propagated to both `capabilities` and `hello`
       command without any other action needed.
     """
     # copy to prevent modification of the global list
     caps = list(wireprotocaps)
     if streamclone.allowservergeneration(repo.ui):
         if repo.ui.configbool('server', 'preferuncompressed', False):
             caps.append('stream-preferred')
         requiredformats = repo.requirements & repo.supportedformats
         # if our local revlogs are just revlogv1, add 'stream' cap
         if not requiredformats - set(('revlogv1',)):
             caps.append('stream')
         # otherwise, add 'streamreqs' detailing our local revlog format
         else:
             caps.append('streamreqs=%s' % ','.join(sorted(requiredformats)))
     if repo.ui.configbool('experimental', 'bundle2-advertise', True):
         capsblob = bundle2.encodecaps(bundle2.getrepocaps(repo))
         caps.append('bundle2=' + urlreq.quote(capsblob))
     caps.append('unbundle=%s' % ','.join(bundle2.bundlepriority))
 
     if proto.name == 'http':
         caps.append('httpheader=%d' %
                     repo.ui.configint('server', 'maxhttpheaderlen', 1024))
         if repo.ui.configbool('experimental', 'httppostargs', False):
             caps.append('httppostargs')
 
         # FUTURE advertise 0.2rx once support is implemented
         # FUTURE advertise minrx and mintx after consulting config option
         caps.append('httpmediatype=0.1rx,0.1tx,0.2tx')
 
         compengines = supportedcompengines(repo.ui, proto, util.SERVERROLE)
         if compengines:
             comptypes = ','.join(urlreq.quote(e.wireprotosupport().name)
                                  for e in compengines)
             caps.append('compression=%s' % comptypes)
 
     return caps
 
 # If you are writing an extension and consider wrapping this function. Wrap
 # `_capabilities` instead.
 @wireprotocommand('capabilities')
 def capabilities(repo, proto):
     return ' '.join(_capabilities(repo, proto))
 
 @wireprotocommand('changegroup', 'roots')
 def changegroup(repo, proto, roots):
     nodes = decodelist(roots)
     cg = changegroupmod.changegroup(repo, nodes, 'serve')
     return streamres(reader=cg, v1compressible=True)
 
 @wireprotocommand('changegroupsubset', 'bases heads')
 def changegroupsubset(repo, proto, bases, heads):
     bases = decodelist(bases)
     heads = decodelist(heads)
     cg = changegroupmod.changegroupsubset(repo, bases, heads, 'serve')
     return streamres(reader=cg, v1compressible=True)
 
 @wireprotocommand('debugwireargs', 'one two *')
 def debugwireargs(repo, proto, one, two, others):
     # only accept optional args from the known set
     opts = options('debugwireargs', ['three', 'four'], others)
     return repo.debugwireargs(one, two, **opts)
 
 @wireprotocommand('getbundle', '*')
 def getbundle(repo, proto, others):
     opts = options('getbundle', gboptsmap.keys(), others)
     for k, v in opts.iteritems():
         keytype = gboptsmap[k]
         if keytype == 'nodes':
             opts[k] = decodelist(v)
         elif keytype == 'csv':
             opts[k] = list(v.split(','))
         elif keytype == 'scsv':
             opts[k] = set(v.split(','))
         elif keytype == 'boolean':
             # Client should serialize False as '0', which is a non-empty string
             # so it evaluates as a True bool.
             if v == '0':
                 opts[k] = False
             else:
                 opts[k] = bool(v)
         elif keytype != 'plain':
             raise KeyError('unknown getbundle option type %s'
                            % keytype)
 
     if not bundle1allowed(repo, 'pull'):
         if not exchange.bundle2requested(opts.get('bundlecaps')):
             if proto.name == 'http':
                 return ooberror(bundle2required)
             raise error.Abort(bundle2requiredmain,
                               hint=bundle2requiredhint)
 
     #chunks = exchange.getbundlechunks(repo, 'serve', **opts)
     try:
         chunks = exchange.getbundlechunks(repo, 'serve', **opts)
     except error.Abort as exc:
         # cleanly forward Abort error to the client
         if not exchange.bundle2requested(opts.get('bundlecaps')):
             if proto.name == 'http':
                 return ooberror(str(exc) + '\n')
             raise # cannot do better for bundle1 + ssh
         # bundle2 request expect a bundle2 reply
         bundler = bundle2.bundle20(repo.ui)
         manargs = [('message', str(exc))]
         advargs = []
         if exc.hint is not None:
             advargs.append(('hint', exc.hint))
         bundler.addpart(bundle2.bundlepart('error:abort',
                                            manargs, advargs))
         return streamres(gen=bundler.getchunks(), v1compressible=True)
     return streamres(gen=chunks, v1compressible=True)
 
 @wireprotocommand('heads')
 def heads(repo, proto):
     h = repo.heads()
     return encodelist(h) + "\n"
 
 @wireprotocommand('hello')
 def hello(repo, proto):
     '''the hello command returns a set of lines describing various
     interesting things about the server, in an RFC822-like format.
     Currently the only one defined is "capabilities", which
     consists of a line in the form:
 
     capabilities: space separated list of tokens
     '''
     return "capabilities: %s\n" % (capabilities(repo, proto))
 
 @wireprotocommand('listkeys', 'namespace')
 def listkeys(repo, proto, namespace):
     d = repo.listkeys(encoding.tolocal(namespace)).items()
     return pushkeymod.encodekeys(d)
 
 @wireprotocommand('lookup', 'key')
 def lookup(repo, proto, key):
     try:
         k = encoding.tolocal(key)
         c = repo[k]
         r = c.hex()
         success = 1
     except Exception as inst:
         r = str(inst)
         success = 0
     return "%s %s\n" % (success, r)
 
 @wireprotocommand('known', 'nodes *')
 def known(repo, proto, nodes, others):
     return ''.join(b and "1" or "0" for b in repo.known(decodelist(nodes)))
 
 @wireprotocommand('pushkey', 'namespace key old new')
 def pushkey(repo, proto, namespace, key, old, new):
     # compatibility with pre-1.8 clients which were accidentally
     # sending raw binary nodes rather than utf-8-encoded hex
     if len(new) == 20 and new.encode('string-escape') != new:
         # looks like it could be a binary node
         try:
             new.decode('utf-8')
             new = encoding.tolocal(new) # but cleanly decodes as UTF-8
         except UnicodeDecodeError:
             pass # binary, leave unmodified
     else:
         new = encoding.tolocal(new) # normal path
 
     if util.safehasattr(proto, 'restore'):
 
         proto.redirect()
 
         try:
             r = repo.pushkey(encoding.tolocal(namespace), encoding.tolocal(key),
                              encoding.tolocal(old), new) or False
         except error.Abort:
             r = False
 
         output = proto.restore()
 
         return '%s\n%s' % (int(r), output)
 
     r = repo.pushkey(encoding.tolocal(namespace), encoding.tolocal(key),
                      encoding.tolocal(old), new)
     return '%s\n' % int(r)
 
 @wireprotocommand('stream_out')
 def stream(repo, proto):
     '''If the server supports streaming clone, it advertises the "stream"
     capability with a value representing the version and flags of the repo
     it is serving. Client checks to see if it understands the format.
     '''
     if not streamclone.allowservergeneration(repo.ui):
         return '1\n'
 
     def getstream(it):
         yield '0\n'
         for chunk in it:
             yield chunk
 
     try:
         # LockError may be raised before the first result is yielded. Don't
         # emit output until we're sure we got the lock successfully.
         it = streamclone.generatev1wireproto(repo)
         return streamres(gen=getstream(it))
     except error.LockError:
         return '2\n'
 
 @wireprotocommand('unbundle', 'heads')
 def unbundle(repo, proto, heads):
     their_heads = decodelist(heads)
 
     try:
         proto.redirect()
 
         exchange.check_heads(repo, their_heads, 'preparing changes')
 
         # write bundle data to temporary file because it can be big
         fd, tempname = tempfile.mkstemp(prefix='hg-unbundle-')
-        fp = os.fdopen(fd, 'wb+')
+        fp = os.fdopen(fd, pycompat.sysstr('wb+'))
         r = 0
         try:
             proto.getfile(fp)
             fp.seek(0)
             gen = exchange.readbundle(repo.ui, fp, None)
             if (isinstance(gen, changegroupmod.cg1unpacker)
                 and not bundle1allowed(repo, 'push')):
                 if proto.name == 'http':
                     # need to special case http because stderr do not get to
                     # the http client on failed push so we need to abuse some
                     # other error type to make sure the message get to the
                     # user.
                     return ooberror(bundle2required)
                 raise error.Abort(bundle2requiredmain,
                                   hint=bundle2requiredhint)
 
             r = exchange.unbundle(repo, gen, their_heads, 'serve',
                                   proto._client())
             if util.safehasattr(r, 'addpart'):
                 # The return looks streamable, we are in the bundle2 case and
                 # should return a stream.
                 return streamres(gen=r.getchunks())
             return pushres(r)
 
         finally:
             fp.close()
             os.unlink(tempname)
 
     except (error.BundleValueError, error.Abort, error.PushRaced) as exc:
         # handle non-bundle2 case first
         if not getattr(exc, 'duringunbundle2', False):
             try:
                 raise
             except error.Abort:
                 # The old code we moved used util.stderr directly.
                 # We did not change it to minimise code change.
                 # This need to be moved to something proper.
                 # Feel free to do it.
                 util.stderr.write("abort: %s\n" % exc)
                 if exc.hint is not None:
                     util.stderr.write("(%s)\n" % exc.hint)
                 return pushres(0)
             except error.PushRaced:
                 return pusherr(str(exc))
 
         bundler = bundle2.bundle20(repo.ui)
         for out in getattr(exc, '_bundle2salvagedoutput', ()):
             bundler.addpart(out)
         try:
             try:
                 raise
             except error.PushkeyFailed as exc:
                 # check client caps
                 remotecaps = getattr(exc, '_replycaps', None)
                 if (remotecaps is not None
                         and 'pushkey' not in remotecaps.get('error', ())):
                     # no support remote side, fallback to Abort handler.
                     raise
                 part = bundler.newpart('error:pushkey')
                 part.addparam('in-reply-to', exc.partid)
                 if exc.namespace is not None:
                     part.addparam('namespace', exc.namespace, mandatory=False)
                 if exc.key is not None:
                     part.addparam('key', exc.key, mandatory=False)
                 if exc.new is not None:
                     part.addparam('new', exc.new, mandatory=False)
                 if exc.old is not None:
                     part.addparam('old', exc.old, mandatory=False)
                 if exc.ret is not None:
                     part.addparam('ret', exc.ret, mandatory=False)
         except error.BundleValueError as exc:
             errpart = bundler.newpart('error:unsupportedcontent')
             if exc.parttype is not None:
                 errpart.addparam('parttype', exc.parttype)
             if exc.params:
                 errpart.addparam('params', '\0'.join(exc.params))
         except error.Abort as exc:
             manargs = [('message', str(exc))]
             advargs = []
             if exc.hint is not None:
                 advargs.append(('hint', exc.hint))
             bundler.addpart(bundle2.bundlepart('error:abort',
                                                manargs, advargs))
         except error.PushRaced as exc:
             bundler.newpart('error:pushraced', [('message', str(exc))])
         return streamres(gen=bundler.getchunks())

mercurial/worker.py

@@ -1,224 +1,224 b''
 # worker.py - master-slave parallelism support
 #
 # Copyright 2013 Facebook, Inc.
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import errno
 import os
 import signal
 import sys
 
 from .i18n import _
 from . import (
     encoding,
     error,
     pycompat,
     scmutil,
     util,
 )
 
 def countcpus():
     '''try to count the number of CPUs on the system'''
 
     # posix
     try:
         n = int(os.sysconf('SC_NPROCESSORS_ONLN'))
         if n > 0:
             return n
     except (AttributeError, ValueError):
         pass
 
     # windows
     try:
         n = int(encoding.environ['NUMBER_OF_PROCESSORS'])
         if n > 0:
             return n
     except (KeyError, ValueError):
         pass
 
     return 1
 
 def _numworkers(ui):
     s = ui.config('worker', 'numcpus')
     if s:
         try:
             n = int(s)
             if n >= 1:
                 return n
         except ValueError:
             raise error.Abort(_('number of cpus must be an integer'))
     return min(max(countcpus(), 4), 32)
 
 if pycompat.osname == 'posix':
     _startupcost = 0.01
 else:
     _startupcost = 1e30
 
 def worthwhile(ui, costperop, nops):
     '''try to determine whether the benefit of multiple processes can
     outweigh the cost of starting them'''
     linear = costperop * nops
     workers = _numworkers(ui)
     benefit = linear - (_startupcost * workers + linear / workers)
     return benefit >= 0.15
 
 def worker(ui, costperarg, func, staticargs, args):
     '''run a function, possibly in parallel in multiple worker
     processes.
 
     returns a progress iterator
 
     costperarg - cost of a single task
 
     func - function to run
 
     staticargs - arguments to pass to every invocation of the function
 
     args - arguments to split into chunks, to pass to individual
     workers
     '''
     if worthwhile(ui, costperarg, len(args)):
         return _platformworker(ui, func, staticargs, args)
     return func(*staticargs + (args,))
 
 def _posixworker(ui, func, staticargs, args):
     rfd, wfd = os.pipe()
     workers = _numworkers(ui)
     oldhandler = signal.getsignal(signal.SIGINT)
     signal.signal(signal.SIGINT, signal.SIG_IGN)
     pids, problem = set(), [0]
     def killworkers():
         # unregister SIGCHLD handler as all children will be killed. This
         # function shouldn't be interrupted by another SIGCHLD; otherwise pids
         # could be updated while iterating, which would cause inconsistency.
         signal.signal(signal.SIGCHLD, oldchldhandler)
         # if one worker bails, there's no good reason to wait for the rest
         for p in pids:
             try:
                 os.kill(p, signal.SIGTERM)
             except OSError as err:
                 if err.errno != errno.ESRCH:
                     raise
     def waitforworkers(blocking=True):
         for pid in pids.copy():
             p = st = 0
             while True:
                 try:
                     p, st = os.waitpid(pid, (0 if blocking else os.WNOHANG))
                     break
                 except OSError as e:
                     if e.errno == errno.EINTR:
                         continue
                     elif e.errno == errno.ECHILD:
                         # child would already be reaped, but pids yet been
                         # updated (maybe interrupted just after waitpid)
                         pids.discard(pid)
                         break
                     else:
                         raise
             if p:
                 pids.discard(p)
                 st = _exitstatus(st)
             if st and not problem[0]:
                 problem[0] = st
     def sigchldhandler(signum, frame):
         waitforworkers(blocking=False)
         if problem[0]:
             killworkers()
     oldchldhandler = signal.signal(signal.SIGCHLD, sigchldhandler)
     for pargs in partition(args, workers):
         pid = os.fork()
         if pid == 0:
             signal.signal(signal.SIGINT, oldhandler)
             signal.signal(signal.SIGCHLD, oldchldhandler)
 
             def workerfunc():
                 os.close(rfd)
                 for i, item in func(*(staticargs + (pargs,))):
                     os.write(wfd, '%d %s\n' % (i, item))
 
             # make sure we use os._exit in all code paths. otherwise the worker
             # may do some clean-ups which could cause surprises like deadlock.
             # see sshpeer.cleanup for example.
             try:
                 scmutil.callcatch(ui, workerfunc)
             except KeyboardInterrupt:
                 os._exit(255)
             except: # never return, therefore no re-raises
                 try:
                     ui.traceback()
                 finally:
                     os._exit(255)
             else:
                 os._exit(0)
         pids.add(pid)
     os.close(wfd)
-    fp = os.fdopen(rfd, 'rb', 0)
+    fp = os.fdopen(rfd, pycompat.sysstr('rb'), 0)
     def cleanup():
         signal.signal(signal.SIGINT, oldhandler)
         waitforworkers()
         signal.signal(signal.SIGCHLD, oldchldhandler)
         status = problem[0]
         if status:
             if status < 0:
                 os.kill(os.getpid(), -status)
             sys.exit(status)
     try:
         for line in util.iterfile(fp):
             l = line.split(' ', 1)
             yield int(l[0]), l[1][:-1]
     except: # re-raises
         killworkers()
         cleanup()
         raise
     cleanup()
 
 def _posixexitstatus(code):
     '''convert a posix exit status into the same form returned by
     os.spawnv
 
     returns None if the process was stopped instead of exiting'''
     if os.WIFEXITED(code):
         return os.WEXITSTATUS(code)
     elif os.WIFSIGNALED(code):
         return -os.WTERMSIG(code)
 
 if pycompat.osname != 'nt':
     _platformworker = _posixworker
     _exitstatus = _posixexitstatus
 
 def partition(lst, nslices):
     '''partition a list into N slices of roughly equal size
 
     The current strategy takes every Nth element from the input. If
     we ever write workers that need to preserve grouping in input
     we should consider allowing callers to specify a partition strategy.
 
     mpm is not a fan of this partitioning strategy when files are involved.
     In his words:
 
         Single-threaded Mercurial makes a point of creating and visiting
         files in a fixed order (alphabetical). When creating files in order,
         a typical filesystem is likely to allocate them on nearby regions on
         disk. Thus, when revisiting in the same order, locality is maximized
         and various forms of OS and disk-level caching and read-ahead get a
         chance to work.
 
         This effect can be quite significant on spinning disks. I discovered it
         circa Mercurial v0.4 when revlogs were named by hashes of filenames.
         Tarring a repo and copying it to another disk effectively randomized
         the revlog ordering on disk by sorting the revlogs by hash and suddenly
         performance of my kernel checkout benchmark dropped by ~10x because the
         "working set" of sectors visited no longer fit in the drive's cache and
         the workload switched from streaming to random I/O.
 
         What we should really be doing is have workers read filenames from a
         ordered queue. This preserves locality and also keeps any worker from
         getting more than one file out of balance.
     '''
     for i in range(nslices):
         yield lst[i::nslices]