subrepos: be smarter about what's an absolute path (issue2808)

mercurial/subrepo.py

@@ -1,1085 +1,1085 b''
 # subrepo.py - sub-repository handling for Mercurial
 #
 # Copyright 2009-2010 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 import errno, os, re, xml.dom.minidom, shutil, posixpath
 import stat, subprocess, tarfile
 from i18n import _
 import config, scmutil, util, node, error, cmdutil, bookmarks
 hg = None
 propertycache = util.propertycache
 
 nullstate = ('', '', 'empty')
 
 def state(ctx, ui):
     """return a state dict, mapping subrepo paths configured in .hgsub
     to tuple: (source from .hgsub, revision from .hgsubstate, kind
     (key in types dict))
     """
     p = config.config()
     def read(f, sections=None, remap=None):
         if f in ctx:
             try:
                 data = ctx[f].data()
             except IOError, err:
                 if err.errno != errno.ENOENT:
                     raise
                 # handle missing subrepo spec files as removed
                 ui.warn(_("warning: subrepo spec file %s not found\n") % f)
                 return
             p.parse(f, data, sections, remap, read)
         else:
             raise util.Abort(_("subrepo spec file %s not found") % f)
 
     if '.hgsub' in ctx:
         read('.hgsub')
 
     for path, src in ui.configitems('subpaths'):
         p.set('subpaths', path, src, ui.configsource('subpaths', path))
 
     rev = {}
     if '.hgsubstate' in ctx:
         try:
             for l in ctx['.hgsubstate'].data().splitlines():
                 revision, path = l.split(" ", 1)
                 rev[path] = revision
         except IOError, err:
             if err.errno != errno.ENOENT:
                 raise
 
     state = {}
     for path, src in p[''].items():
         kind = 'hg'
         if src.startswith('['):
             if ']' not in src:
                 raise util.Abort(_('missing ] in subrepo source'))
             kind, src = src.split(']', 1)
             kind = kind[1:]
 
         for pattern, repl in p.items('subpaths'):
             # Turn r'C:\foo\bar' into r'C:\\foo\\bar' since re.sub
             # does a string decode.
             repl = repl.encode('string-escape')
             # However, we still want to allow back references to go
             # through unharmed, so we turn r'\\1' into r'\1'. Again,
             # extra escapes are needed because re.sub string decodes.
             repl = re.sub(r'\\\\([0-9]+)', r'\\\1', repl)
             try:
                 src = re.sub(pattern, repl, src, 1)
             except re.error, e:
                 raise util.Abort(_("bad subrepository pattern in %s: %s")
                                  % (p.source('subpaths', pattern), e))
 
         state[path] = (src.strip(), rev.get(path, ''), kind)
 
     return state
 
 def writestate(repo, state):
     """rewrite .hgsubstate in (outer) repo with these subrepo states"""
     lines = ['%s %s\n' % (state[s][1], s) for s in sorted(state)]
     repo.wwrite('.hgsubstate', ''.join(lines), '')
 
 def submerge(repo, wctx, mctx, actx, overwrite):
     """delegated from merge.applyupdates: merging of .hgsubstate file
     in working context, merging context and ancestor context"""
     if mctx == actx: # backwards?
         actx = wctx.p1()
     s1 = wctx.substate
     s2 = mctx.substate
     sa = actx.substate
     sm = {}
 
     repo.ui.debug("subrepo merge %s %s %s\n" % (wctx, mctx, actx))
 
     def debug(s, msg, r=""):
         if r:
             r = "%s:%s:%s" % r
         repo.ui.debug("  subrepo %s: %s %s\n" % (s, msg, r))
 
     for s, l in s1.items():
         a = sa.get(s, nullstate)
         ld = l # local state with possible dirty flag for compares
         if wctx.sub(s).dirty():
             ld = (l[0], l[1] + "+")
         if wctx == actx: # overwrite
             a = ld
 
         if s in s2:
             r = s2[s]
             if ld == r or r == a: # no change or local is newer
                 sm[s] = l
                 continue
             elif ld == a: # other side changed
                 debug(s, "other changed, get", r)
                 wctx.sub(s).get(r, overwrite)
                 sm[s] = r
             elif ld[0] != r[0]: # sources differ
                 if repo.ui.promptchoice(
                     _(' subrepository sources for %s differ\n'
                       'use (l)ocal source (%s) or (r)emote source (%s)?')
                       % (s, l[0], r[0]),
                       (_('&Local'), _('&Remote')), 0):
                     debug(s, "prompt changed, get", r)
                     wctx.sub(s).get(r, overwrite)
                     sm[s] = r
             elif ld[1] == a[1]: # local side is unchanged
                 debug(s, "other side changed, get", r)
                 wctx.sub(s).get(r, overwrite)
                 sm[s] = r
             else:
                 debug(s, "both sides changed, merge with", r)
                 wctx.sub(s).merge(r)
                 sm[s] = l
         elif ld == a: # remote removed, local unchanged
             debug(s, "remote removed, remove")
             wctx.sub(s).remove()
         elif a == nullstate: # not present in remote or ancestor
             debug(s, "local added, keep")
             sm[s] = l
             continue
         else:
             if repo.ui.promptchoice(
                 _(' local changed subrepository %s which remote removed\n'
                   'use (c)hanged version or (d)elete?') % s,
                 (_('&Changed'), _('&Delete')), 0):
                 debug(s, "prompt remove")
                 wctx.sub(s).remove()
 
     for s, r in sorted(s2.items()):
         if s in s1:
             continue
         elif s not in sa:
             debug(s, "remote added, get", r)
             mctx.sub(s).get(r)
             sm[s] = r
         elif r != sa[s]:
             if repo.ui.promptchoice(
                 _(' remote changed subrepository %s which local removed\n'
                   'use (c)hanged version or (d)elete?') % s,
                 (_('&Changed'), _('&Delete')), 0) == 0:
                 debug(s, "prompt recreate", r)
                 wctx.sub(s).get(r)
                 sm[s] = r
 
     # record merged .hgsubstate
     writestate(repo, sm)
 
 def _updateprompt(ui, sub, dirty, local, remote):
     if dirty:
         msg = (_(' subrepository sources for %s differ\n'
                  'use (l)ocal source (%s) or (r)emote source (%s)?\n')
                % (subrelpath(sub), local, remote))
     else:
         msg = (_(' subrepository sources for %s differ (in checked out version)\n'
                  'use (l)ocal source (%s) or (r)emote source (%s)?\n')
                % (subrelpath(sub), local, remote))
     return ui.promptchoice(msg, (_('&Local'), _('&Remote')), 0)
 
 def reporelpath(repo):
     """return path to this (sub)repo as seen from outermost repo"""
     parent = repo
     while hasattr(parent, '_subparent'):
         parent = parent._subparent
     return repo.root[len(parent.root)+1:]
 
 def subrelpath(sub):
     """return path to this subrepo as seen from outermost repo"""
     if hasattr(sub, '_relpath'):
         return sub._relpath
     if not hasattr(sub, '_repo'):
         return sub._path
     return reporelpath(sub._repo)
 
 def _abssource(repo, push=False, abort=True):
     """return pull/push path of repo - either based on parent repo .hgsub info
     or on the top repo config. Abort or return None if no source found."""
     if hasattr(repo, '_subparent'):
         source = util.url(repo._subsource)
+        if source.isabs():
+            return str(source)
         source.path = posixpath.normpath(source.path)
-        if posixpath.isabs(source.path) or source.scheme:
-            return str(source)
         parent = _abssource(repo._subparent, push, abort=False)
         if parent:
             parent = util.url(parent)
             parent.path = posixpath.join(parent.path, source.path)
             parent.path = posixpath.normpath(parent.path)
             return str(parent)
     else: # recursion reached top repo
         if hasattr(repo, '_subtoppath'):
             return repo._subtoppath
         if push and repo.ui.config('paths', 'default-push'):
             return repo.ui.config('paths', 'default-push')
         if repo.ui.config('paths', 'default'):
             return repo.ui.config('paths', 'default')
     if abort:
         raise util.Abort(_("default path for subrepository %s not found") %
             reporelpath(repo))
 
 def itersubrepos(ctx1, ctx2):
     """find subrepos in ctx1 or ctx2"""
     # Create a (subpath, ctx) mapping where we prefer subpaths from
     # ctx1. The subpaths from ctx2 are important when the .hgsub file
     # has been modified (in ctx2) but not yet committed (in ctx1).
     subpaths = dict.fromkeys(ctx2.substate, ctx2)
     subpaths.update(dict.fromkeys(ctx1.substate, ctx1))
     for subpath, ctx in sorted(subpaths.iteritems()):
         yield subpath, ctx.sub(subpath)
 
 def subrepo(ctx, path):
     """return instance of the right subrepo class for subrepo in path"""
     # subrepo inherently violates our import layering rules
     # because it wants to make repo objects from deep inside the stack
     # so we manually delay the circular imports to not break
     # scripts that don't use our demand-loading
     global hg
     import hg as h
     hg = h
 
     scmutil.pathauditor(ctx._repo.root)(path)
     state = ctx.substate.get(path, nullstate)
     if state[2] not in types:
         raise util.Abort(_('unknown subrepo type %s') % state[2])
     return types[state[2]](ctx, path, state[:2])
 
 # subrepo classes need to implement the following abstract class:
 
 class abstractsubrepo(object):
 
     def dirty(self, ignoreupdate=False):
         """returns true if the dirstate of the subrepo is dirty or does not
         match current stored state. If ignoreupdate is true, only check
         whether the subrepo has uncommitted changes in its dirstate.
         """
         raise NotImplementedError
 
     def checknested(self, path):
         """check if path is a subrepository within this repository"""
         return False
 
     def commit(self, text, user, date):
         """commit the current changes to the subrepo with the given
         log message. Use given user and date if possible. Return the
         new state of the subrepo.
         """
         raise NotImplementedError
 
     def remove(self):
         """remove the subrepo
 
         (should verify the dirstate is not dirty first)
         """
         raise NotImplementedError
 
     def get(self, state, overwrite=False):
         """run whatever commands are needed to put the subrepo into
         this state
         """
         raise NotImplementedError
 
     def merge(self, state):
         """merge currently-saved state with the new state."""
         raise NotImplementedError
 
     def push(self, force):
         """perform whatever action is analogous to 'hg push'
 
         This may be a no-op on some systems.
         """
         raise NotImplementedError
 
     def add(self, ui, match, dryrun, prefix):
         return []
 
     def status(self, rev2, **opts):
         return [], [], [], [], [], [], []
 
     def diff(self, diffopts, node2, match, prefix, **opts):
         pass
 
     def outgoing(self, ui, dest, opts):
         return 1
 
     def incoming(self, ui, source, opts):
         return 1
 
     def files(self):
         """return filename iterator"""
         raise NotImplementedError
 
     def filedata(self, name):
         """return file data"""
         raise NotImplementedError
 
     def fileflags(self, name):
         """return file flags"""
         return ''
 
     def archive(self, ui, archiver, prefix):
         files = self.files()
         total = len(files)
         relpath = subrelpath(self)
         ui.progress(_('archiving (%s)') % relpath, 0,
                     unit=_('files'), total=total)
         for i, name in enumerate(files):
             flags = self.fileflags(name)
             mode = 'x' in flags and 0755 or 0644
             symlink = 'l' in flags
             archiver.addfile(os.path.join(prefix, self._path, name),
                              mode, symlink, self.filedata(name))
             ui.progress(_('archiving (%s)') % relpath, i + 1,
                         unit=_('files'), total=total)
         ui.progress(_('archiving (%s)') % relpath, None)
 
 
 class hgsubrepo(abstractsubrepo):
     def __init__(self, ctx, path, state):
         self._path = path
         self._state = state
         r = ctx._repo
         root = r.wjoin(path)
         create = False
         if not os.path.exists(os.path.join(root, '.hg')):
             create = True
             util.makedirs(root)
         self._repo = hg.repository(r.ui, root, create=create)
         self._initrepo(r, state[0], create)
 
     def _initrepo(self, parentrepo, source, create):
         self._repo._subparent = parentrepo
         self._repo._subsource = source
 
         if create:
             fp = self._repo.opener("hgrc", "w", text=True)
             fp.write('[paths]\n')
 
             def addpathconfig(key, value):
                 if value:
                     fp.write('%s = %s\n' % (key, value))
                     self._repo.ui.setconfig('paths', key, value)
 
             defpath = _abssource(self._repo, abort=False)
             defpushpath = _abssource(self._repo, True, abort=False)
             addpathconfig('default', defpath)
             if defpath != defpushpath:
                 addpathconfig('default-push', defpushpath)
             fp.close()
 
     def add(self, ui, match, dryrun, prefix):
         return cmdutil.add(ui, self._repo, match, dryrun, True,
                            os.path.join(prefix, self._path))
 
     def status(self, rev2, **opts):
         try:
             rev1 = self._state[1]
             ctx1 = self._repo[rev1]
             ctx2 = self._repo[rev2]
             return self._repo.status(ctx1, ctx2, **opts)
         except error.RepoLookupError, inst:
             self._repo.ui.warn(_('warning: error "%s" in subrepository "%s"\n')
                                % (inst, subrelpath(self)))
             return [], [], [], [], [], [], []
 
     def diff(self, diffopts, node2, match, prefix, **opts):
         try:
             node1 = node.bin(self._state[1])
             # We currently expect node2 to come from substate and be
             # in hex format
             if node2 is not None:
                 node2 = node.bin(node2)
             cmdutil.diffordiffstat(self._repo.ui, self._repo, diffopts,
                                    node1, node2, match,
                                    prefix=os.path.join(prefix, self._path),
                                    listsubrepos=True, **opts)
         except error.RepoLookupError, inst:
             self._repo.ui.warn(_('warning: error "%s" in subrepository "%s"\n')
                                % (inst, subrelpath(self)))
 
     def archive(self, ui, archiver, prefix):
         abstractsubrepo.archive(self, ui, archiver, prefix)
 
         rev = self._state[1]
         ctx = self._repo[rev]
         for subpath in ctx.substate:
             s = subrepo(ctx, subpath)
             s.archive(ui, archiver, os.path.join(prefix, self._path))
 
     def dirty(self, ignoreupdate=False):
         r = self._state[1]
         if r == '' and not ignoreupdate: # no state recorded
             return True
         w = self._repo[None]
         if r != w.p1().hex() and not ignoreupdate:
             # different version checked out
             return True
         return w.dirty() # working directory changed
 
     def checknested(self, path):
         return self._repo._checknested(self._repo.wjoin(path))
 
     def commit(self, text, user, date):
         self._repo.ui.debug("committing subrepo %s\n" % subrelpath(self))
         n = self._repo.commit(text, user, date)
         if not n:
             return self._repo['.'].hex() # different version checked out
         return node.hex(n)
 
     def remove(self):
         # we can't fully delete the repository as it may contain
         # local-only history
         self._repo.ui.note(_('removing subrepo %s\n') % subrelpath(self))
         hg.clean(self._repo, node.nullid, False)
 
     def _get(self, state):
         source, revision, kind = state
         if revision not in self._repo:
             self._repo._subsource = source
             srcurl = _abssource(self._repo)
             other = hg.peer(self._repo.ui, {}, srcurl)
             if len(self._repo) == 0:
                 self._repo.ui.status(_('cloning subrepo %s from %s\n')
                                      % (subrelpath(self), srcurl))
                 parentrepo = self._repo._subparent
                 shutil.rmtree(self._repo.root)
                 other, self._repo = hg.clone(self._repo._subparent.ui, {}, other,
                                          self._repo.root, update=False)
                 self._initrepo(parentrepo, source, create=True)
             else:
                 self._repo.ui.status(_('pulling subrepo %s from %s\n')
                                      % (subrelpath(self), srcurl))
                 self._repo.pull(other)
             bookmarks.updatefromremote(self._repo.ui, self._repo, other)
 
     def get(self, state, overwrite=False):
         self._get(state)
         source, revision, kind = state
         self._repo.ui.debug("getting subrepo %s\n" % self._path)
         hg.clean(self._repo, revision, False)
 
     def merge(self, state):
         self._get(state)
         cur = self._repo['.']
         dst = self._repo[state[1]]
         anc = dst.ancestor(cur)
 
         def mergefunc():
             if anc == cur:
                 self._repo.ui.debug("updating subrepo %s\n" % subrelpath(self))
                 hg.update(self._repo, state[1])
             elif anc == dst:
                 self._repo.ui.debug("skipping subrepo %s\n" % subrelpath(self))
             else:
                 self._repo.ui.debug("merging subrepo %s\n" % subrelpath(self))
                 hg.merge(self._repo, state[1], remind=False)
 
         wctx = self._repo[None]
         if self.dirty():
             if anc != dst:
                 if _updateprompt(self._repo.ui, self, wctx.dirty(), cur, dst):
                     mergefunc()
             else:
                 mergefunc()
         else:
             mergefunc()
 
     def push(self, force):
         # push subrepos depth-first for coherent ordering
         c = self._repo['']
         subs = c.substate # only repos that are committed
         for s in sorted(subs):
             if not c.sub(s).push(force):
                 return False
 
         dsturl = _abssource(self._repo, True)
         self._repo.ui.status(_('pushing subrepo %s to %s\n') %
             (subrelpath(self), dsturl))
         other = hg.peer(self._repo.ui, {}, dsturl)
         return self._repo.push(other, force)
 
     def outgoing(self, ui, dest, opts):
         return hg.outgoing(ui, self._repo, _abssource(self._repo, True), opts)
 
     def incoming(self, ui, source, opts):
         return hg.incoming(ui, self._repo, _abssource(self._repo, False), opts)
 
     def files(self):
         rev = self._state[1]
         ctx = self._repo[rev]
         return ctx.manifest()
 
     def filedata(self, name):
         rev = self._state[1]
         return self._repo[rev][name].data()
 
     def fileflags(self, name):
         rev = self._state[1]
         ctx = self._repo[rev]
         return ctx.flags(name)
 
 
 class svnsubrepo(abstractsubrepo):
     def __init__(self, ctx, path, state):
         self._path = path
         self._state = state
         self._ctx = ctx
         self._ui = ctx._repo.ui
 
     def _svncommand(self, commands, filename='', failok=False):
         cmd = ['svn']
         extrakw = {}
         if not self._ui.interactive():
             # Making stdin be a pipe should prevent svn from behaving
             # interactively even if we can't pass --non-interactive.
             extrakw['stdin'] = subprocess.PIPE
             # Starting in svn 1.5 --non-interactive is a global flag
             # instead of being per-command, but we need to support 1.4 so
             # we have to be intelligent about what commands take
             # --non-interactive.
             if commands[0] in ('update', 'checkout', 'commit'):
                 cmd.append('--non-interactive')
         cmd.extend(commands)
         if filename is not None:
             path = os.path.join(self._ctx._repo.origroot, self._path, filename)
             cmd.append(path)
         env = dict(os.environ)
         # Avoid localized output, preserve current locale for everything else.
         env['LC_MESSAGES'] = 'C'
         p = subprocess.Popen(cmd, bufsize=-1, close_fds=util.closefds,
                              stdout=subprocess.PIPE, stderr=subprocess.PIPE,
                               universal_newlines=True, env=env, **extrakw)
         stdout, stderr = p.communicate()
         stderr = stderr.strip()
         if not failok:
             if p.returncode:
                 raise util.Abort(stderr or 'exited with code %d' % p.returncode)
             if stderr:
                 self._ui.warn(stderr + '\n')
         return stdout, stderr
 
     @propertycache
     def _svnversion(self):
         output, err = self._svncommand(['--version'], filename=None)
         m = re.search(r'^svn,\s+version\s+(\d+)\.(\d+)', output)
         if not m:
             raise util.Abort(_('cannot retrieve svn tool version'))
         return (int(m.group(1)), int(m.group(2)))
 
     def _wcrevs(self):
         # Get the working directory revision as well as the last
         # commit revision so we can compare the subrepo state with
         # both. We used to store the working directory one.
         output, err = self._svncommand(['info', '--xml'])
         doc = xml.dom.minidom.parseString(output)
         entries = doc.getElementsByTagName('entry')
         lastrev, rev = '0', '0'
         if entries:
             rev = str(entries[0].getAttribute('revision')) or '0'
             commits = entries[0].getElementsByTagName('commit')
             if commits:
                 lastrev = str(commits[0].getAttribute('revision')) or '0'
         return (lastrev, rev)
 
     def _wcrev(self):
         return self._wcrevs()[0]
 
     def _wcchanged(self):
         """Return (changes, extchanges) where changes is True
         if the working directory was changed, and extchanges is
         True if any of these changes concern an external entry.
         """
         output, err = self._svncommand(['status', '--xml'])
         externals, changes = [], []
         doc = xml.dom.minidom.parseString(output)
         for e in doc.getElementsByTagName('entry'):
             s = e.getElementsByTagName('wc-status')
             if not s:
                 continue
             item = s[0].getAttribute('item')
             props = s[0].getAttribute('props')
             path = e.getAttribute('path')
             if item == 'external':
                 externals.append(path)
             if (item not in ('', 'normal', 'unversioned', 'external')
                 or props not in ('', 'none')):
                 changes.append(path)
         for path in changes:
             for ext in externals:
                 if path == ext or path.startswith(ext + os.sep):
                     return True, True
         return bool(changes), False
 
     def dirty(self, ignoreupdate=False):
         if not self._wcchanged()[0]:
             if self._state[1] in self._wcrevs() or ignoreupdate:
                 return False
         return True
 
     def commit(self, text, user, date):
         # user and date are out of our hands since svn is centralized
         changed, extchanged = self._wcchanged()
         if not changed:
             return self._wcrev()
         if extchanged:
             # Do not try to commit externals
             raise util.Abort(_('cannot commit svn externals'))
         commitinfo, err = self._svncommand(['commit', '-m', text])
         self._ui.status(commitinfo)
         newrev = re.search('Committed revision ([0-9]+).', commitinfo)
         if not newrev:
             raise util.Abort(commitinfo.splitlines()[-1])
         newrev = newrev.groups()[0]
         self._ui.status(self._svncommand(['update', '-r', newrev])[0])
         return newrev
 
     def remove(self):
         if self.dirty():
             self._ui.warn(_('not removing repo %s because '
                             'it has changes.\n' % self._path))
             return
         self._ui.note(_('removing subrepo %s\n') % self._path)
 
         def onerror(function, path, excinfo):
             if function is not os.remove:
                 raise
             # read-only files cannot be unlinked under Windows
             s = os.stat(path)
             if (s.st_mode & stat.S_IWRITE) != 0:
                 raise
             os.chmod(path, stat.S_IMODE(s.st_mode) | stat.S_IWRITE)
             os.remove(path)
 
         path = self._ctx._repo.wjoin(self._path)
         shutil.rmtree(path, onerror=onerror)
         try:
             os.removedirs(os.path.dirname(path))
         except OSError:
             pass
 
     def get(self, state, overwrite=False):
         if overwrite:
             self._svncommand(['revert', '--recursive'])
         args = ['checkout']
         if self._svnversion >= (1, 5):
             args.append('--force')
         args.extend([state[0], '--revision', state[1]])
         status, err = self._svncommand(args, failok=True)
         if not re.search('Checked out revision [0-9]+.', status):
             if ('is already a working copy for a different URL' in err
                 and (self._wcchanged() == (False, False))):
                 # obstructed but clean working copy, so just blow it away.
                 self.remove()
                 self.get(state, overwrite=False)
                 return
             raise util.Abort((status or err).splitlines()[-1])
         self._ui.status(status)
 
     def merge(self, state):
         old = self._state[1]
         new = state[1]
         if new != self._wcrev():
             dirty = old == self._wcrev() or self._wcchanged()[0]
             if _updateprompt(self._ui, self, dirty, self._wcrev(), new):
                 self.get(state, False)
 
     def push(self, force):
         # push is a no-op for SVN
         return True
 
     def files(self):
         output = self._svncommand(['list'])
         # This works because svn forbids \n in filenames.
         return output.splitlines()
 
     def filedata(self, name):
         return self._svncommand(['cat'], name)
 
 
 class gitsubrepo(abstractsubrepo):
     def __init__(self, ctx, path, state):
         # TODO add git version check.
         self._state = state
         self._ctx = ctx
         self._path = path
         self._relpath = os.path.join(reporelpath(ctx._repo), path)
         self._abspath = ctx._repo.wjoin(path)
         self._subparent = ctx._repo
         self._ui = ctx._repo.ui
 
     def _gitcommand(self, commands, env=None, stream=False):
         return self._gitdir(commands, env=env, stream=stream)[0]
 
     def _gitdir(self, commands, env=None, stream=False):
         return self._gitnodir(commands, env=env, stream=stream,
                               cwd=self._abspath)
 
     def _gitnodir(self, commands, env=None, stream=False, cwd=None):
         """Calls the git command
 
         The methods tries to call the git command. versions previor to 1.6.0
         are not supported and very probably fail.
         """
         self._ui.debug('%s: git %s\n' % (self._relpath, ' '.join(commands)))
         # unless ui.quiet is set, print git's stderr,
         # which is mostly progress and useful info
         errpipe = None
         if self._ui.quiet:
             errpipe = open(os.devnull, 'w')
         p = subprocess.Popen(['git'] + commands, bufsize=-1, cwd=cwd, env=env,
                              close_fds=util.closefds,
                              stdout=subprocess.PIPE, stderr=errpipe)
         if stream:
             return p.stdout, None
 
         retdata = p.stdout.read().strip()
         # wait for the child to exit to avoid race condition.
         p.wait()
 
         if p.returncode != 0 and p.returncode != 1:
             # there are certain error codes that are ok
             command = commands[0]
             if command in ('cat-file', 'symbolic-ref'):
                 return retdata, p.returncode
             # for all others, abort
             raise util.Abort('git %s error %d in %s' %
                              (command, p.returncode, self._relpath))
 
         return retdata, p.returncode
 
     def _gitmissing(self):
         return not os.path.exists(os.path.join(self._abspath, '.git'))
 
     def _gitstate(self):
         return self._gitcommand(['rev-parse', 'HEAD'])
 
     def _gitcurrentbranch(self):
         current, err = self._gitdir(['symbolic-ref', 'HEAD', '--quiet'])
         if err:
             current = None
         return current
 
     def _gitremote(self, remote):
         out = self._gitcommand(['remote', 'show', '-n', remote])
         line = out.split('\n')[1]
         i = line.index('URL: ') + len('URL: ')
         return line[i:]
 
     def _githavelocally(self, revision):
         out, code = self._gitdir(['cat-file', '-e', revision])
         return code == 0
 
     def _gitisancestor(self, r1, r2):
         base = self._gitcommand(['merge-base', r1, r2])
         return base == r1
 
     def _gitisbare(self):
         return self._gitcommand(['config', '--bool', 'core.bare']) == 'true'
 
     def _gitbranchmap(self):
         '''returns 2 things:
         a map from git branch to revision
         a map from revision to branches'''
         branch2rev = {}
         rev2branch = {}
 
         out = self._gitcommand(['for-each-ref', '--format',
                                 '%(objectname) %(refname)'])
         for line in out.split('\n'):
             revision, ref = line.split(' ')
             if (not ref.startswith('refs/heads/') and
                 not ref.startswith('refs/remotes/')):
                 continue
             if ref.startswith('refs/remotes/') and ref.endswith('/HEAD'):
                 continue # ignore remote/HEAD redirects
             branch2rev[ref] = revision
             rev2branch.setdefault(revision, []).append(ref)
         return branch2rev, rev2branch
 
     def _gittracking(self, branches):
         'return map of remote branch to local tracking branch'
         # assumes no more than one local tracking branch for each remote
         tracking = {}
         for b in branches:
             if b.startswith('refs/remotes/'):
                 continue
             remote = self._gitcommand(['config', 'branch.%s.remote' % b])
             if remote:
                 ref = self._gitcommand(['config', 'branch.%s.merge' % b])
                 tracking['refs/remotes/%s/%s' %
                          (remote, ref.split('/', 2)[2])] = b
         return tracking
 
     def _abssource(self, source):
         if '://' not in source:
             # recognize the scp syntax as an absolute source
             colon = source.find(':')
             if colon != -1 and '/' not in source[:colon]:
                 return source
         self._subsource = source
         return _abssource(self)
 
     def _fetch(self, source, revision):
         if self._gitmissing():
             source = self._abssource(source)
             self._ui.status(_('cloning subrepo %s from %s\n') %
                             (self._relpath, source))
             self._gitnodir(['clone', source, self._abspath])
         if self._githavelocally(revision):
             return
         self._ui.status(_('pulling subrepo %s from %s\n') %
                         (self._relpath, self._gitremote('origin')))
         # try only origin: the originally cloned repo
         self._gitcommand(['fetch'])
         if not self._githavelocally(revision):
             raise util.Abort(_("revision %s does not exist in subrepo %s\n") %
                                (revision, self._relpath))
 
     def dirty(self, ignoreupdate=False):
         if self._gitmissing():
             return self._state[1] != ''
         if self._gitisbare():
             return True
         if not ignoreupdate and self._state[1] != self._gitstate():
             # different version checked out
             return True
         # check for staged changes or modified files; ignore untracked files
         out, code = self._gitdir(['diff-index', '--quiet', 'HEAD'])
         return code == 1
 
     def get(self, state, overwrite=False):
         source, revision, kind = state
         if not revision:
             self.remove()
             return
         self._fetch(source, revision)
         # if the repo was set to be bare, unbare it
         if self._gitisbare():
             self._gitcommand(['config', 'core.bare', 'false'])
             if self._gitstate() == revision:
                 self._gitcommand(['reset', '--hard', 'HEAD'])
                 return
         elif self._gitstate() == revision:
             if overwrite:
                 # first reset the index to unmark new files for commit, because
                 # reset --hard will otherwise throw away files added for commit,
                 # not just unmark them.
                 self._gitcommand(['reset', 'HEAD'])
                 self._gitcommand(['reset', '--hard', 'HEAD'])
             return
         branch2rev, rev2branch = self._gitbranchmap()
 
         def checkout(args):
             cmd = ['checkout']
             if overwrite:
                 # first reset the index to unmark new files for commit, because
                 # the -f option will otherwise throw away files added for
                 # commit, not just unmark them.
                 self._gitcommand(['reset', 'HEAD'])
                 cmd.append('-f')
             self._gitcommand(cmd + args)
 
         def rawcheckout():
             # no branch to checkout, check it out with no branch
             self._ui.warn(_('checking out detached HEAD in subrepo %s\n') %
                           self._relpath)
             self._ui.warn(_('check out a git branch if you intend '
                             'to make changes\n'))
             checkout(['-q', revision])
 
         if revision not in rev2branch:
             rawcheckout()
             return
         branches = rev2branch[revision]
         firstlocalbranch = None
         for b in branches:
             if b == 'refs/heads/master':
                 # master trumps all other branches
                 checkout(['refs/heads/master'])
                 return
             if not firstlocalbranch and not b.startswith('refs/remotes/'):
                 firstlocalbranch = b
         if firstlocalbranch:
             checkout([firstlocalbranch])
             return
 
         tracking = self._gittracking(branch2rev.keys())
         # choose a remote branch already tracked if possible
         remote = branches[0]
         if remote not in tracking:
             for b in branches:
                 if b in tracking:
                     remote = b
                     break
 
         if remote not in tracking:
             # create a new local tracking branch
             local = remote.split('/', 2)[2]
             checkout(['-b', local, remote])
         elif self._gitisancestor(branch2rev[tracking[remote]], remote):
             # When updating to a tracked remote branch,
             # if the local tracking branch is downstream of it,
             # a normal `git pull` would have performed a "fast-forward merge"
             # which is equivalent to updating the local branch to the remote.
             # Since we are only looking at branching at update, we need to
             # detect this situation and perform this action lazily.
             if tracking[remote] != self._gitcurrentbranch():
                 checkout([tracking[remote]])
             self._gitcommand(['merge', '--ff', remote])
         else:
             # a real merge would be required, just checkout the revision
             rawcheckout()
 
     def commit(self, text, user, date):
         if self._gitmissing():
             raise util.Abort(_("subrepo %s is missing") % self._relpath)
         cmd = ['commit', '-a', '-m', text]
         env = os.environ.copy()
         if user:
             cmd += ['--author', user]
         if date:
             # git's date parser silently ignores when seconds < 1e9
             # convert to ISO8601
             env['GIT_AUTHOR_DATE'] = util.datestr(date,
                                                   '%Y-%m-%dT%H:%M:%S %1%2')
         self._gitcommand(cmd, env=env)
         # make sure commit works otherwise HEAD might not exist under certain
         # circumstances
         return self._gitstate()
 
     def merge(self, state):
         source, revision, kind = state
         self._fetch(source, revision)
         base = self._gitcommand(['merge-base', revision, self._state[1]])
         out, code = self._gitdir(['diff-index', '--quiet', 'HEAD'])
 
         def mergefunc():
             if base == revision:
                 self.get(state) # fast forward merge
             elif base != self._state[1]:
                 self._gitcommand(['merge', '--no-commit', revision])
 
         if self.dirty():
             if self._gitstate() != revision:
                 dirty = self._gitstate() == self._state[1] or code != 0
                 if _updateprompt(self._ui, self, dirty,
                                  self._state[1][:7], revision[:7]):
                     mergefunc()
         else:
             mergefunc()
 
     def push(self, force):
         if not self._state[1]:
             return True
         if self._gitmissing():
             raise util.Abort(_("subrepo %s is missing") % self._relpath)
         # if a branch in origin contains the revision, nothing to do
         branch2rev, rev2branch = self._gitbranchmap()
         if self._state[1] in rev2branch:
             for b in rev2branch[self._state[1]]:
                 if b.startswith('refs/remotes/origin/'):
                     return True
         for b, revision in branch2rev.iteritems():
             if b.startswith('refs/remotes/origin/'):
                 if self._gitisancestor(self._state[1], revision):
                     return True
         # otherwise, try to push the currently checked out branch
         cmd = ['push']
         if force:
             cmd.append('--force')
 
         current = self._gitcurrentbranch()
         if current:
             # determine if the current branch is even useful
             if not self._gitisancestor(self._state[1], current):
                 self._ui.warn(_('unrelated git branch checked out '
                                 'in subrepo %s\n') % self._relpath)
                 return False
             self._ui.status(_('pushing branch %s of subrepo %s\n') %
                             (current.split('/', 2)[2], self._relpath))
             self._gitcommand(cmd + ['origin', current])
             return True
         else:
             self._ui.warn(_('no branch checked out in subrepo %s\n'
                             'cannot push revision %s') %
                           (self._relpath, self._state[1]))
             return False
 
     def remove(self):
         if self._gitmissing():
             return
         if self.dirty():
             self._ui.warn(_('not removing repo %s because '
                             'it has changes.\n') % self._relpath)
             return
         # we can't fully delete the repository as it may contain
         # local-only history
         self._ui.note(_('removing subrepo %s\n') % self._relpath)
         self._gitcommand(['config', 'core.bare', 'true'])
         for f in os.listdir(self._abspath):
             if f == '.git':
                 continue
             path = os.path.join(self._abspath, f)
             if os.path.isdir(path) and not os.path.islink(path):
                 shutil.rmtree(path)
             else:
                 os.remove(path)
 
     def archive(self, ui, archiver, prefix):
         source, revision = self._state
         if not revision:
             return
         self._fetch(source, revision)
 
         # Parse git's native archive command.
         # This should be much faster than manually traversing the trees
         # and objects with many subprocess calls.
         tarstream = self._gitcommand(['archive', revision], stream=True)
         tar = tarfile.open(fileobj=tarstream, mode='r|')
         relpath = subrelpath(self)
         ui.progress(_('archiving (%s)') % relpath, 0, unit=_('files'))
         for i, info in enumerate(tar):
             if info.isdir():
                 continue
             if info.issym():
                 data = info.linkname
             else:
                 data = tar.extractfile(info).read()
             archiver.addfile(os.path.join(prefix, self._path, info.name),
                              info.mode, info.issym(), data)
             ui.progress(_('archiving (%s)') % relpath, i + 1,
                         unit=_('files'))
         ui.progress(_('archiving (%s)') % relpath, None)
 
 
     def status(self, rev2, **opts):
         rev1 = self._state[1]
         if self._gitmissing() or not rev1:
             # if the repo is missing, return no results
             return [], [], [], [], [], [], []
         modified, added, removed = [], [], []
         if rev2:
             command = ['diff-tree', rev1, rev2]
         else:
             command = ['diff-index', rev1]
         out = self._gitcommand(command)
         for line in out.split('\n'):
             tab = line.find('\t')
             if tab == -1:
                 continue
             status, f = line[tab - 1], line[tab + 1:]
             if status == 'M':
                 modified.append(f)
             elif status == 'A':
                 added.append(f)
             elif status == 'D':
                 removed.append(f)
 
         deleted = unknown = ignored = clean = []
         return modified, added, removed, deleted, unknown, ignored, clean
 
 types = {
     'hg': hgsubrepo,
     'svn': svnsubrepo,
     'git': gitsubrepo,
     }

mercurial/util.py

@@ -1,1601 +1,1612 b''
 # util.py - Mercurial utility functions and platform specfic implementations
 #
 #  Copyright 2005 K. Thananchayan <[e-mail]>
 #  Copyright 2005-2007 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.
 
 """Mercurial utility functions and platform specfic implementations.
 
 This contains helper routines that are independent of the SCM core and
 hide platform-specific details from the core.
 """
 
 from i18n import _
 import error, osutil, encoding
 import errno, re, shutil, sys, tempfile, traceback
 import os, time, calendar, textwrap, unicodedata, signal
 import imp, socket, urllib
 
 # Python compatibility
 
 def sha1(s):
     return _fastsha1(s)
 
 def _fastsha1(s):
     # This function will import sha1 from hashlib or sha (whichever is
     # available) and overwrite itself with it on the first call.
     # Subsequent calls will go directly to the imported function.
     if sys.version_info >= (2, 5):
         from hashlib import sha1 as _sha1
     else:
         from sha import sha as _sha1
     global _fastsha1, sha1
     _fastsha1 = sha1 = _sha1
     return _sha1(s)
 
 import __builtin__
 
 if sys.version_info[0] < 3:
     def fakebuffer(sliceable, offset=0):
         return sliceable[offset:]
 else:
     def fakebuffer(sliceable, offset=0):
         return memoryview(sliceable)[offset:]
 try:
     buffer
 except NameError:
     __builtin__.buffer = fakebuffer
 
 import subprocess
 closefds = os.name == 'posix'
 
 def popen2(cmd, env=None, newlines=False):
     # Setting bufsize to -1 lets the system decide the buffer size.
     # The default for bufsize is 0, meaning unbuffered. This leads to
     # poor performance on Mac OS X: http://bugs.python.org/issue4194
     p = subprocess.Popen(cmd, shell=True, bufsize=-1,
                          close_fds=closefds,
                          stdin=subprocess.PIPE, stdout=subprocess.PIPE,
                          universal_newlines=newlines,
                          env=env)
     return p.stdin, p.stdout
 
 def popen3(cmd, env=None, newlines=False):
     p = subprocess.Popen(cmd, shell=True, bufsize=-1,
                          close_fds=closefds,
                          stdin=subprocess.PIPE, stdout=subprocess.PIPE,
                          stderr=subprocess.PIPE,
                          universal_newlines=newlines,
                          env=env)
     return p.stdin, p.stdout, p.stderr
 
 def version():
     """Return version information if available."""
     try:
         import __version__
         return __version__.version
     except ImportError:
         return 'unknown'
 
 # used by parsedate
 defaultdateformats = (
     '%Y-%m-%d %H:%M:%S',
     '%Y-%m-%d %I:%M:%S%p',
     '%Y-%m-%d %H:%M',
     '%Y-%m-%d %I:%M%p',
     '%Y-%m-%d',
     '%m-%d',
     '%m/%d',
     '%m/%d/%y',
     '%m/%d/%Y',
     '%a %b %d %H:%M:%S %Y',
     '%a %b %d %I:%M:%S%p %Y',
     '%a, %d %b %Y %H:%M:%S',        #  GNU coreutils "/bin/date --rfc-2822"
     '%b %d %H:%M:%S %Y',
     '%b %d %I:%M:%S%p %Y',
     '%b %d %H:%M:%S',
     '%b %d %I:%M:%S%p',
     '%b %d %H:%M',
     '%b %d %I:%M%p',
     '%b %d %Y',
     '%b %d',
     '%H:%M:%S',
     '%I:%M:%S%p',
     '%H:%M',
     '%I:%M%p',
 )
 
 extendeddateformats = defaultdateformats + (
     "%Y",
     "%Y-%m",
     "%b",
     "%b %Y",
     )
 
 def cachefunc(func):
     '''cache the result of function calls'''
     # XXX doesn't handle keywords args
     cache = {}
     if func.func_code.co_argcount == 1:
         # we gain a small amount of time because
         # we don't need to pack/unpack the list
         def f(arg):
             if arg not in cache:
                 cache[arg] = func(arg)
             return cache[arg]
     else:
         def f(*args):
             if args not in cache:
                 cache[args] = func(*args)
             return cache[args]
 
     return f
 
 def lrucachefunc(func):
     '''cache most recent results of function calls'''
     cache = {}
     order = []
     if func.func_code.co_argcount == 1:
         def f(arg):
             if arg not in cache:
                 if len(cache) > 20:
                     del cache[order.pop(0)]
                 cache[arg] = func(arg)
             else:
                 order.remove(arg)
             order.append(arg)
             return cache[arg]
     else:
         def f(*args):
             if args not in cache:
                 if len(cache) > 20:
                     del cache[order.pop(0)]
                 cache[args] = func(*args)
             else:
                 order.remove(args)
             order.append(args)
             return cache[args]
 
     return f
 
 class propertycache(object):
     def __init__(self, func):
         self.func = func
         self.name = func.__name__
     def __get__(self, obj, type=None):
         result = self.func(obj)
         setattr(obj, self.name, result)
         return result
 
 def pipefilter(s, cmd):
     '''filter string S through command CMD, returning its output'''
     p = subprocess.Popen(cmd, shell=True, close_fds=closefds,
                          stdin=subprocess.PIPE, stdout=subprocess.PIPE)
     pout, perr = p.communicate(s)
     return pout
 
 def tempfilter(s, cmd):
     '''filter string S through a pair of temporary files with CMD.
     CMD is used as a template to create the real command to be run,
     with the strings INFILE and OUTFILE replaced by the real names of
     the temporary files generated.'''
     inname, outname = None, None
     try:
         infd, inname = tempfile.mkstemp(prefix='hg-filter-in-')
         fp = os.fdopen(infd, 'wb')
         fp.write(s)
         fp.close()
         outfd, outname = tempfile.mkstemp(prefix='hg-filter-out-')
         os.close(outfd)
         cmd = cmd.replace('INFILE', inname)
         cmd = cmd.replace('OUTFILE', outname)
         code = os.system(cmd)
         if sys.platform == 'OpenVMS' and code & 1:
             code = 0
         if code:
             raise Abort(_("command '%s' failed: %s") %
                         (cmd, explainexit(code)))
         fp = open(outname, 'rb')
         r = fp.read()
         fp.close()
         return r
     finally:
         try:
             if inname:
                 os.unlink(inname)
         except OSError:
             pass
         try:
             if outname:
                 os.unlink(outname)
         except OSError:
             pass
 
 filtertable = {
     'tempfile:': tempfilter,
     'pipe:': pipefilter,
     }
 
 def filter(s, cmd):
     "filter a string through a command that transforms its input to its output"
     for name, fn in filtertable.iteritems():
         if cmd.startswith(name):
             return fn(s, cmd[len(name):].lstrip())
     return pipefilter(s, cmd)
 
 def binary(s):
     """return true if a string is binary data"""
     return bool(s and '\0' in s)
 
 def increasingchunks(source, min=1024, max=65536):
     '''return no less than min bytes per chunk while data remains,
     doubling min after each chunk until it reaches max'''
     def log2(x):
         if not x:
             return 0
         i = 0
         while x:
             x >>= 1
             i += 1
         return i - 1
 
     buf = []
     blen = 0
     for chunk in source:
         buf.append(chunk)
         blen += len(chunk)
         if blen >= min:
             if min < max:
                 min = min << 1
                 nmin = 1 << log2(blen)
                 if nmin > min:
                     min = nmin
                 if min > max:
                     min = max
             yield ''.join(buf)
             blen = 0
             buf = []
     if buf:
         yield ''.join(buf)
 
 Abort = error.Abort
 
 def always(fn):
     return True
 
 def never(fn):
     return False
 
 def pathto(root, n1, n2):
     '''return the relative path from one place to another.
     root should use os.sep to separate directories
     n1 should use os.sep to separate directories
     n2 should use "/" to separate directories
     returns an os.sep-separated path.
 
     If n1 is a relative path, it's assumed it's
     relative to root.
     n2 should always be relative to root.
     '''
     if not n1:
         return localpath(n2)
     if os.path.isabs(n1):
         if os.path.splitdrive(root)[0] != os.path.splitdrive(n1)[0]:
             return os.path.join(root, localpath(n2))
         n2 = '/'.join((pconvert(root), n2))
     a, b = splitpath(n1), n2.split('/')
     a.reverse()
     b.reverse()
     while a and b and a[-1] == b[-1]:
         a.pop()
         b.pop()
     b.reverse()
     return os.sep.join((['..'] * len(a)) + b) or '.'
 
 _hgexecutable = None
 
 def mainfrozen():
     """return True if we are a frozen executable.
 
     The code supports py2exe (most common, Windows only) and tools/freeze
     (portable, not much used).
     """
     return (hasattr(sys, "frozen") or # new py2exe
             hasattr(sys, "importers") or # old py2exe
             imp.is_frozen("__main__")) # tools/freeze
 
 def hgexecutable():
     """return location of the 'hg' executable.
 
     Defaults to $HG or 'hg' in the search path.
     """
     if _hgexecutable is None:
         hg = os.environ.get('HG')
         if hg:
             _sethgexecutable(hg)
         elif mainfrozen():
             _sethgexecutable(sys.executable)
         else:
             exe = findexe('hg') or os.path.basename(sys.argv[0])
             _sethgexecutable(exe)
     return _hgexecutable
 
 def _sethgexecutable(path):
     """set location of the 'hg' executable"""
     global _hgexecutable
     _hgexecutable = path
 
 def system(cmd, environ={}, cwd=None, onerr=None, errprefix=None, out=None):
     '''enhanced shell command execution.
     run with environment maybe modified, maybe in different dir.
 
     if command fails and onerr is None, return status.  if ui object,
     print error message and return status, else raise onerr object as
     exception.
 
     if out is specified, it is assumed to be a file-like object that has a
     write() method. stdout and stderr will be redirected to out.'''
     try:
         sys.stdout.flush()
     except Exception:
         pass
     def py2shell(val):
         'convert python object into string that is useful to shell'
         if val is None or val is False:
             return '0'
         if val is True:
             return '1'
         return str(val)
     origcmd = cmd
     cmd = quotecommand(cmd)
     env = dict(os.environ)
     env.update((k, py2shell(v)) for k, v in environ.iteritems())
     env['HG'] = hgexecutable()
     if out is None or out == sys.__stdout__:
         rc = subprocess.call(cmd, shell=True, close_fds=closefds,
                              env=env, cwd=cwd)
     else:
         proc = subprocess.Popen(cmd, shell=True, close_fds=closefds,
                                 env=env, cwd=cwd, stdout=subprocess.PIPE,
                                 stderr=subprocess.STDOUT)
         for line in proc.stdout:
             out.write(line)
         proc.wait()
         rc = proc.returncode
     if sys.platform == 'OpenVMS' and rc & 1:
         rc = 0
     if rc and onerr:
         errmsg = '%s %s' % (os.path.basename(origcmd.split(None, 1)[0]),
                             explainexit(rc)[0])
         if errprefix:
             errmsg = '%s: %s' % (errprefix, errmsg)
         try:
             onerr.warn(errmsg + '\n')
         except AttributeError:
             raise onerr(errmsg)
     return rc
 
 def checksignature(func):
     '''wrap a function with code to check for calling errors'''
     def check(*args, **kwargs):
         try:
             return func(*args, **kwargs)
         except TypeError:
             if len(traceback.extract_tb(sys.exc_info()[2])) == 1:
                 raise error.SignatureError
             raise
 
     return check
 
 def makedir(path, notindexed):
     os.mkdir(path)
 
 def unlinkpath(f):
     """unlink and remove the directory if it is empty"""
     os.unlink(f)
     # try removing directories that might now be empty
     try:
         os.removedirs(os.path.dirname(f))
     except OSError:
         pass
 
 def copyfile(src, dest):
     "copy a file, preserving mode and atime/mtime"
     if os.path.islink(src):
         try:
             os.unlink(dest)
         except OSError:
             pass
         os.symlink(os.readlink(src), dest)
     else:
         try:
             shutil.copyfile(src, dest)
             shutil.copymode(src, dest)
         except shutil.Error, inst:
             raise Abort(str(inst))
 
 def copyfiles(src, dst, hardlink=None):
     """Copy a directory tree using hardlinks if possible"""
 
     if hardlink is None:
         hardlink = (os.stat(src).st_dev ==
                     os.stat(os.path.dirname(dst)).st_dev)
 
     num = 0
     if os.path.isdir(src):
         os.mkdir(dst)
         for name, kind in osutil.listdir(src):
             srcname = os.path.join(src, name)
             dstname = os.path.join(dst, name)
             hardlink, n = copyfiles(srcname, dstname, hardlink)
             num += n
     else:
         if hardlink:
             try:
                 oslink(src, dst)
             except (IOError, OSError):
                 hardlink = False
                 shutil.copy(src, dst)
         else:
             shutil.copy(src, dst)
         num += 1
 
     return hardlink, num
 
 _winreservednames = '''con prn aux nul
     com1 com2 com3 com4 com5 com6 com7 com8 com9
     lpt1 lpt2 lpt3 lpt4 lpt5 lpt6 lpt7 lpt8 lpt9'''.split()
 _winreservedchars = ':*?"<>|'
 def checkwinfilename(path):
     '''Check that the base-relative path is a valid filename on Windows.
     Returns None if the path is ok, or a UI string describing the problem.
 
     >>> checkwinfilename("just/a/normal/path")
     >>> checkwinfilename("foo/bar/con.xml")
     "filename contains 'con', which is reserved on Windows"
     >>> checkwinfilename("foo/con.xml/bar")
     "filename contains 'con', which is reserved on Windows"
     >>> checkwinfilename("foo/bar/xml.con")
     >>> checkwinfilename("foo/bar/AUX/bla.txt")
     "filename contains 'AUX', which is reserved on Windows"
     >>> checkwinfilename("foo/bar/bla:.txt")
     "filename contains ':', which is reserved on Windows"
     >>> checkwinfilename("foo/bar/b\07la.txt")
     "filename contains '\\\\x07', which is invalid on Windows"
     >>> checkwinfilename("foo/bar/bla ")
     "filename ends with ' ', which is not allowed on Windows"
     '''
     for n in path.replace('\\', '/').split('/'):
         if not n:
             continue
         for c in n:
             if c in _winreservedchars:
                 return _("filename contains '%s', which is reserved "
                          "on Windows") % c
             if ord(c) <= 31:
                 return _("filename contains %r, which is invalid "
                          "on Windows") % c
         base = n.split('.')[0]
         if base and base.lower() in _winreservednames:
             return _("filename contains '%s', which is reserved "
                      "on Windows") % base
         t = n[-1]
         if t in '. ':
             return _("filename ends with '%s', which is not allowed "
                      "on Windows") % t
 
 def lookupreg(key, name=None, scope=None):
     return None
 
 def hidewindow():
     """Hide current shell window.
 
     Used to hide the window opened when starting asynchronous
     child process under Windows, unneeded on other systems.
     """
     pass
 
 if os.name == 'nt':
     checkosfilename = checkwinfilename
     from windows import *
 else:
     from posix import *
 
 def makelock(info, pathname):
     try:
         return os.symlink(info, pathname)
     except OSError, why:
         if why.errno == errno.EEXIST:
             raise
     except AttributeError: # no symlink in os
         pass
 
     ld = os.open(pathname, os.O_CREAT | os.O_WRONLY | os.O_EXCL)
     os.write(ld, info)
     os.close(ld)
 
 def readlock(pathname):
     try:
         return os.readlink(pathname)
     except OSError, why:
         if why.errno not in (errno.EINVAL, errno.ENOSYS):
             raise
     except AttributeError: # no symlink in os
         pass
     fp = posixfile(pathname)
     r = fp.read()
     fp.close()
     return r
 
 def fstat(fp):
     '''stat file object that may not have fileno method.'''
     try:
         return os.fstat(fp.fileno())
     except AttributeError:
         return os.stat(fp.name)
 
 # File system features
 
 def checkcase(path):
     """
     Check whether the given path is on a case-sensitive filesystem
 
     Requires a path (like /foo/.hg) ending with a foldable final
     directory component.
     """
     s1 = os.stat(path)
     d, b = os.path.split(path)
     p2 = os.path.join(d, b.upper())
     if path == p2:
         p2 = os.path.join(d, b.lower())
     try:
         s2 = os.stat(p2)
         if s2 == s1:
             return False
         return True
     except OSError:
         return True
 
 _fspathcache = {}
 def fspath(name, root):
     '''Get name in the case stored in the filesystem
 
     The name is either relative to root, or it is an absolute path starting
     with root. Note that this function is unnecessary, and should not be
     called, for case-sensitive filesystems (simply because it's expensive).
     '''
     # If name is absolute, make it relative
     if name.lower().startswith(root.lower()):
         l = len(root)
         if name[l] == os.sep or name[l] == os.altsep:
             l = l + 1
         name = name[l:]
 
     if not os.path.lexists(os.path.join(root, name)):
         return None
 
     seps = os.sep
     if os.altsep:
         seps = seps + os.altsep
     # Protect backslashes. This gets silly very quickly.
     seps.replace('\\','\\\\')
     pattern = re.compile(r'([^%s]+)|([%s]+)' % (seps, seps))
     dir = os.path.normcase(os.path.normpath(root))
     result = []
     for part, sep in pattern.findall(name):
         if sep:
             result.append(sep)
             continue
 
         if dir not in _fspathcache:
             _fspathcache[dir] = os.listdir(dir)
         contents = _fspathcache[dir]
 
         lpart = part.lower()
         lenp = len(part)
         for n in contents:
             if lenp == len(n) and n.lower() == lpart:
                 result.append(n)
                 break
         else:
             # Cannot happen, as the file exists!
             result.append(part)
         dir = os.path.join(dir, lpart)
 
     return ''.join(result)
 
 def checknlink(testfile):
     '''check whether hardlink count reporting works properly'''
 
     # testfile may be open, so we need a separate file for checking to
     # work around issue2543 (or testfile may get lost on Samba shares)
     f1 = testfile + ".hgtmp1"
     if os.path.lexists(f1):
         return False
     try:
         posixfile(f1, 'w').close()
     except IOError:
         return False
 
     f2 = testfile + ".hgtmp2"
     fd = None
     try:
         try:
             oslink(f1, f2)
         except OSError:
             return False
 
         # nlinks() may behave differently for files on Windows shares if
         # the file is open.
         fd = posixfile(f2)
         return nlinks(f2) > 1
     finally:
         if fd is not None:
             fd.close()
         for f in (f1, f2):
             try:
                 os.unlink(f)
             except OSError:
                 pass
 
     return False
 
 def endswithsep(path):
     '''Check path ends with os.sep or os.altsep.'''
     return path.endswith(os.sep) or os.altsep and path.endswith(os.altsep)
 
 def splitpath(path):
     '''Split path by os.sep.
     Note that this function does not use os.altsep because this is
     an alternative of simple "xxx.split(os.sep)".
     It is recommended to use os.path.normpath() before using this
     function if need.'''
     return path.split(os.sep)
 
 def gui():
     '''Are we running in a GUI?'''
     if sys.platform == 'darwin':
         if 'SSH_CONNECTION' in os.environ:
             # handle SSH access to a box where the user is logged in
             return False
         elif getattr(osutil, 'isgui', None):
             # check if a CoreGraphics session is available
             return osutil.isgui()
         else:
             # pure build; use a safe default
             return True
     else:
         return os.name == "nt" or os.environ.get("DISPLAY")
 
 def mktempcopy(name, emptyok=False, createmode=None):
     """Create a temporary file with the same contents from name
 
     The permission bits are copied from the original file.
 
     If the temporary file is going to be truncated immediately, you
     can use emptyok=True as an optimization.
 
     Returns the name of the temporary file.
     """
     d, fn = os.path.split(name)
     fd, temp = tempfile.mkstemp(prefix='.%s-' % fn, dir=d)
     os.close(fd)
     # Temporary files are created with mode 0600, which is usually not
     # what we want.  If the original file already exists, just copy
     # its mode.  Otherwise, manually obey umask.
     try:
         st_mode = os.lstat(name).st_mode & 0777
     except OSError, inst:
         if inst.errno != errno.ENOENT:
             raise
         st_mode = createmode
         if st_mode is None:
             st_mode = ~umask
         st_mode &= 0666
     os.chmod(temp, st_mode)
     if emptyok:
         return temp
     try:
         try:
             ifp = posixfile(name, "rb")
         except IOError, inst:
             if inst.errno == errno.ENOENT:
                 return temp
             if not getattr(inst, 'filename', None):
                 inst.filename = name
             raise
         ofp = posixfile(temp, "wb")
         for chunk in filechunkiter(ifp):
             ofp.write(chunk)
         ifp.close()
         ofp.close()
     except:
         try: os.unlink(temp)
         except: pass
         raise
     return temp
 
 class atomictempfile(object):
     '''writeable file object that atomically updates a file
 
     All writes will go to a temporary copy of the original file. Call
     rename() when you are done writing, and atomictempfile will rename
     the temporary copy to the original name, making the changes visible.
 
     Unlike other file-like objects, close() discards your writes by
     simply deleting the temporary file.
     '''
     def __init__(self, name, mode='w+b', createmode=None):
         self.__name = name      # permanent name
         self._tempname = mktempcopy(name, emptyok=('w' in mode),
                                     createmode=createmode)
         self._fp = posixfile(self._tempname, mode)
 
         # delegated methods
         self.write = self._fp.write
         self.fileno = self._fp.fileno
 
     def rename(self):
         if not self._fp.closed:
             self._fp.close()
             rename(self._tempname, localpath(self.__name))
 
     def close(self):
         if not self._fp.closed:
             try:
                 os.unlink(self._tempname)
             except OSError:
                 pass
             self._fp.close()
 
     def __del__(self):
         if hasattr(self, '_fp'): # constructor actually did something
             self.close()
 
 def makedirs(name, mode=None):
     """recursive directory creation with parent mode inheritance"""
     parent = os.path.abspath(os.path.dirname(name))
     try:
         os.mkdir(name)
         if mode is not None:
             os.chmod(name, mode)
         return
     except OSError, err:
         if err.errno == errno.EEXIST:
             return
         if not name or parent == name or err.errno != errno.ENOENT:
             raise
     makedirs(parent, mode)
     makedirs(name, mode)
 
 def readfile(path):
     fp = open(path, 'rb')
     try:
         return fp.read()
     finally:
         fp.close()
 
 def writefile(path, text):
     fp = open(path, 'wb')
     try:
         fp.write(text)
     finally:
         fp.close()
 
 def appendfile(path, text):
     fp = open(path, 'ab')
     try:
         fp.write(text)
     finally:
         fp.close()
 
 class chunkbuffer(object):
     """Allow arbitrary sized chunks of data to be efficiently read from an
     iterator over chunks of arbitrary size."""
 
     def __init__(self, in_iter):
         """in_iter is the iterator that's iterating over the input chunks.
         targetsize is how big a buffer to try to maintain."""
         def splitbig(chunks):
             for chunk in chunks:
                 if len(chunk) > 2**20:
                     pos = 0
                     while pos < len(chunk):
                         end = pos + 2 ** 18
                         yield chunk[pos:end]
                         pos = end
                 else:
                     yield chunk
         self.iter = splitbig(in_iter)
         self._queue = []
 
     def read(self, l):
         """Read L bytes of data from the iterator of chunks of data.
         Returns less than L bytes if the iterator runs dry."""
         left = l
         buf = ''
         queue = self._queue
         while left > 0:
             # refill the queue
             if not queue:
                 target = 2**18
                 for chunk in self.iter:
                     queue.append(chunk)
                     target -= len(chunk)
                     if target <= 0:
                         break
                 if not queue:
                     break
 
             chunk = queue.pop(0)
             left -= len(chunk)
             if left < 0:
                 queue.insert(0, chunk[left:])
                 buf += chunk[:left]
             else:
                 buf += chunk
 
         return buf
 
 def filechunkiter(f, size=65536, limit=None):
     """Create a generator that produces the data in the file size
     (default 65536) bytes at a time, up to optional limit (default is
     to read all data).  Chunks may be less than size bytes if the
     chunk is the last chunk in the file, or the file is a socket or
     some other type of file that sometimes reads less data than is
     requested."""
     assert size >= 0
     assert limit is None or limit >= 0
     while True:
         if limit is None:
             nbytes = size
         else:
             nbytes = min(limit, size)
         s = nbytes and f.read(nbytes)
         if not s:
             break
         if limit:
             limit -= len(s)
         yield s
 
 def makedate():
     lt = time.localtime()
     if lt[8] == 1 and time.daylight:
         tz = time.altzone
     else:
         tz = time.timezone
     t = time.mktime(lt)
     if t < 0:
         hint = _("check your clock")
         raise Abort(_("negative timestamp: %d") % t, hint=hint)
     return t, tz
 
 def datestr(date=None, format='%a %b %d %H:%M:%S %Y %1%2'):
     """represent a (unixtime, offset) tuple as a localized time.
     unixtime is seconds since the epoch, and offset is the time zone's
     number of seconds away from UTC. if timezone is false, do not
     append time zone to string."""
     t, tz = date or makedate()
     if t < 0:
         t = 0   # time.gmtime(lt) fails on Windows for lt < -43200
         tz = 0
     if "%1" in format or "%2" in format:
         sign = (tz > 0) and "-" or "+"
         minutes = abs(tz) // 60
         format = format.replace("%1", "%c%02d" % (sign, minutes // 60))
         format = format.replace("%2", "%02d" % (minutes % 60))
     s = time.strftime(format, time.gmtime(float(t) - tz))
     return s
 
 def shortdate(date=None):
     """turn (timestamp, tzoff) tuple into iso 8631 date."""
     return datestr(date, format='%Y-%m-%d')
 
 def strdate(string, format, defaults=[]):
     """parse a localized time string and return a (unixtime, offset) tuple.
     if the string cannot be parsed, ValueError is raised."""
     def timezone(string):
         tz = string.split()[-1]
         if tz[0] in "+-" and len(tz) == 5 and tz[1:].isdigit():
             sign = (tz[0] == "+") and 1 or -1
             hours = int(tz[1:3])
             minutes = int(tz[3:5])
             return -sign * (hours * 60 + minutes) * 60
         if tz == "GMT" or tz == "UTC":
             return 0
         return None
 
     # NOTE: unixtime = localunixtime + offset
     offset, date = timezone(string), string
     if offset is not None:
         date = " ".join(string.split()[:-1])
 
     # add missing elements from defaults
     usenow = False # default to using biased defaults
     for part in ("S", "M", "HI", "d", "mb", "yY"): # decreasing specificity
         found = [True for p in part if ("%"+p) in format]
         if not found:
             date += "@" + defaults[part][usenow]
             format += "@%" + part[0]
         else:
             # We've found a specific time element, less specific time
             # elements are relative to today
             usenow = True
 
     timetuple = time.strptime(date, format)
     localunixtime = int(calendar.timegm(timetuple))
     if offset is None:
         # local timezone
         unixtime = int(time.mktime(timetuple))
         offset = unixtime - localunixtime
     else:
         unixtime = localunixtime + offset
     return unixtime, offset
 
 def parsedate(date, formats=None, bias={}):
     """parse a localized date/time and return a (unixtime, offset) tuple.
 
     The date may be a "unixtime offset" string or in one of the specified
     formats. If the date already is a (unixtime, offset) tuple, it is returned.
     """
     if not date:
         return 0, 0
     if isinstance(date, tuple) and len(date) == 2:
         return date
     if not formats:
         formats = defaultdateformats
     date = date.strip()
     try:
         when, offset = map(int, date.split(' '))
     except ValueError:
         # fill out defaults
         now = makedate()
         defaults = {}
         for part in ("d", "mb", "yY", "HI", "M", "S"):
             # this piece is for rounding the specific end of unknowns
             b = bias.get(part)
             if b is None:
                 if part[0] in "HMS":
                     b = "00"
                 else:
                     b = "0"
 
             # this piece is for matching the generic end to today's date
             n = datestr(now, "%" + part[0])
 
             defaults[part] = (b, n)
 
         for format in formats:
             try:
                 when, offset = strdate(date, format, defaults)
             except (ValueError, OverflowError):
                 pass
             else:
                 break
         else:
             raise Abort(_('invalid date: %r') % date)
     # validate explicit (probably user-specified) date and
     # time zone offset. values must fit in signed 32 bits for
     # current 32-bit linux runtimes. timezones go from UTC-12
     # to UTC+14
     if abs(when) > 0x7fffffff:
         raise Abort(_('date exceeds 32 bits: %d') % when)
     if when < 0:
         raise Abort(_('negative date value: %d') % when)
     if offset < -50400 or offset > 43200:
         raise Abort(_('impossible time zone offset: %d') % offset)
     return when, offset
 
 def matchdate(date):
     """Return a function that matches a given date match specifier
 
     Formats include:
 
     '{date}' match a given date to the accuracy provided
 
     '<{date}' on or before a given date
 
     '>{date}' on or after a given date
 
     >>> p1 = parsedate("10:29:59")
     >>> p2 = parsedate("10:30:00")
     >>> p3 = parsedate("10:30:59")
     >>> p4 = parsedate("10:31:00")
     >>> p5 = parsedate("Sep 15 10:30:00 1999")
     >>> f = matchdate("10:30")
     >>> f(p1[0])
     False
     >>> f(p2[0])
     True
     >>> f(p3[0])
     True
     >>> f(p4[0])
     False
     >>> f(p5[0])
     False
     """
 
     def lower(date):
         d = dict(mb="1", d="1")
         return parsedate(date, extendeddateformats, d)[0]
 
     def upper(date):
         d = dict(mb="12", HI="23", M="59", S="59")
         for days in ("31", "30", "29"):
             try:
                 d["d"] = days
                 return parsedate(date, extendeddateformats, d)[0]
             except:
                 pass
         d["d"] = "28"
         return parsedate(date, extendeddateformats, d)[0]
 
     date = date.strip()
 
     if not date:
         raise Abort(_("dates cannot consist entirely of whitespace"))
     elif date[0] == "<":
         if not date[1:]:
             raise Abort(_("invalid day spec, use '<DATE'"))
         when = upper(date[1:])
         return lambda x: x <= when
     elif date[0] == ">":
         if not date[1:]:
             raise Abort(_("invalid day spec, use '>DATE'"))
         when = lower(date[1:])
         return lambda x: x >= when
     elif date[0] == "-":
         try:
             days = int(date[1:])
         except ValueError:
             raise Abort(_("invalid day spec: %s") % date[1:])
         if days < 0:
             raise Abort(_("%s must be nonnegative (see 'hg help dates')")
                 % date[1:])
         when = makedate()[0] - days * 3600 * 24
         return lambda x: x >= when
     elif " to " in date:
         a, b = date.split(" to ")
         start, stop = lower(a), upper(b)
         return lambda x: x >= start and x <= stop
     else:
         start, stop = lower(date), upper(date)
         return lambda x: x >= start and x <= stop
 
 def shortuser(user):
     """Return a short representation of a user name or email address."""
     f = user.find('@')
     if f >= 0:
         user = user[:f]
     f = user.find('<')
     if f >= 0:
         user = user[f + 1:]
     f = user.find(' ')
     if f >= 0:
         user = user[:f]
     f = user.find('.')
     if f >= 0:
         user = user[:f]
     return user
 
 def email(author):
     '''get email of author.'''
     r = author.find('>')
     if r == -1:
         r = None
     return author[author.find('<') + 1:r]
 
 def _ellipsis(text, maxlength):
     if len(text) <= maxlength:
         return text, False
     else:
         return "%s..." % (text[:maxlength - 3]), True
 
 def ellipsis(text, maxlength=400):
     """Trim string to at most maxlength (default: 400) characters."""
     try:
         # use unicode not to split at intermediate multi-byte sequence
         utext, truncated = _ellipsis(text.decode(encoding.encoding),
                                      maxlength)
         if not truncated:
             return text
         return utext.encode(encoding.encoding)
     except (UnicodeDecodeError, UnicodeEncodeError):
         return _ellipsis(text, maxlength)[0]
 
 def bytecount(nbytes):
     '''return byte count formatted as readable string, with units'''
 
     units = (
         (100, 1 << 30, _('%.0f GB')),
         (10, 1 << 30, _('%.1f GB')),
         (1, 1 << 30, _('%.2f GB')),
         (100, 1 << 20, _('%.0f MB')),
         (10, 1 << 20, _('%.1f MB')),
         (1, 1 << 20, _('%.2f MB')),
         (100, 1 << 10, _('%.0f KB')),
         (10, 1 << 10, _('%.1f KB')),
         (1, 1 << 10, _('%.2f KB')),
         (1, 1, _('%.0f bytes')),
         )
 
     for multiplier, divisor, format in units:
         if nbytes >= divisor * multiplier:
             return format % (nbytes / float(divisor))
     return units[-1][2] % nbytes
 
 def uirepr(s):
     # Avoid double backslash in Windows path repr()
     return repr(s).replace('\\\\', '\\')
 
 # delay import of textwrap
 def MBTextWrapper(**kwargs):
     class tw(textwrap.TextWrapper):
         """
         Extend TextWrapper for double-width characters.
 
         Some Asian characters use two terminal columns instead of one.
         A good example of this behavior can be seen with u'\u65e5\u672c',
         the two Japanese characters for "Japan":
         len() returns 2, but when printed to a terminal, they eat 4 columns.
 
         (Note that this has nothing to do whatsoever with unicode
         representation, or encoding of the underlying string)
         """
         def __init__(self, **kwargs):
             textwrap.TextWrapper.__init__(self, **kwargs)
 
         def _cutdown(self, str, space_left):
             l = 0
             ucstr = unicode(str, encoding.encoding)
             colwidth = unicodedata.east_asian_width
             for i in xrange(len(ucstr)):
                 l += colwidth(ucstr[i]) in 'WFA' and 2 or 1
                 if space_left < l:
                     return (ucstr[:i].encode(encoding.encoding),
                             ucstr[i:].encode(encoding.encoding))
             return str, ''
 
         # overriding of base class
         def _handle_long_word(self, reversed_chunks, cur_line, cur_len, width):
             space_left = max(width - cur_len, 1)
 
             if self.break_long_words:
                 cut, res = self._cutdown(reversed_chunks[-1], space_left)
                 cur_line.append(cut)
                 reversed_chunks[-1] = res
             elif not cur_line:
                 cur_line.append(reversed_chunks.pop())
 
     global MBTextWrapper
     MBTextWrapper = tw
     return tw(**kwargs)
 
 def wrap(line, width, initindent='', hangindent=''):
     maxindent = max(len(hangindent), len(initindent))
     if width <= maxindent:
         # adjust for weird terminal size
         width = max(78, maxindent + 1)
     wrapper = MBTextWrapper(width=width,
                             initial_indent=initindent,
                             subsequent_indent=hangindent)
     return wrapper.fill(line)
 
 def iterlines(iterator):
     for chunk in iterator:
         for line in chunk.splitlines():
             yield line
 
 def expandpath(path):
     return os.path.expanduser(os.path.expandvars(path))
 
 def hgcmd():
     """Return the command used to execute current hg
 
     This is different from hgexecutable() because on Windows we want
     to avoid things opening new shell windows like batch files, so we
     get either the python call or current executable.
     """
     if mainfrozen():
         return [sys.executable]
     return gethgcmd()
 
 def rundetached(args, condfn):
     """Execute the argument list in a detached process.
 
     condfn is a callable which is called repeatedly and should return
     True once the child process is known to have started successfully.
     At this point, the child process PID is returned. If the child
     process fails to start or finishes before condfn() evaluates to
     True, return -1.
     """
     # Windows case is easier because the child process is either
     # successfully starting and validating the condition or exiting
     # on failure. We just poll on its PID. On Unix, if the child
     # process fails to start, it will be left in a zombie state until
     # the parent wait on it, which we cannot do since we expect a long
     # running process on success. Instead we listen for SIGCHLD telling
     # us our child process terminated.
     terminated = set()
     def handler(signum, frame):
         terminated.add(os.wait())
     prevhandler = None
     if hasattr(signal, 'SIGCHLD'):
         prevhandler = signal.signal(signal.SIGCHLD, handler)
     try:
         pid = spawndetached(args)
         while not condfn():
             if ((pid in terminated or not testpid(pid))
                 and not condfn()):
                 return -1
             time.sleep(0.1)
         return pid
     finally:
         if prevhandler is not None:
             signal.signal(signal.SIGCHLD, prevhandler)
 
 try:
     any, all = any, all
 except NameError:
     def any(iterable):
         for i in iterable:
             if i:
                 return True
         return False
 
     def all(iterable):
         for i in iterable:
             if not i:
                 return False
         return True
 
 def interpolate(prefix, mapping, s, fn=None, escape_prefix=False):
     """Return the result of interpolating items in the mapping into string s.
 
     prefix is a single character string, or a two character string with
     a backslash as the first character if the prefix needs to be escaped in
     a regular expression.
 
     fn is an optional function that will be applied to the replacement text
     just before replacement.
 
     escape_prefix is an optional flag that allows using doubled prefix for
     its escaping.
     """
     fn = fn or (lambda s: s)
     patterns = '|'.join(mapping.keys())
     if escape_prefix:
         patterns += '|' + prefix
         if len(prefix) > 1:
             prefix_char = prefix[1:]
         else:
             prefix_char = prefix
         mapping[prefix_char] = prefix_char
     r = re.compile(r'%s(%s)' % (prefix, patterns))
     return r.sub(lambda x: fn(mapping[x.group()[1:]]), s)
 
 def getport(port):
     """Return the port for a given network service.
 
     If port is an integer, it's returned as is. If it's a string, it's
     looked up using socket.getservbyname(). If there's no matching
     service, util.Abort is raised.
     """
     try:
         return int(port)
     except ValueError:
         pass
 
     try:
         return socket.getservbyname(port)
     except socket.error:
         raise Abort(_("no port number associated with service '%s'") % port)
 
 _booleans = {'1': True, 'yes': True, 'true': True, 'on': True, 'always': True,
              '0': False, 'no': False, 'false': False, 'off': False,
              'never': False}
 
 def parsebool(s):
     """Parse s into a boolean.
 
     If s is not a valid boolean, returns None.
     """
     return _booleans.get(s.lower(), None)
 
 _hexdig = '0123456789ABCDEFabcdef'
 _hextochr = dict((a + b, chr(int(a + b, 16)))
                  for a in _hexdig for b in _hexdig)
 
 def _urlunquote(s):
     """unquote('abc%20def') -> 'abc def'."""
     res = s.split('%')
     # fastpath
     if len(res) == 1:
         return s
     s = res[0]
     for item in res[1:]:
         try:
             s += _hextochr[item[:2]] + item[2:]
         except KeyError:
             s += '%' + item
         except UnicodeDecodeError:
             s += unichr(int(item[:2], 16)) + item[2:]
     return s
 
 class url(object):
     r"""Reliable URL parser.
 
     This parses URLs and provides attributes for the following
     components:
 
     <scheme>://<user>:<passwd>@<host>:<port>/<path>?<query>#<fragment>
 
     Missing components are set to None. The only exception is
     fragment, which is set to '' if present but empty.
 
     If parsefragment is False, fragment is included in query. If
     parsequery is False, query is included in path. If both are
     False, both fragment and query are included in path.
 
     See http://www.ietf.org/rfc/rfc2396.txt for more information.
 
     Note that for backward compatibility reasons, bundle URLs do not
     take host names. That means 'bundle://../' has a path of '../'.
 
     Examples:
 
     >>> url('http://www.ietf.org/rfc/rfc2396.txt')
     <url scheme: 'http', host: 'www.ietf.org', path: 'rfc/rfc2396.txt'>
     >>> url('ssh://[::1]:2200//home/joe/repo')
     <url scheme: 'ssh', host: '[::1]', port: '2200', path: '/home/joe/repo'>
     >>> url('file:///home/joe/repo')
     <url scheme: 'file', path: '/home/joe/repo'>
     >>> url('bundle:foo')
     <url scheme: 'bundle', path: 'foo'>
     >>> url('bundle://../foo')
     <url scheme: 'bundle', path: '../foo'>
     >>> url(r'c:\foo\bar')
     <url path: 'c:\\foo\\bar'>
     >>> url(r'\\blah\blah\blah')
     <url path: '\\\\blah\\blah\\blah'>
 
     Authentication credentials:
 
     >>> url('ssh://joe:xyz@x/repo')
     <url scheme: 'ssh', user: 'joe', passwd: 'xyz', host: 'x', path: 'repo'>
     >>> url('ssh://joe@x/repo')
     <url scheme: 'ssh', user: 'joe', host: 'x', path: 'repo'>
 
     Query strings and fragments:
 
     >>> url('http://host/a?b#c')
     <url scheme: 'http', host: 'host', path: 'a', query: 'b', fragment: 'c'>
     >>> url('http://host/a?b#c', parsequery=False, parsefragment=False)
     <url scheme: 'http', host: 'host', path: 'a?b#c'>
     """
 
     _safechars = "!~*'()+"
     _safepchars = "/!~*'()+"
     _matchscheme = re.compile(r'^[a-zA-Z0-9+.\-]+:').match
 
     def __init__(self, path, parsequery=True, parsefragment=True):
         # We slowly chomp away at path until we have only the path left
         self.scheme = self.user = self.passwd = self.host = None
         self.port = self.path = self.query = self.fragment = None
         self._localpath = True
         self._hostport = ''
         self._origpath = path
 
         # special case for Windows drive letters and UNC paths
         if hasdriveletter(path) or path.startswith(r'\\'):
             self.path = path
             return
 
         # For compatibility reasons, we can't handle bundle paths as
         # normal URLS
         if path.startswith('bundle:'):
             self.scheme = 'bundle'
             path = path[7:]
             if path.startswith('//'):
                 path = path[2:]
             self.path = path
             return
 
         if self._matchscheme(path):
             parts = path.split(':', 1)
             if parts[0]:
                 self.scheme, path = parts
                 self._localpath = False
 
         if not path:
             path = None
             if self._localpath:
                 self.path = ''
                 return
         else:
             if parsefragment and '#' in path:
                 path, self.fragment = path.split('#', 1)
                 if not path:
                     path = None
             if self._localpath:
                 self.path = path
                 return
 
             if parsequery and '?' in path:
                 path, self.query = path.split('?', 1)
                 if not path:
                     path = None
                 if not self.query:
                     self.query = None
 
             # // is required to specify a host/authority
             if path and path.startswith('//'):
                 parts = path[2:].split('/', 1)
                 if len(parts) > 1:
                     self.host, path = parts
                     path = path
                 else:
                     self.host = parts[0]
                     path = None
                 if not self.host:
                     self.host = None
                     if path:
                         path = '/' + path
 
             if self.host and '@' in self.host:
                 self.user, self.host = self.host.rsplit('@', 1)
                 if ':' in self.user:
                     self.user, self.passwd = self.user.split(':', 1)
                 if not self.host:
                     self.host = None
 
             # Don't split on colons in IPv6 addresses without ports
             if (self.host and ':' in self.host and
                 not (self.host.startswith('[') and self.host.endswith(']'))):
                 self._hostport = self.host
                 self.host, self.port = self.host.rsplit(':', 1)
                 if not self.host:
                     self.host = None
 
             if (self.host and self.scheme == 'file' and
                 self.host not in ('localhost', '127.0.0.1', '[::1]')):
                 raise Abort(_('file:// URLs can only refer to localhost'))
 
         self.path = path
 
         for a in ('user', 'passwd', 'host', 'port',
                   'path', 'query', 'fragment'):
             v = getattr(self, a)
             if v is not None:
                 setattr(self, a, _urlunquote(v))
 
     def __repr__(self):
         attrs = []
         for a in ('scheme', 'user', 'passwd', 'host', 'port', 'path',
                   'query', 'fragment'):
             v = getattr(self, a)
             if v is not None:
                 attrs.append('%s: %r' % (a, v))
         return '<url %s>' % ', '.join(attrs)
 
     def __str__(self):
         r"""Join the URL's components back into a URL string.
 
         Examples:
 
         >>> str(url('http://user:pw@host:80/?foo#bar'))
         'http://user:pw@host:80/?foo#bar'
         >>> str(url('ssh://user:pw@[::1]:2200//home/joe#'))
         'ssh://user:pw@[::1]:2200//home/joe#'
         >>> str(url('http://localhost:80//'))
         'http://localhost:80//'
         >>> str(url('http://localhost:80/'))
         'http://localhost:80/'
         >>> str(url('http://localhost:80'))
         'http://localhost:80/'
         >>> str(url('bundle:foo'))
         'bundle:foo'
         >>> str(url('bundle://../foo'))
         'bundle:../foo'
         >>> str(url('path'))
         'path'
         >>> str(url('file:///tmp/foo/bar'))
         'file:///tmp/foo/bar'
         >>> print url(r'bundle:foo\bar')
         bundle:foo\bar
         """
         if self._localpath:
             s = self.path
             if self.scheme == 'bundle':
                 s = 'bundle:' + s
             if self.fragment:
                 s += '#' + self.fragment
             return s
 
         s = self.scheme + ':'
         if self.user or self.passwd or self.host:
             s += '//'
         elif self.scheme and (not self.path or self.path.startswith('/')):
             s += '//'
         if self.user:
             s += urllib.quote(self.user, safe=self._safechars)
         if self.passwd:
             s += ':' + urllib.quote(self.passwd, safe=self._safechars)
         if self.user or self.passwd:
             s += '@'
         if self.host:
             if not (self.host.startswith('[') and self.host.endswith(']')):
                 s += urllib.quote(self.host)
             else:
                 s += self.host
         if self.port:
             s += ':' + urllib.quote(self.port)
         if self.host:
             s += '/'
         if self.path:
             s += urllib.quote(self.path, safe=self._safepchars)
         if self.query:
             s += '?' + urllib.quote(self.query, safe=self._safepchars)
         if self.fragment is not None:
             s += '#' + urllib.quote(self.fragment, safe=self._safepchars)
         return s
 
     def authinfo(self):
         user, passwd = self.user, self.passwd
         try:
             self.user, self.passwd = None, None
             s = str(self)
         finally:
             self.user, self.passwd = user, passwd
         if not self.user:
             return (s, None)
         return (s, (None, (str(self), self.host),
                     self.user, self.passwd or ''))
 
+    def isabs(self):
+        if self.scheme and self.scheme != 'file':
+            return True # remote URL
+        if hasdriveletter(self.path):
+            return True # absolute for our purposes - can't be joined()
+        if self.path.startswith(r'\\'):
+            return True # Windows UNC path
+        if self.path.startswith('/'):
+            return True # POSIX-style
+        return False
+
     def localpath(self):
         if self.scheme == 'file' or self.scheme == 'bundle':
             path = self.path or '/'
             # For Windows, we need to promote hosts containing drive
             # letters to paths with drive letters.
             if hasdriveletter(self._hostport):
                 path = self._hostport + '/' + self.path
             elif self.host is not None and self.path:
                 path = '/' + path
             # We also need to handle the case of file:///C:/, which
             # should return C:/, not /C:/.
             elif hasdriveletter(path):
                 # Strip leading slash from paths with drive names
                 return path[1:]
             return path
         return self._origpath
 
 def hasscheme(path):
     return bool(url(path).scheme)
 
 def hasdriveletter(path):
     return path[1:2] == ':' and path[0:1].isalpha()
 
 def localpath(path):
     return url(path, parsequery=False, parsefragment=False).localpath()
 
 def hidepassword(u):
     '''hide user credential in a url string'''
     u = url(u)
     if u.passwd:
         u.passwd = '***'
     return str(u)
 
 def removeauth(u):
     '''remove all authentication information from a url string'''
     u = url(u)
     u.user = u.passwd = None
     return str(u)
 
 def isatty(fd):
     try:
         return fd.isatty()
     except AttributeError:
         return False