obsolete: introduce caches for all meaningful sets

This changeset introduces caches on the `obsstore` that keeps track of sets of
revisions meaningful for obsolescence related logics. For now they are:

- obsolete: changesets used as precursors (and not public),
- extinct:  obsolete changesets with osbolete descendants only,
- unstable: non obsolete changesets with obsolete ancestors.

The cache is accessed using the `getobscache(repo, '<set-name>')` function which
builds the cache on demand. The `clearobscaches(repo)` function takes care of
clearing the caches if any.

Caches are cleared when one of these events happens:

- a new marker is added,
- a new changeset is added,
- some changesets are made public,
- some public changesets are demoted to draft or secret.

Declaration of more sets is made easy because we will have to handle at least
two other "troubles" (latecomer and conflicting).

Caches are now used by revset and changectx. It is usually not much more
expensive to compute the whole set than to check the property of a few elements.
The performance boost is welcome in case we apply obsolescence logic on a lot of
revisions. This makes the feature usable!

mercurial/context.py

@@ -1,1313 +1,1291 b''
 # context.py - changeset and file context objects for mercurial
 #
 # Copyright 2006, 2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from node import nullid, nullrev, short, hex, bin
 from i18n import _
 import ancestor, mdiff, error, util, scmutil, subrepo, patch, encoding, phases
 import copies
 import match as matchmod
 import os, errno, stat
+import obsolete as obsmod
 
 propertycache = util.propertycache
 
 class changectx(object):
     """A changecontext object makes access to data related to a particular
     changeset convenient."""
     def __init__(self, repo, changeid=''):
         """changeid is a revision number, node, or tag"""
         if changeid == '':
             changeid = '.'
         self._repo = repo
 
         if isinstance(changeid, int):
             self._rev = changeid
             self._node = repo.changelog.node(changeid)
             return
         if isinstance(changeid, long):
             changeid = str(changeid)
         if changeid == '.':
             self._node = repo.dirstate.p1()
             self._rev = repo.changelog.rev(self._node)
             return
         if changeid == 'null':
             self._node = nullid
             self._rev = nullrev
             return
         if changeid == 'tip':
             self._rev = len(repo.changelog) - 1
             self._node = repo.changelog.node(self._rev)
             return
         if len(changeid) == 20:
             try:
                 self._node = changeid
                 self._rev = repo.changelog.rev(changeid)
                 return
             except LookupError:
                 pass
 
         try:
             r = int(changeid)
             if str(r) != changeid:
                 raise ValueError
             l = len(repo.changelog)
             if r < 0:
                 r += l
             if r < 0 or r >= l:
                 raise ValueError
             self._rev = r
             self._node = repo.changelog.node(r)
             return
         except (ValueError, OverflowError):
             pass
 
         if len(changeid) == 40:
             try:
                 self._node = bin(changeid)
                 self._rev = repo.changelog.rev(self._node)
                 return
             except (TypeError, LookupError):
                 pass
 
         if changeid in repo._bookmarks:
             self._node = repo._bookmarks[changeid]
             self._rev = repo.changelog.rev(self._node)
             return
         if changeid in repo._tagscache.tags:
             self._node = repo._tagscache.tags[changeid]
             self._rev = repo.changelog.rev(self._node)
             return
         try:
             self._node = repo.branchtip(changeid)
             self._rev = repo.changelog.rev(self._node)
             return
         except error.RepoLookupError:
             pass
 
         self._node = repo.changelog._partialmatch(changeid)
         if self._node is not None:
             self._rev = repo.changelog.rev(self._node)
             return
 
         # lookup failed
         # check if it might have come from damaged dirstate
         if changeid in repo.dirstate.parents():
             raise error.Abort(_("working directory has unknown parent '%s'!")
                               % short(changeid))
         try:
             if len(changeid) == 20:
                 changeid = hex(changeid)
         except TypeError:
             pass
         raise error.RepoLookupError(
             _("unknown revision '%s'") % changeid)
 
     def __str__(self):
         return short(self.node())
 
     def __int__(self):
         return self.rev()
 
     def __repr__(self):
         return "<changectx %s>" % str(self)
 
     def __hash__(self):
         try:
             return hash(self._rev)
         except AttributeError:
             return id(self)
 
     def __eq__(self, other):
         try:
             return self._rev == other._rev
         except AttributeError:
             return False
 
     def __ne__(self, other):
         return not (self == other)
 
     def __nonzero__(self):
         return self._rev != nullrev
 
     @propertycache
     def _changeset(self):
         return self._repo.changelog.read(self.rev())
 
     @propertycache
     def _manifest(self):
         return self._repo.manifest.read(self._changeset[0])
 
     @propertycache
     def _manifestdelta(self):
         return self._repo.manifest.readdelta(self._changeset[0])
 
     @propertycache
     def _parents(self):
         p = self._repo.changelog.parentrevs(self._rev)
         if p[1] == nullrev:
             p = p[:-1]
         return [changectx(self._repo, x) for x in p]
 
     @propertycache
     def substate(self):
         return subrepo.state(self, self._repo.ui)
 
     def __contains__(self, key):
         return key in self._manifest
 
     def __getitem__(self, key):
         return self.filectx(key)
 
     def __iter__(self):
         for f in sorted(self._manifest):
             yield f
 
     def changeset(self):
         return self._changeset
     def manifest(self):
         return self._manifest
     def manifestnode(self):
         return self._changeset[0]
 
     def rev(self):
         return self._rev
     def node(self):
         return self._node
     def hex(self):
         return hex(self._node)
     def user(self):
         return self._changeset[1]
     def date(self):
         return self._changeset[2]
     def files(self):
         return self._changeset[3]
     def description(self):
         return self._changeset[4]
     def branch(self):
         return encoding.tolocal(self._changeset[5].get("branch"))
     def closesbranch(self):
         return 'close' in self._changeset[5]
     def extra(self):
         return self._changeset[5]
     def tags(self):
         return self._repo.nodetags(self._node)
     def bookmarks(self):
         return self._repo.nodebookmarks(self._node)
     def phase(self):
         return self._repo._phasecache.phase(self._repo, self._rev)
     def phasestr(self):
         return phases.phasenames[self.phase()]
     def mutable(self):
         return self.phase() > phases.public
     def hidden(self):
         return self._rev in self._repo.hiddenrevs
 
     def parents(self):
         """return contexts for each parent changeset"""
         return self._parents
 
     def p1(self):
         return self._parents[0]
 
     def p2(self):
         if len(self._parents) == 2:
             return self._parents[1]
         return changectx(self._repo, -1)
 
     def children(self):
         """return contexts for each child changeset"""
         c = self._repo.changelog.children(self._node)
         return [changectx(self._repo, x) for x in c]
 
     def ancestors(self):
         for a in self._repo.changelog.ancestors([self._rev]):
             yield changectx(self._repo, a)
 
     def descendants(self):
         for d in self._repo.changelog.descendants([self._rev]):
             yield changectx(self._repo, d)
 
     def obsolete(self):
         """True if the changeset is obsolete"""
-        return (self.node() in self._repo.obsstore.precursors
-                and self.phase() > phases.public)
+        return self.rev() in obsmod.getobscache(self._repo, 'obsolete')
 
     def extinct(self):
         """True if the changeset is extinct"""
-        # We should just compute a cache and check against it.
-        # See revset implementation for details.
-        #
-        # But this naive implementation does not require cache
-        if self.phase() <= phases.public:
-            return False
-        if not self.obsolete():
-            return False
-        for desc in self.descendants():
-            if not desc.obsolete():
-                return False
-        return True
+        return self.rev() in obsmod.getobscache(self._repo, 'extinct')
 
     def unstable(self):
         """True if the changeset is not obsolete but it's ancestor are"""
-        # We should just compute /(obsolete()::) - obsolete()/
-        # and keep it in a cache.
-        #
-        # But this naive implementation does not require cache
-        if self.phase() <= phases.public:
-            return False
-        if self.obsolete():
-            return False
-        for anc in self.ancestors():
-            if anc.obsolete():
-                return True
-        return False
+        return self.rev() in obsmod.getobscache(self._repo, 'unstable')
 
     def _fileinfo(self, path):
         if '_manifest' in self.__dict__:
             try:
                 return self._manifest[path], self._manifest.flags(path)
             except KeyError:
                 raise error.LookupError(self._node, path,
                                         _('not found in manifest'))
         if '_manifestdelta' in self.__dict__ or path in self.files():
             if path in self._manifestdelta:
                 return (self._manifestdelta[path],
                         self._manifestdelta.flags(path))
         node, flag = self._repo.manifest.find(self._changeset[0], path)
         if not node:
             raise error.LookupError(self._node, path,
                                     _('not found in manifest'))
 
         return node, flag
 
     def filenode(self, path):
         return self._fileinfo(path)[0]
 
     def flags(self, path):
         try:
             return self._fileinfo(path)[1]
         except error.LookupError:
             return ''
 
     def filectx(self, path, fileid=None, filelog=None):
         """get a file context from this changeset"""
         if fileid is None:
             fileid = self.filenode(path)
         return filectx(self._repo, path, fileid=fileid,
                        changectx=self, filelog=filelog)
 
     def ancestor(self, c2):
         """
         return the ancestor context of self and c2
         """
         # deal with workingctxs
         n2 = c2._node
         if n2 is None:
             n2 = c2._parents[0]._node
         n = self._repo.changelog.ancestor(self._node, n2)
         return changectx(self._repo, n)
 
     def walk(self, match):
         fset = set(match.files())
         # for dirstate.walk, files=['.'] means "walk the whole tree".
         # follow that here, too
         fset.discard('.')
         for fn in self:
             if fn in fset:
                 # specified pattern is the exact name
                 fset.remove(fn)
             if match(fn):
                 yield fn
         for fn in sorted(fset):
             if fn in self._dirs:
                 # specified pattern is a directory
                 continue
             if match.bad(fn, _('no such file in rev %s') % self) and match(fn):
                 yield fn
 
     def sub(self, path):
         return subrepo.subrepo(self, path)
 
     def match(self, pats=[], include=None, exclude=None, default='glob'):
         r = self._repo
         return matchmod.match(r.root, r.getcwd(), pats,
                               include, exclude, default,
                               auditor=r.auditor, ctx=self)
 
     def diff(self, ctx2=None, match=None, **opts):
         """Returns a diff generator for the given contexts and matcher"""
         if ctx2 is None:
             ctx2 = self.p1()
         if ctx2 is not None and not isinstance(ctx2, changectx):
             ctx2 = self._repo[ctx2]
         diffopts = patch.diffopts(self._repo.ui, opts)
         return patch.diff(self._repo, ctx2.node(), self.node(),
                           match=match, opts=diffopts)
 
     @propertycache
     def _dirs(self):
         dirs = set()
         for f in self._manifest:
             pos = f.rfind('/')
             while pos != -1:
                 f = f[:pos]
                 if f in dirs:
                     break # dirs already contains this and above
                 dirs.add(f)
                 pos = f.rfind('/')
         return dirs
 
     def dirs(self):
         return self._dirs
 
 class filectx(object):
     """A filecontext object makes access to data related to a particular
        filerevision convenient."""
     def __init__(self, repo, path, changeid=None, fileid=None,
                  filelog=None, changectx=None):
         """changeid can be a changeset revision, node, or tag.
            fileid can be a file revision or node."""
         self._repo = repo
         self._path = path
 
         assert (changeid is not None
                 or fileid is not None
                 or changectx is not None), \
                 ("bad args: changeid=%r, fileid=%r, changectx=%r"
                  % (changeid, fileid, changectx))
 
         if filelog:
             self._filelog = filelog
 
         if changeid is not None:
             self._changeid = changeid
         if changectx is not None:
             self._changectx = changectx
         if fileid is not None:
             self._fileid = fileid
 
     @propertycache
     def _changectx(self):
         return changectx(self._repo, self._changeid)
 
     @propertycache
     def _filelog(self):
         return self._repo.file(self._path)
 
     @propertycache
     def _changeid(self):
         if '_changectx' in self.__dict__:
             return self._changectx.rev()
         else:
             return self._filelog.linkrev(self._filerev)
 
     @propertycache
     def _filenode(self):
         if '_fileid' in self.__dict__:
             return self._filelog.lookup(self._fileid)
         else:
             return self._changectx.filenode(self._path)
 
     @propertycache
     def _filerev(self):
         return self._filelog.rev(self._filenode)
 
     @propertycache
     def _repopath(self):
         return self._path
 
     def __nonzero__(self):
         try:
             self._filenode
             return True
         except error.LookupError:
             # file is missing
             return False
 
     def __str__(self):
         return "%s@%s" % (self.path(), short(self.node()))
 
     def __repr__(self):
         return "<filectx %s>" % str(self)
 
     def __hash__(self):
         try:
             return hash((self._path, self._filenode))
         except AttributeError:
             return id(self)
 
     def __eq__(self, other):
         try:
             return (self._path == other._path
                     and self._filenode == other._filenode)
         except AttributeError:
             return False
 
     def __ne__(self, other):
         return not (self == other)
 
     def filectx(self, fileid):
         '''opens an arbitrary revision of the file without
         opening a new filelog'''
         return filectx(self._repo, self._path, fileid=fileid,
                        filelog=self._filelog)
 
     def filerev(self):
         return self._filerev
     def filenode(self):
         return self._filenode
     def flags(self):
         return self._changectx.flags(self._path)
     def filelog(self):
         return self._filelog
 
     def rev(self):
         if '_changectx' in self.__dict__:
             return self._changectx.rev()
         if '_changeid' in self.__dict__:
             return self._changectx.rev()
         return self._filelog.linkrev(self._filerev)
 
     def linkrev(self):
         return self._filelog.linkrev(self._filerev)
     def node(self):
         return self._changectx.node()
     def hex(self):
         return hex(self.node())
     def user(self):
         return self._changectx.user()
     def date(self):
         return self._changectx.date()
     def files(self):
         return self._changectx.files()
     def description(self):
         return self._changectx.description()
     def branch(self):
         return self._changectx.branch()
     def extra(self):
         return self._changectx.extra()
     def manifest(self):
         return self._changectx.manifest()
     def changectx(self):
         return self._changectx
 
     def data(self):
         return self._filelog.read(self._filenode)
     def path(self):
         return self._path
     def size(self):
         return self._filelog.size(self._filerev)
 
     def isbinary(self):
         try:
             return util.binary(self.data())
         except IOError:
             return False
 
     def cmp(self, fctx):
         """compare with other file context
 
         returns True if different than fctx.
         """
         if (fctx._filerev is None
             and (self._repo._encodefilterpats
                  # if file data starts with '\1\n', empty metadata block is
                  # prepended, which adds 4 bytes to filelog.size().
                  or self.size() - 4 == fctx.size())
             or self.size() == fctx.size()):
             return self._filelog.cmp(self._filenode, fctx.data())
 
         return True
 
     def renamed(self):
         """check if file was actually renamed in this changeset revision
 
         If rename logged in file revision, we report copy for changeset only
         if file revisions linkrev points back to the changeset in question
         or both changeset parents contain different file revisions.
         """
 
         renamed = self._filelog.renamed(self._filenode)
         if not renamed:
             return renamed
 
         if self.rev() == self.linkrev():
             return renamed
 
         name = self.path()
         fnode = self._filenode
         for p in self._changectx.parents():
             try:
                 if fnode == p.filenode(name):
                     return None
             except error.LookupError:
                 pass
         return renamed
 
     def parents(self):
         p = self._path
         fl = self._filelog
         pl = [(p, n, fl) for n in self._filelog.parents(self._filenode)]
 
         r = self._filelog.renamed(self._filenode)
         if r:
             pl[0] = (r[0], r[1], None)
 
         return [filectx(self._repo, p, fileid=n, filelog=l)
                 for p, n, l in pl if n != nullid]
 
     def p1(self):
         return self.parents()[0]
 
     def p2(self):
         p = self.parents()
         if len(p) == 2:
             return p[1]
         return filectx(self._repo, self._path, fileid=-1, filelog=self._filelog)
 
     def children(self):
         # hard for renames
         c = self._filelog.children(self._filenode)
         return [filectx(self._repo, self._path, fileid=x,
                         filelog=self._filelog) for x in c]
 
     def annotate(self, follow=False, linenumber=None, diffopts=None):
         '''returns a list of tuples of (ctx, line) for each line
         in the file, where ctx is the filectx of the node where
         that line was last changed.
         This returns tuples of ((ctx, linenumber), line) for each line,
         if "linenumber" parameter is NOT "None".
         In such tuples, linenumber means one at the first appearance
         in the managed file.
         To reduce annotation cost,
         this returns fixed value(False is used) as linenumber,
         if "linenumber" parameter is "False".'''
 
         def decorate_compat(text, rev):
             return ([rev] * len(text.splitlines()), text)
 
         def without_linenumber(text, rev):
             return ([(rev, False)] * len(text.splitlines()), text)
 
         def with_linenumber(text, rev):
             size = len(text.splitlines())
             return ([(rev, i) for i in xrange(1, size + 1)], text)
 
         decorate = (((linenumber is None) and decorate_compat) or
                     (linenumber and with_linenumber) or
                     without_linenumber)
 
         def pair(parent, child):
             blocks = mdiff.allblocks(parent[1], child[1], opts=diffopts,
                                      refine=True)
             for (a1, a2, b1, b2), t in blocks:
                 # Changed blocks ('!') or blocks made only of blank lines ('~')
                 # belong to the child.
                 if t == '=':
                     child[0][b1:b2] = parent[0][a1:a2]
             return child
 
         getlog = util.lrucachefunc(lambda x: self._repo.file(x))
         def getctx(path, fileid):
             log = path == self._path and self._filelog or getlog(path)
             return filectx(self._repo, path, fileid=fileid, filelog=log)
         getctx = util.lrucachefunc(getctx)
 
         def parents(f):
             # we want to reuse filectx objects as much as possible
             p = f._path
             if f._filerev is None: # working dir
                 pl = [(n.path(), n.filerev()) for n in f.parents()]
             else:
                 pl = [(p, n) for n in f._filelog.parentrevs(f._filerev)]
 
             if follow:
                 r = f.renamed()
                 if r:
                     pl[0] = (r[0], getlog(r[0]).rev(r[1]))
 
             return [getctx(p, n) for p, n in pl if n != nullrev]
 
         # use linkrev to find the first changeset where self appeared
         if self.rev() != self.linkrev():
             base = self.filectx(self.filerev())
         else:
             base = self
 
         # This algorithm would prefer to be recursive, but Python is a
         # bit recursion-hostile. Instead we do an iterative
         # depth-first search.
 
         visit = [base]
         hist = {}
         pcache = {}
         needed = {base: 1}
         while visit:
             f = visit[-1]
             if f not in pcache:
                 pcache[f] = parents(f)
 
             ready = True
             pl = pcache[f]
             for p in pl:
                 if p not in hist:
                     ready = False
                     visit.append(p)
                     needed[p] = needed.get(p, 0) + 1
             if ready:
                 visit.pop()
                 curr = decorate(f.data(), f)
                 for p in pl:
                     curr = pair(hist[p], curr)
                     if needed[p] == 1:
                         del hist[p]
                     else:
                         needed[p] -= 1
 
                 hist[f] = curr
                 pcache[f] = []
 
         return zip(hist[base][0], hist[base][1].splitlines(True))
 
     def ancestor(self, fc2, actx):
         """
         find the common ancestor file context, if any, of self, and fc2
 
         actx must be the changectx of the common ancestor
         of self's and fc2's respective changesets.
         """
 
         # the easy case: no (relevant) renames
         if fc2.path() == self.path() and self.path() in actx:
             return actx[self.path()]
 
         # the next easiest cases: unambiguous predecessor (name trumps
         # history)
         if self.path() in actx and fc2.path() not in actx:
             return actx[self.path()]
         if fc2.path() in actx and self.path() not in actx:
             return actx[fc2.path()]
 
         # prime the ancestor cache for the working directory
         acache = {}
         for c in (self, fc2):
             if c._filerev is None:
                 pl = [(n.path(), n.filenode()) for n in c.parents()]
                 acache[(c._path, None)] = pl
 
         flcache = {self._repopath:self._filelog, fc2._repopath:fc2._filelog}
         def parents(vertex):
             if vertex in acache:
                 return acache[vertex]
             f, n = vertex
             if f not in flcache:
                 flcache[f] = self._repo.file(f)
             fl = flcache[f]
             pl = [(f, p) for p in fl.parents(n) if p != nullid]
             re = fl.renamed(n)
             if re:
                 pl.append(re)
             acache[vertex] = pl
             return pl
 
         a, b = (self._path, self._filenode), (fc2._path, fc2._filenode)
         v = ancestor.ancestor(a, b, parents)
         if v:
             f, n = v
             return filectx(self._repo, f, fileid=n, filelog=flcache[f])
 
         return None
 
     def ancestors(self, followfirst=False):
         visit = {}
         c = self
         cut = followfirst and 1 or None
         while True:
             for parent in c.parents()[:cut]:
                 visit[(parent.rev(), parent.node())] = parent
             if not visit:
                 break
             c = visit.pop(max(visit))
             yield c
 
     def copies(self, c2):
         if not util.safehasattr(self, "_copycache"):
             self._copycache = {}
         sc2 = str(c2)
         if sc2 not in self._copycache:
             self._copycache[sc2] = copies.pathcopies(c2)
         return self._copycache[sc2]
 
 class workingctx(changectx):
     """A workingctx object makes access to data related to
     the current working directory convenient.
     date - any valid date string or (unixtime, offset), or None.
     user - username string, or None.
     extra - a dictionary of extra values, or None.
     changes - a list of file lists as returned by localrepo.status()
                or None to use the repository status.
     """
     def __init__(self, repo, text="", user=None, date=None, extra=None,
                  changes=None):
         self._repo = repo
         self._rev = None
         self._node = None
         self._text = text
         if date:
             self._date = util.parsedate(date)
         if user:
             self._user = user
         if changes:
             self._status = list(changes[:4])
             self._unknown = changes[4]
             self._ignored = changes[5]
             self._clean = changes[6]
         else:
             self._unknown = None
             self._ignored = None
             self._clean = None
 
         self._extra = {}
         if extra:
             self._extra = extra.copy()
         if 'branch' not in self._extra:
             try:
                 branch = encoding.fromlocal(self._repo.dirstate.branch())
             except UnicodeDecodeError:
                 raise util.Abort(_('branch name not in UTF-8!'))
             self._extra['branch'] = branch
         if self._extra['branch'] == '':
             self._extra['branch'] = 'default'
 
     def __str__(self):
         return str(self._parents[0]) + "+"
 
     def __repr__(self):
         return "<workingctx %s>" % str(self)
 
     def __nonzero__(self):
         return True
 
     def __contains__(self, key):
         return self._repo.dirstate[key] not in "?r"
 
     def _buildflagfunc(self):
         # Create a fallback function for getting file flags when the
         # filesystem doesn't support them
 
         copiesget = self._repo.dirstate.copies().get
 
         if len(self._parents) < 2:
             # when we have one parent, it's easy: copy from parent
             man = self._parents[0].manifest()
             def func(f):
                 f = copiesget(f, f)
                 return man.flags(f)
         else:
             # merges are tricky: we try to reconstruct the unstored
             # result from the merge (issue1802)
             p1, p2 = self._parents
             pa = p1.ancestor(p2)
             m1, m2, ma = p1.manifest(), p2.manifest(), pa.manifest()
 
             def func(f):
                 f = copiesget(f, f) # may be wrong for merges with copies
                 fl1, fl2, fla = m1.flags(f), m2.flags(f), ma.flags(f)
                 if fl1 == fl2:
                     return fl1
                 if fl1 == fla:
                     return fl2
                 if fl2 == fla:
                     return fl1
                 return '' # punt for conflicts
 
         return func
 
     @propertycache
     def _flagfunc(self):
         return self._repo.dirstate.flagfunc(self._buildflagfunc)
 
     @propertycache
     def _manifest(self):
         """generate a manifest corresponding to the working directory"""
 
         man = self._parents[0].manifest().copy()
         if len(self._parents) > 1:
             man2 = self.p2().manifest()
             def getman(f):
                 if f in man:
                     return man
                 return man2
         else:
             getman = lambda f: man
 
         copied = self._repo.dirstate.copies()
         ff = self._flagfunc
         modified, added, removed, deleted = self._status
         for i, l in (("a", added), ("m", modified)):
             for f in l:
                 orig = copied.get(f, f)
                 man[f] = getman(orig).get(orig, nullid) + i
                 try:
                     man.set(f, ff(f))
                 except OSError:
                     pass
 
         for f in deleted + removed:
             if f in man:
                 del man[f]
 
         return man
 
     def __iter__(self):
         d = self._repo.dirstate
         for f in d:
             if d[f] != 'r':
                 yield f
 
     @propertycache
     def _status(self):
         return self._repo.status()[:4]
 
     @propertycache
     def _user(self):
         return self._repo.ui.username()
 
     @propertycache
     def _date(self):
         return util.makedate()
 
     @propertycache
     def _parents(self):
         p = self._repo.dirstate.parents()
         if p[1] == nullid:
             p = p[:-1]
         return [changectx(self._repo, x) for x in p]
 
     def status(self, ignored=False, clean=False, unknown=False):
         """Explicit status query
         Unless this method is used to query the working copy status, the
         _status property will implicitly read the status using its default
         arguments."""
         stat = self._repo.status(ignored=ignored, clean=clean, unknown=unknown)
         self._unknown = self._ignored = self._clean = None
         if unknown:
             self._unknown = stat[4]
         if ignored:
             self._ignored = stat[5]
         if clean:
             self._clean = stat[6]
         self._status = stat[:4]
         return stat
 
     def manifest(self):
         return self._manifest
     def user(self):
         return self._user or self._repo.ui.username()
     def date(self):
         return self._date
     def description(self):
         return self._text
     def files(self):
         return sorted(self._status[0] + self._status[1] + self._status[2])
 
     def modified(self):
         return self._status[0]
     def added(self):
         return self._status[1]
     def removed(self):
         return self._status[2]
     def deleted(self):
         return self._status[3]
     def unknown(self):
         assert self._unknown is not None  # must call status first
         return self._unknown
     def ignored(self):
         assert self._ignored is not None  # must call status first
         return self._ignored
     def clean(self):
         assert self._clean is not None  # must call status first
         return self._clean
     def branch(self):
         return encoding.tolocal(self._extra['branch'])
     def closesbranch(self):
         return 'close' in self._extra
     def extra(self):
         return self._extra
 
     def tags(self):
         t = []
         for p in self.parents():
             t.extend(p.tags())
         return t
 
     def bookmarks(self):
         b = []
         for p in self.parents():
             b.extend(p.bookmarks())
         return b
 
     def phase(self):
         phase = phases.draft # default phase to draft
         for p in self.parents():
             phase = max(phase, p.phase())
         return phase
 
     def hidden(self):
         return False
 
     def children(self):
         return []
 
     def flags(self, path):
         if '_manifest' in self.__dict__:
             try:
                 return self._manifest.flags(path)
             except KeyError:
                 return ''
 
         try:
             return self._flagfunc(path)
         except OSError:
             return ''
 
     def filectx(self, path, filelog=None):
         """get a file context from the working directory"""
         return workingfilectx(self._repo, path, workingctx=self,
                               filelog=filelog)
 
     def ancestor(self, c2):
         """return the ancestor context of self and c2"""
         return self._parents[0].ancestor(c2) # punt on two parents for now
 
     def walk(self, match):
         return sorted(self._repo.dirstate.walk(match, self.substate.keys(),
                                                True, False))
 
     def dirty(self, missing=False, merge=True, branch=True):
         "check whether a working directory is modified"
         # check subrepos first
         for s in self.substate:
             if self.sub(s).dirty():
                 return True
         # check current working dir
         return ((merge and self.p2()) or
                 (branch and self.branch() != self.p1().branch()) or
                 self.modified() or self.added() or self.removed() or
                 (missing and self.deleted()))
 
     def add(self, list, prefix=""):
         join = lambda f: os.path.join(prefix, f)
         wlock = self._repo.wlock()
         ui, ds = self._repo.ui, self._repo.dirstate
         try:
             rejected = []
             for f in list:
                 scmutil.checkportable(ui, join(f))
                 p = self._repo.wjoin(f)
                 try:
                     st = os.lstat(p)
                 except OSError:
                     ui.warn(_("%s does not exist!\n") % join(f))
                     rejected.append(f)
                     continue
                 if st.st_size > 10000000:
                     ui.warn(_("%s: up to %d MB of RAM may be required "
                               "to manage this file\n"
                               "(use 'hg revert %s' to cancel the "
                               "pending addition)\n")
                               % (f, 3 * st.st_size // 1000000, join(f)))
                 if not (stat.S_ISREG(st.st_mode) or stat.S_ISLNK(st.st_mode)):
                     ui.warn(_("%s not added: only files and symlinks "
                               "supported currently\n") % join(f))
                     rejected.append(p)
                 elif ds[f] in 'amn':
                     ui.warn(_("%s already tracked!\n") % join(f))
                 elif ds[f] == 'r':
                     ds.normallookup(f)
                 else:
                     ds.add(f)
             return rejected
         finally:
             wlock.release()
 
     def forget(self, files, prefix=""):
         join = lambda f: os.path.join(prefix, f)
         wlock = self._repo.wlock()
         try:
             rejected = []
             for f in files:
                 if f not in self._repo.dirstate:
                     self._repo.ui.warn(_("%s not tracked!\n") % join(f))
                     rejected.append(f)
                 elif self._repo.dirstate[f] != 'a':
                     self._repo.dirstate.remove(f)
                 else:
                     self._repo.dirstate.drop(f)
             return rejected
         finally:
             wlock.release()
 
     def ancestors(self):
         for a in self._repo.changelog.ancestors(
             [p.rev() for p in self._parents]):
             yield changectx(self._repo, a)
 
     def undelete(self, list):
         pctxs = self.parents()
         wlock = self._repo.wlock()
         try:
             for f in list:
                 if self._repo.dirstate[f] != 'r':
                     self._repo.ui.warn(_("%s not removed!\n") % f)
                 else:
                     fctx = f in pctxs[0] and pctxs[0][f] or pctxs[1][f]
                     t = fctx.data()
                     self._repo.wwrite(f, t, fctx.flags())
                     self._repo.dirstate.normal(f)
         finally:
             wlock.release()
 
     def copy(self, source, dest):
         p = self._repo.wjoin(dest)
         if not os.path.lexists(p):
             self._repo.ui.warn(_("%s does not exist!\n") % dest)
         elif not (os.path.isfile(p) or os.path.islink(p)):
             self._repo.ui.warn(_("copy failed: %s is not a file or a "
                                  "symbolic link\n") % dest)
         else:
             wlock = self._repo.wlock()
             try:
                 if self._repo.dirstate[dest] in '?r':
                     self._repo.dirstate.add(dest)
                 self._repo.dirstate.copy(source, dest)
             finally:
                 wlock.release()
 
     def dirs(self):
         return set(self._repo.dirstate.dirs())
 
 class workingfilectx(filectx):
     """A workingfilectx object makes access to data related to a particular
        file in the working directory convenient."""
     def __init__(self, repo, path, filelog=None, workingctx=None):
         """changeid can be a changeset revision, node, or tag.
            fileid can be a file revision or node."""
         self._repo = repo
         self._path = path
         self._changeid = None
         self._filerev = self._filenode = None
 
         if filelog:
             self._filelog = filelog
         if workingctx:
             self._changectx = workingctx
 
     @propertycache
     def _changectx(self):
         return workingctx(self._repo)
 
     def __nonzero__(self):
         return True
 
     def __str__(self):
         return "%s@%s" % (self.path(), self._changectx)
 
     def __repr__(self):
         return "<workingfilectx %s>" % str(self)
 
     def data(self):
         return self._repo.wread(self._path)
     def renamed(self):
         rp = self._repo.dirstate.copied(self._path)
         if not rp:
             return None
         return rp, self._changectx._parents[0]._manifest.get(rp, nullid)
 
     def parents(self):
         '''return parent filectxs, following copies if necessary'''
         def filenode(ctx, path):
             return ctx._manifest.get(path, nullid)
 
         path = self._path
         fl = self._filelog
         pcl = self._changectx._parents
         renamed = self.renamed()
 
         if renamed:
             pl = [renamed + (None,)]
         else:
             pl = [(path, filenode(pcl[0], path), fl)]
 
         for pc in pcl[1:]:
             pl.append((path, filenode(pc, path), fl))
 
         return [filectx(self._repo, p, fileid=n, filelog=l)
                 for p, n, l in pl if n != nullid]
 
     def children(self):
         return []
 
     def size(self):
         return os.lstat(self._repo.wjoin(self._path)).st_size
     def date(self):
         t, tz = self._changectx.date()
         try:
             return (int(os.lstat(self._repo.wjoin(self._path)).st_mtime), tz)
         except OSError, err:
             if err.errno != errno.ENOENT:
                 raise
             return (t, tz)
 
     def cmp(self, fctx):
         """compare with other file context
 
         returns True if different than fctx.
         """
         # fctx should be a filectx (not a workingfilectx)
         # invert comparison to reuse the same code path
         return fctx.cmp(self)
 
 class memctx(object):
     """Use memctx to perform in-memory commits via localrepo.commitctx().
 
     Revision information is supplied at initialization time while
     related files data and is made available through a callback
     mechanism.  'repo' is the current localrepo, 'parents' is a
     sequence of two parent revisions identifiers (pass None for every
     missing parent), 'text' is the commit message and 'files' lists
     names of files touched by the revision (normalized and relative to
     repository root).
 
     filectxfn(repo, memctx, path) is a callable receiving the
     repository, the current memctx object and the normalized path of
     requested file, relative to repository root. It is fired by the
     commit function for every file in 'files', but calls order is
     undefined. If the file is available in the revision being
     committed (updated or added), filectxfn returns a memfilectx
     object. If the file was removed, filectxfn raises an
     IOError. Moved files are represented by marking the source file
     removed and the new file added with copy information (see
     memfilectx).
 
     user receives the committer name and defaults to current
     repository username, date is the commit date in any format
     supported by util.parsedate() and defaults to current date, extra
     is a dictionary of metadata or is left empty.
     """
     def __init__(self, repo, parents, text, files, filectxfn, user=None,
                  date=None, extra=None):
         self._repo = repo
         self._rev = None
         self._node = None
         self._text = text
         self._date = date and util.parsedate(date) or util.makedate()
         self._user = user
         parents = [(p or nullid) for p in parents]
         p1, p2 = parents
         self._parents = [changectx(self._repo, p) for p in (p1, p2)]
         files = sorted(set(files))
         self._status = [files, [], [], [], []]
         self._filectxfn = filectxfn
 
         self._extra = extra and extra.copy() or {}
         if self._extra.get('branch', '') == '':
             self._extra['branch'] = 'default'
 
     def __str__(self):
         return str(self._parents[0]) + "+"
 
     def __int__(self):
         return self._rev
 
     def __nonzero__(self):
         return True
 
     def __getitem__(self, key):
         return self.filectx(key)
 
     def p1(self):
         return self._parents[0]
     def p2(self):
         return self._parents[1]
 
     def user(self):
         return self._user or self._repo.ui.username()
     def date(self):
         return self._date
     def description(self):
         return self._text
     def files(self):
         return self.modified()
     def modified(self):
         return self._status[0]
     def added(self):
         return self._status[1]
     def removed(self):
         return self._status[2]
     def deleted(self):
         return self._status[3]
     def unknown(self):
         return self._status[4]
     def ignored(self):
         return self._status[5]
     def clean(self):
         return self._status[6]
     def branch(self):
         return encoding.tolocal(self._extra['branch'])
     def extra(self):
         return self._extra
     def flags(self, f):
         return self[f].flags()
 
     def parents(self):
         """return contexts for each parent changeset"""
         return self._parents
 
     def filectx(self, path, filelog=None):
         """get a file context from the working directory"""
         return self._filectxfn(self._repo, self, path)
 
     def commit(self):
         """commit context to the repo"""
         return self._repo.commitctx(self)
 
 class memfilectx(object):
     """memfilectx represents an in-memory file to commit.
 
     See memctx for more details.
     """
     def __init__(self, path, data, islink=False, isexec=False, copied=None):
         """
         path is the normalized file path relative to repository root.
         data is the file content as a string.
         islink is True if the file is a symbolic link.
         isexec is True if the file is executable.
         copied is the source file path if current file was copied in the
         revision being committed, or None."""
         self._path = path
         self._data = data
         self._flags = (islink and 'l' or '') + (isexec and 'x' or '')
         self._copied = None
         if copied:
             self._copied = (copied, nullid)
 
     def __nonzero__(self):
         return True
     def __str__(self):
         return "%s@%s" % (self.path(), self._changectx)
     def path(self):
         return self._path
     def data(self):
         return self._data
     def flags(self):
         return self._flags
     def isexec(self):
         return 'x' in self._flags
     def islink(self):
         return 'l' in self._flags
     def renamed(self):
         return self._copied

mercurial/localrepo.py

@@ -1,2608 +1,2610 b''
 # localrepo.py - read/write repository class for mercurial
 #
 # Copyright 2005-2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 from node import bin, hex, nullid, nullrev, short
 from i18n import _
 import peer, changegroup, subrepo, discovery, pushkey, obsolete
 import changelog, dirstate, filelog, manifest, context, bookmarks, phases
 import lock, transaction, store, encoding, base85
 import scmutil, util, extensions, hook, error, revset
 import match as matchmod
 import merge as mergemod
 import tags as tagsmod
 from lock import release
 import weakref, errno, os, time, inspect
 propertycache = util.propertycache
 filecache = scmutil.filecache
 
 class storecache(filecache):
     """filecache for files in the store"""
     def join(self, obj, fname):
         return obj.sjoin(fname)
 
 MODERNCAPS = set(('lookup', 'branchmap', 'pushkey', 'known', 'getbundle'))
 LEGACYCAPS = MODERNCAPS.union(set(['changegroupsubset']))
 
 class localpeer(peer.peerrepository):
     '''peer for a local repo; reflects only the most recent API'''
 
     def __init__(self, repo, caps=MODERNCAPS):
         peer.peerrepository.__init__(self)
         self._repo = repo
         self.ui = repo.ui
         self._caps = repo._restrictcapabilities(caps)
         self.requirements = repo.requirements
         self.supportedformats = repo.supportedformats
 
     def close(self):
         self._repo.close()
 
     def _capabilities(self):
         return self._caps
 
     def local(self):
         return self._repo
 
     def canpush(self):
         return True
 
     def url(self):
         return self._repo.url()
 
     def lookup(self, key):
         return self._repo.lookup(key)
 
     def branchmap(self):
         return discovery.visiblebranchmap(self._repo)
 
     def heads(self):
         return discovery.visibleheads(self._repo)
 
     def known(self, nodes):
         return self._repo.known(nodes)
 
     def getbundle(self, source, heads=None, common=None):
         return self._repo.getbundle(source, heads=heads, common=common)
 
     # TODO We might want to move the next two calls into legacypeer and add
     # unbundle instead.
 
     def lock(self):
         return self._repo.lock()
 
     def addchangegroup(self, cg, source, url):
         return self._repo.addchangegroup(cg, source, url)
 
     def pushkey(self, namespace, key, old, new):
         return self._repo.pushkey(namespace, key, old, new)
 
     def listkeys(self, namespace):
         return self._repo.listkeys(namespace)
 
     def debugwireargs(self, one, two, three=None, four=None, five=None):
         '''used to test argument passing over the wire'''
         return "%s %s %s %s %s" % (one, two, three, four, five)
 
 class locallegacypeer(localpeer):
     '''peer extension which implements legacy methods too; used for tests with
     restricted capabilities'''
 
     def __init__(self, repo):
         localpeer.__init__(self, repo, caps=LEGACYCAPS)
 
     def branches(self, nodes):
         return self._repo.branches(nodes)
 
     def between(self, pairs):
         return self._repo.between(pairs)
 
     def changegroup(self, basenodes, source):
         return self._repo.changegroup(basenodes, source)
 
     def changegroupsubset(self, bases, heads, source):
         return self._repo.changegroupsubset(bases, heads, source)
 
 class localrepository(object):
 
     supportedformats = set(('revlogv1', 'generaldelta'))
     supported = supportedformats | set(('store', 'fncache', 'shared',
                                         'dotencode'))
     openerreqs = set(('revlogv1', 'generaldelta'))
     requirements = ['revlogv1']
 
     def _baserequirements(self, create):
         return self.requirements[:]
 
     def __init__(self, baseui, path=None, create=False):
         self.wopener = scmutil.opener(path, expand=True)
         self.wvfs = self.wopener
         self.root = self.wvfs.base
         self.path = self.wvfs.join(".hg")
         self.origroot = path
         self.auditor = scmutil.pathauditor(self.root, self._checknested)
         self.opener = scmutil.opener(self.path)
         self.vfs = self.opener
         self.baseui = baseui
         self.ui = baseui.copy()
         # A list of callback to shape the phase if no data were found.
         # Callback are in the form: func(repo, roots) --> processed root.
         # This list it to be filled by extension during repo setup
         self._phasedefaults = []
         try:
             self.ui.readconfig(self.join("hgrc"), self.root)
             extensions.loadall(self.ui)
         except IOError:
             pass
 
         if not self.vfs.isdir():
             if create:
                 if not self.wvfs.exists():
                     self.wvfs.makedirs()
                 self.vfs.makedir(notindexed=True)
                 requirements = self._baserequirements(create)
                 if self.ui.configbool('format', 'usestore', True):
                     self.vfs.mkdir("store")
                     requirements.append("store")
                     if self.ui.configbool('format', 'usefncache', True):
                         requirements.append("fncache")
                         if self.ui.configbool('format', 'dotencode', True):
                             requirements.append('dotencode')
                     # create an invalid changelog
                     self.vfs.append(
                         "00changelog.i",
                         '\0\0\0\2' # represents revlogv2
                         ' dummy changelog to prevent using the old repo layout'
                     )
                 if self.ui.configbool('format', 'generaldelta', False):
                     requirements.append("generaldelta")
                 requirements = set(requirements)
             else:
                 raise error.RepoError(_("repository %s not found") % path)
         elif create:
             raise error.RepoError(_("repository %s already exists") % path)
         else:
             try:
                 requirements = scmutil.readrequires(self.vfs, self.supported)
             except IOError, inst:
                 if inst.errno != errno.ENOENT:
                     raise
                 requirements = set()
 
         self.sharedpath = self.path
         try:
             s = os.path.realpath(self.opener.read("sharedpath").rstrip('\n'))
             if not os.path.exists(s):
                 raise error.RepoError(
                     _('.hg/sharedpath points to nonexistent directory %s') % s)
             self.sharedpath = s
         except IOError, inst:
             if inst.errno != errno.ENOENT:
                 raise
 
         self.store = store.store(requirements, self.sharedpath, scmutil.opener)
         self.spath = self.store.path
         self.sopener = self.store.opener
         self.svfs = self.sopener
         self.sjoin = self.store.join
         self.opener.createmode = self.store.createmode
         self._applyrequirements(requirements)
         if create:
             self._writerequirements()
 
 
         self._branchcache = None
         self._branchcachetip = None
         self.filterpats = {}
         self._datafilters = {}
         self._transref = self._lockref = self._wlockref = None
 
         # A cache for various files under .hg/ that tracks file changes,
         # (used by the filecache decorator)
         #
         # Maps a property name to its util.filecacheentry
         self._filecache = {}
 
     def close(self):
         pass
 
     def _restrictcapabilities(self, caps):
         return caps
 
     def _applyrequirements(self, requirements):
         self.requirements = requirements
         self.sopener.options = dict((r, 1) for r in requirements
                                            if r in self.openerreqs)
 
     def _writerequirements(self):
         reqfile = self.opener("requires", "w")
         for r in self.requirements:
             reqfile.write("%s\n" % r)
         reqfile.close()
 
     def _checknested(self, path):
         """Determine if path is a legal nested repository."""
         if not path.startswith(self.root):
             return False
         subpath = path[len(self.root) + 1:]
         normsubpath = util.pconvert(subpath)
 
         # XXX: Checking against the current working copy is wrong in
         # the sense that it can reject things like
         #
         #   $ hg cat -r 10 sub/x.txt
         #
         # if sub/ is no longer a subrepository in the working copy
         # parent revision.
         #
         # However, it can of course also allow things that would have
         # been rejected before, such as the above cat command if sub/
         # is a subrepository now, but was a normal directory before.
         # The old path auditor would have rejected by mistake since it
         # panics when it sees sub/.hg/.
         #
         # All in all, checking against the working copy seems sensible
         # since we want to prevent access to nested repositories on
         # the filesystem *now*.
         ctx = self[None]
         parts = util.splitpath(subpath)
         while parts:
             prefix = '/'.join(parts)
             if prefix in ctx.substate:
                 if prefix == normsubpath:
                     return True
                 else:
                     sub = ctx.sub(prefix)
                     return sub.checknested(subpath[len(prefix) + 1:])
             else:
                 parts.pop()
         return False
 
     def peer(self):
         return localpeer(self) # not cached to avoid reference cycle
 
     @filecache('bookmarks')
     def _bookmarks(self):
         return bookmarks.read(self)
 
     @filecache('bookmarks.current')
     def _bookmarkcurrent(self):
         return bookmarks.readcurrent(self)
 
     def _writebookmarks(self, marks):
         bookmarks.write(self)
 
     def bookmarkheads(self, bookmark):
         name = bookmark.split('@', 1)[0]
         heads = []
         for mark, n in self._bookmarks.iteritems():
             if mark.split('@', 1)[0] == name:
                 heads.append(n)
         return heads
 
     @storecache('phaseroots')
     def _phasecache(self):
         return phases.phasecache(self, self._phasedefaults)
 
     @storecache('obsstore')
     def obsstore(self):
         store = obsolete.obsstore(self.sopener)
         if store and not obsolete._enabled:
             # message is rare enough to not be translated
             msg = 'obsolete feature not enabled but %i markers found!\n'
             self.ui.warn(msg % len(list(store)))
         return store
 
     @propertycache
     def hiddenrevs(self):
         """hiddenrevs: revs that should be hidden by command and tools
 
         This set is carried on the repo to ease initialisation and lazy
         loading; it'll probably move back to changelog for efficiency and
         consistency reason
 
         Note that the hiddenrevs will needs invalidations when
         - a new changesets is added (possible unstable above extinct)
         - a new obsolete marker is added (possible new extinct changeset)
 
         hidden changesets cannot have non-hidden descendants
         """
         hidden = set()
         if self.obsstore:
             ### hide extinct changeset that are not accessible by any mean
             hiddenquery = 'extinct() - ::(. + bookmark() + tagged())'
             hidden.update(self.revs(hiddenquery))
         return hidden
 
     @storecache('00changelog.i')
     def changelog(self):
         c = changelog.changelog(self.sopener)
         if 'HG_PENDING' in os.environ:
             p = os.environ['HG_PENDING']
             if p.startswith(self.root):
                 c.readpending('00changelog.i.a')
         return c
 
     @storecache('00manifest.i')
     def manifest(self):
         return manifest.manifest(self.sopener)
 
     @filecache('dirstate')
     def dirstate(self):
         warned = [0]
         def validate(node):
             try:
                 self.changelog.rev(node)
                 return node
             except error.LookupError:
                 if not warned[0]:
                     warned[0] = True
                     self.ui.warn(_("warning: ignoring unknown"
                                    " working parent %s!\n") % short(node))
                 return nullid
 
         return dirstate.dirstate(self.opener, self.ui, self.root, validate)
 
     def __getitem__(self, changeid):
         if changeid is None:
             return context.workingctx(self)
         return context.changectx(self, changeid)
 
     def __contains__(self, changeid):
         try:
             return bool(self.lookup(changeid))
         except error.RepoLookupError:
             return False
 
     def __nonzero__(self):
         return True
 
     def __len__(self):
         return len(self.changelog)
 
     def __iter__(self):
         for i in xrange(len(self)):
             yield i
 
     def revs(self, expr, *args):
         '''Return a list of revisions matching the given revset'''
         expr = revset.formatspec(expr, *args)
         m = revset.match(None, expr)
         return [r for r in m(self, range(len(self)))]
 
     def set(self, expr, *args):
         '''
         Yield a context for each matching revision, after doing arg
         replacement via revset.formatspec
         '''
         for r in self.revs(expr, *args):
             yield self[r]
 
     def url(self):
         return 'file:' + self.root
 
     def hook(self, name, throw=False, **args):
         return hook.hook(self.ui, self, name, throw, **args)
 
     tag_disallowed = ':\r\n'
 
     def _tag(self, names, node, message, local, user, date, extra={}):
         if isinstance(names, str):
             allchars = names
             names = (names,)
         else:
             allchars = ''.join(names)
         for c in self.tag_disallowed:
             if c in allchars:
                 raise util.Abort(_('%r cannot be used in a tag name') % c)
 
         branches = self.branchmap()
         for name in names:
             self.hook('pretag', throw=True, node=hex(node), tag=name,
                       local=local)
             if name in branches:
                 self.ui.warn(_("warning: tag %s conflicts with existing"
                 " branch name\n") % name)
 
         def writetags(fp, names, munge, prevtags):
             fp.seek(0, 2)
             if prevtags and prevtags[-1] != '\n':
                 fp.write('\n')
             for name in names:
                 m = munge and munge(name) or name
                 if (self._tagscache.tagtypes and
                     name in self._tagscache.tagtypes):
                     old = self.tags().get(name, nullid)
                     fp.write('%s %s\n' % (hex(old), m))
                 fp.write('%s %s\n' % (hex(node), m))
             fp.close()
 
         prevtags = ''
         if local:
             try:
                 fp = self.opener('localtags', 'r+')
             except IOError:
                 fp = self.opener('localtags', 'a')
             else:
                 prevtags = fp.read()
 
             # local tags are stored in the current charset
             writetags(fp, names, None, prevtags)
             for name in names:
                 self.hook('tag', node=hex(node), tag=name, local=local)
             return
 
         try:
             fp = self.wfile('.hgtags', 'rb+')
         except IOError, e:
             if e.errno != errno.ENOENT:
                 raise
             fp = self.wfile('.hgtags', 'ab')
         else:
             prevtags = fp.read()
 
         # committed tags are stored in UTF-8
         writetags(fp, names, encoding.fromlocal, prevtags)
 
         fp.close()
 
         self.invalidatecaches()
 
         if '.hgtags' not in self.dirstate:
             self[None].add(['.hgtags'])
 
         m = matchmod.exact(self.root, '', ['.hgtags'])
         tagnode = self.commit(message, user, date, extra=extra, match=m)
 
         for name in names:
             self.hook('tag', node=hex(node), tag=name, local=local)
 
         return tagnode
 
     def tag(self, names, node, message, local, user, date):
         '''tag a revision with one or more symbolic names.
 
         names is a list of strings or, when adding a single tag, names may be a
         string.
 
         if local is True, the tags are stored in a per-repository file.
         otherwise, they are stored in the .hgtags file, and a new
         changeset is committed with the change.
 
         keyword arguments:
 
         local: whether to store tags in non-version-controlled file
         (default False)
 
         message: commit message to use if committing
 
         user: name of user to use if committing
 
         date: date tuple to use if committing'''
 
         if not local:
             for x in self.status()[:5]:
                 if '.hgtags' in x:
                     raise util.Abort(_('working copy of .hgtags is changed '
                                        '(please commit .hgtags manually)'))
 
         self.tags() # instantiate the cache
         self._tag(names, node, message, local, user, date)
 
     @propertycache
     def _tagscache(self):
         '''Returns a tagscache object that contains various tags related
         caches.'''
 
         # This simplifies its cache management by having one decorated
         # function (this one) and the rest simply fetch things from it.
         class tagscache(object):
             def __init__(self):
                 # These two define the set of tags for this repository. tags
                 # maps tag name to node; tagtypes maps tag name to 'global' or
                 # 'local'. (Global tags are defined by .hgtags across all
                 # heads, and local tags are defined in .hg/localtags.)
                 # They constitute the in-memory cache of tags.
                 self.tags = self.tagtypes = None
 
                 self.nodetagscache = self.tagslist = None
 
         cache = tagscache()
         cache.tags, cache.tagtypes = self._findtags()
 
         return cache
 
     def tags(self):
         '''return a mapping of tag to node'''
         t = {}
         for k, v in self._tagscache.tags.iteritems():
             try:
                 # ignore tags to unknown nodes
                 self.changelog.rev(v)
                 t[k] = v
             except (error.LookupError, ValueError):
                 pass
         return t
 
     def _findtags(self):
         '''Do the hard work of finding tags.  Return a pair of dicts
         (tags, tagtypes) where tags maps tag name to node, and tagtypes
         maps tag name to a string like \'global\' or \'local\'.
         Subclasses or extensions are free to add their own tags, but
         should be aware that the returned dicts will be retained for the
         duration of the localrepo object.'''
 
         # XXX what tagtype should subclasses/extensions use?  Currently
         # mq and bookmarks add tags, but do not set the tagtype at all.
         # Should each extension invent its own tag type?  Should there
         # be one tagtype for all such "virtual" tags?  Or is the status
         # quo fine?
 
         alltags = {}                    # map tag name to (node, hist)
         tagtypes = {}
 
         tagsmod.findglobaltags(self.ui, self, alltags, tagtypes)
         tagsmod.readlocaltags(self.ui, self, alltags, tagtypes)
 
         # Build the return dicts.  Have to re-encode tag names because
         # the tags module always uses UTF-8 (in order not to lose info
         # writing to the cache), but the rest of Mercurial wants them in
         # local encoding.
         tags = {}
         for (name, (node, hist)) in alltags.iteritems():
             if node != nullid:
                 tags[encoding.tolocal(name)] = node
         tags['tip'] = self.changelog.tip()
         tagtypes = dict([(encoding.tolocal(name), value)
                          for (name, value) in tagtypes.iteritems()])
         return (tags, tagtypes)
 
     def tagtype(self, tagname):
         '''
         return the type of the given tag. result can be:
 
         'local'  : a local tag
         'global' : a global tag
         None     : tag does not exist
         '''
 
         return self._tagscache.tagtypes.get(tagname)
 
     def tagslist(self):
         '''return a list of tags ordered by revision'''
         if not self._tagscache.tagslist:
             l = []
             for t, n in self.tags().iteritems():
                 r = self.changelog.rev(n)
                 l.append((r, t, n))
             self._tagscache.tagslist = [(t, n) for r, t, n in sorted(l)]
 
         return self._tagscache.tagslist
 
     def nodetags(self, node):
         '''return the tags associated with a node'''
         if not self._tagscache.nodetagscache:
             nodetagscache = {}
             for t, n in self._tagscache.tags.iteritems():
                 nodetagscache.setdefault(n, []).append(t)
             for tags in nodetagscache.itervalues():
                 tags.sort()
             self._tagscache.nodetagscache = nodetagscache
         return self._tagscache.nodetagscache.get(node, [])
 
     def nodebookmarks(self, node):
         marks = []
         for bookmark, n in self._bookmarks.iteritems():
             if n == node:
                 marks.append(bookmark)
         return sorted(marks)
 
     def _branchtags(self, partial, lrev):
         # TODO: rename this function?
         tiprev = len(self) - 1
         if lrev != tiprev:
             ctxgen = (self[r] for r in xrange(lrev + 1, tiprev + 1))
             self._updatebranchcache(partial, ctxgen)
             self._writebranchcache(partial, self.changelog.tip(), tiprev)
 
         return partial
 
     def updatebranchcache(self):
         tip = self.changelog.tip()
         if self._branchcache is not None and self._branchcachetip == tip:
             return
 
         oldtip = self._branchcachetip
         self._branchcachetip = tip
         if oldtip is None or oldtip not in self.changelog.nodemap:
             partial, last, lrev = self._readbranchcache()
         else:
             lrev = self.changelog.rev(oldtip)
             partial = self._branchcache
 
         self._branchtags(partial, lrev)
         # this private cache holds all heads (not just the branch tips)
         self._branchcache = partial
 
     def branchmap(self):
         '''returns a dictionary {branch: [branchheads]}'''
         self.updatebranchcache()
         return self._branchcache
 
     def _branchtip(self, heads):
         '''return the tipmost branch head in heads'''
         tip = heads[-1]
         for h in reversed(heads):
             if not self[h].closesbranch():
                 tip = h
                 break
         return tip
 
     def branchtip(self, branch):
         '''return the tip node for a given branch'''
         if branch not in self.branchmap():
             raise error.RepoLookupError(_("unknown branch '%s'") % branch)
         return self._branchtip(self.branchmap()[branch])
 
     def branchtags(self):
         '''return a dict where branch names map to the tipmost head of
         the branch, open heads come before closed'''
         bt = {}
         for bn, heads in self.branchmap().iteritems():
             bt[bn] = self._branchtip(heads)
         return bt
 
     def _readbranchcache(self):
         partial = {}
         try:
             f = self.opener("cache/branchheads")
             lines = f.read().split('\n')
             f.close()
         except (IOError, OSError):
             return {}, nullid, nullrev
 
         try:
             last, lrev = lines.pop(0).split(" ", 1)
             last, lrev = bin(last), int(lrev)
             if lrev >= len(self) or self[lrev].node() != last:
                 # invalidate the cache
                 raise ValueError('invalidating branch cache (tip differs)')
             for l in lines:
                 if not l:
                     continue
                 node, label = l.split(" ", 1)
                 label = encoding.tolocal(label.strip())
                 if not node in self:
                     raise ValueError('invalidating branch cache because node '+
                                      '%s does not exist' % node)
                 partial.setdefault(label, []).append(bin(node))
         except KeyboardInterrupt:
             raise
         except Exception, inst:
             if self.ui.debugflag:
                 self.ui.warn(str(inst), '\n')
             partial, last, lrev = {}, nullid, nullrev
         return partial, last, lrev
 
     def _writebranchcache(self, branches, tip, tiprev):
         try:
             f = self.opener("cache/branchheads", "w", atomictemp=True)
             f.write("%s %s\n" % (hex(tip), tiprev))
             for label, nodes in branches.iteritems():
                 for node in nodes:
                     f.write("%s %s\n" % (hex(node), encoding.fromlocal(label)))
             f.close()
         except (IOError, OSError):
             pass
 
     def _updatebranchcache(self, partial, ctxgen):
         """Given a branchhead cache, partial, that may have extra nodes or be
         missing heads, and a generator of nodes that are at least a superset of
         heads missing, this function updates partial to be correct.
         """
         # collect new branch entries
         newbranches = {}
         for c in ctxgen:
             newbranches.setdefault(c.branch(), []).append(c.node())
         # if older branchheads are reachable from new ones, they aren't
         # really branchheads. Note checking parents is insufficient:
         # 1 (branch a) -> 2 (branch b) -> 3 (branch a)
         for branch, newnodes in newbranches.iteritems():
             bheads = partial.setdefault(branch, [])
             # Remove candidate heads that no longer are in the repo (e.g., as
             # the result of a strip that just happened).  Avoid using 'node in
             # self' here because that dives down into branchcache code somewhat
             # recursively.
             bheadrevs = [self.changelog.rev(node) for node in bheads
                          if self.changelog.hasnode(node)]
             newheadrevs = [self.changelog.rev(node) for node in newnodes
                            if self.changelog.hasnode(node)]
             ctxisnew = bheadrevs and min(newheadrevs) > max(bheadrevs)
             # Remove duplicates - nodes that are in newheadrevs and are already
             # in bheadrevs.  This can happen if you strip a node whose parent
             # was already a head (because they're on different branches).
             bheadrevs = sorted(set(bheadrevs).union(newheadrevs))
 
             # Starting from tip means fewer passes over reachable.  If we know
             # the new candidates are not ancestors of existing heads, we don't
             # have to examine ancestors of existing heads
             if ctxisnew:
                 iterrevs = sorted(newheadrevs)
             else:
                 iterrevs = list(bheadrevs)
 
             # This loop prunes out two kinds of heads - heads that are
             # superseded by a head in newheadrevs, and newheadrevs that are not
             # heads because an existing head is their descendant.
             while iterrevs:
                 latest = iterrevs.pop()
                 if latest not in bheadrevs:
                     continue
                 ancestors = set(self.changelog.ancestors([latest],
                                                          bheadrevs[0]))
                 if ancestors:
                     bheadrevs = [b for b in bheadrevs if b not in ancestors]
             partial[branch] = [self.changelog.node(rev) for rev in bheadrevs]
 
         # There may be branches that cease to exist when the last commit in the
         # branch was stripped.  This code filters them out.  Note that the
         # branch that ceased to exist may not be in newbranches because
         # newbranches is the set of candidate heads, which when you strip the
         # last commit in a branch will be the parent branch.
         for branch in partial.keys():
             nodes = [head for head in partial[branch]
                      if self.changelog.hasnode(head)]
             if not nodes:
                 del partial[branch]
 
     def lookup(self, key):
         return self[key].node()
 
     def lookupbranch(self, key, remote=None):
         repo = remote or self
         if key in repo.branchmap():
             return key
 
         repo = (remote and remote.local()) and remote or self
         return repo[key].branch()
 
     def known(self, nodes):
         nm = self.changelog.nodemap
         pc = self._phasecache
         result = []
         for n in nodes:
             r = nm.get(n)
             resp = not (r is None or pc.phase(self, r) >= phases.secret)
             result.append(resp)
         return result
 
     def local(self):
         return self
 
     def cancopy(self):
         return self.local() # so statichttprepo's override of local() works
 
     def join(self, f):
         return os.path.join(self.path, f)
 
     def wjoin(self, f):
         return os.path.join(self.root, f)
 
     def file(self, f):
         if f[0] == '/':
             f = f[1:]
         return filelog.filelog(self.sopener, f)
 
     def changectx(self, changeid):
         return self[changeid]
 
     def parents(self, changeid=None):
         '''get list of changectxs for parents of changeid'''
         return self[changeid].parents()
 
     def setparents(self, p1, p2=nullid):
         copies = self.dirstate.setparents(p1, p2)
         if copies:
             # Adjust copy records, the dirstate cannot do it, it
             # requires access to parents manifests. Preserve them
             # only for entries added to first parent.
             pctx = self[p1]
             for f in copies:
                 if f not in pctx and copies[f] in pctx:
                     self.dirstate.copy(copies[f], f)
 
     def filectx(self, path, changeid=None, fileid=None):
         """changeid can be a changeset revision, node, or tag.
            fileid can be a file revision or node."""
         return context.filectx(self, path, changeid, fileid)
 
     def getcwd(self):
         return self.dirstate.getcwd()
 
     def pathto(self, f, cwd=None):
         return self.dirstate.pathto(f, cwd)
 
     def wfile(self, f, mode='r'):
         return self.wopener(f, mode)
 
     def _link(self, f):
         return os.path.islink(self.wjoin(f))
 
     def _loadfilter(self, filter):
         if filter not in self.filterpats:
             l = []
             for pat, cmd in self.ui.configitems(filter):
                 if cmd == '!':
                     continue
                 mf = matchmod.match(self.root, '', [pat])
                 fn = None
                 params = cmd
                 for name, filterfn in self._datafilters.iteritems():
                     if cmd.startswith(name):
                         fn = filterfn
                         params = cmd[len(name):].lstrip()
                         break
                 if not fn:
                     fn = lambda s, c, **kwargs: util.filter(s, c)
                 # Wrap old filters not supporting keyword arguments
                 if not inspect.getargspec(fn)[2]:
                     oldfn = fn
                     fn = lambda s, c, **kwargs: oldfn(s, c)
                 l.append((mf, fn, params))
             self.filterpats[filter] = l
         return self.filterpats[filter]
 
     def _filter(self, filterpats, filename, data):
         for mf, fn, cmd in filterpats:
             if mf(filename):
                 self.ui.debug("filtering %s through %s\n" % (filename, cmd))
                 data = fn(data, cmd, ui=self.ui, repo=self, filename=filename)
                 break
 
         return data
 
     @propertycache
     def _encodefilterpats(self):
         return self._loadfilter('encode')
 
     @propertycache
     def _decodefilterpats(self):
         return self._loadfilter('decode')
 
     def adddatafilter(self, name, filter):
         self._datafilters[name] = filter
 
     def wread(self, filename):
         if self._link(filename):
             data = os.readlink(self.wjoin(filename))
         else:
             data = self.wopener.read(filename)
         return self._filter(self._encodefilterpats, filename, data)
 
     def wwrite(self, filename, data, flags):
         data = self._filter(self._decodefilterpats, filename, data)
         if 'l' in flags:
             self.wopener.symlink(data, filename)
         else:
             self.wopener.write(filename, data)
             if 'x' in flags:
                 util.setflags(self.wjoin(filename), False, True)
 
     def wwritedata(self, filename, data):
         return self._filter(self._decodefilterpats, filename, data)
 
     def transaction(self, desc):
         tr = self._transref and self._transref() or None
         if tr and tr.running():
             return tr.nest()
 
         # abort here if the journal already exists
         if os.path.exists(self.sjoin("journal")):
             raise error.RepoError(
                 _("abandoned transaction found - run hg recover"))
 
         self._writejournal(desc)
         renames = [(x, undoname(x)) for x in self._journalfiles()]
 
         tr = transaction.transaction(self.ui.warn, self.sopener,
                                      self.sjoin("journal"),
                                      aftertrans(renames),
                                      self.store.createmode)
         self._transref = weakref.ref(tr)
         return tr
 
     def _journalfiles(self):
         return (self.sjoin('journal'), self.join('journal.dirstate'),
                 self.join('journal.branch'), self.join('journal.desc'),
                 self.join('journal.bookmarks'),
                 self.sjoin('journal.phaseroots'))
 
     def undofiles(self):
         return [undoname(x) for x in self._journalfiles()]
 
     def _writejournal(self, desc):
         self.opener.write("journal.dirstate",
                           self.opener.tryread("dirstate"))
         self.opener.write("journal.branch",
                           encoding.fromlocal(self.dirstate.branch()))
         self.opener.write("journal.desc",
                           "%d\n%s\n" % (len(self), desc))
         self.opener.write("journal.bookmarks",
                           self.opener.tryread("bookmarks"))
         self.sopener.write("journal.phaseroots",
                            self.sopener.tryread("phaseroots"))
 
     def recover(self):
         lock = self.lock()
         try:
             if os.path.exists(self.sjoin("journal")):
                 self.ui.status(_("rolling back interrupted transaction\n"))
                 transaction.rollback(self.sopener, self.sjoin("journal"),
                                      self.ui.warn)
                 self.invalidate()
                 return True
             else:
                 self.ui.warn(_("no interrupted transaction available\n"))
                 return False
         finally:
             lock.release()
 
     def rollback(self, dryrun=False, force=False):
         wlock = lock = None
         try:
             wlock = self.wlock()
             lock = self.lock()
             if os.path.exists(self.sjoin("undo")):
                 return self._rollback(dryrun, force)
             else:
                 self.ui.warn(_("no rollback information available\n"))
                 return 1
         finally:
             release(lock, wlock)
 
     def _rollback(self, dryrun, force):
         ui = self.ui
         try:
             args = self.opener.read('undo.desc').splitlines()
             (oldlen, desc, detail) = (int(args[0]), args[1], None)
             if len(args) >= 3:
                 detail = args[2]
             oldtip = oldlen - 1
 
             if detail and ui.verbose:
                 msg = (_('repository tip rolled back to revision %s'
                          ' (undo %s: %s)\n')
                        % (oldtip, desc, detail))
             else:
                 msg = (_('repository tip rolled back to revision %s'
                          ' (undo %s)\n')
                        % (oldtip, desc))
         except IOError:
             msg = _('rolling back unknown transaction\n')
             desc = None
 
         if not force and self['.'] != self['tip'] and desc == 'commit':
             raise util.Abort(
                 _('rollback of last commit while not checked out '
                   'may lose data'), hint=_('use -f to force'))
 
         ui.status(msg)
         if dryrun:
             return 0
 
         parents = self.dirstate.parents()
         transaction.rollback(self.sopener, self.sjoin('undo'), ui.warn)
         if os.path.exists(self.join('undo.bookmarks')):
             util.rename(self.join('undo.bookmarks'),
                         self.join('bookmarks'))
         if os.path.exists(self.sjoin('undo.phaseroots')):
             util.rename(self.sjoin('undo.phaseroots'),
                         self.sjoin('phaseroots'))
         self.invalidate()
 
         # Discard all cache entries to force reloading everything.
         self._filecache.clear()
 
         parentgone = (parents[0] not in self.changelog.nodemap or
                       parents[1] not in self.changelog.nodemap)
         if parentgone:
             util.rename(self.join('undo.dirstate'), self.join('dirstate'))
             try:
                 branch = self.opener.read('undo.branch')
                 self.dirstate.setbranch(encoding.tolocal(branch))
             except IOError:
                 ui.warn(_('named branch could not be reset: '
                           'current branch is still \'%s\'\n')
                         % self.dirstate.branch())
 
             self.dirstate.invalidate()
             parents = tuple([p.rev() for p in self.parents()])
             if len(parents) > 1:
                 ui.status(_('working directory now based on '
                             'revisions %d and %d\n') % parents)
             else:
                 ui.status(_('working directory now based on '
                             'revision %d\n') % parents)
         # TODO: if we know which new heads may result from this rollback, pass
         # them to destroy(), which will prevent the branchhead cache from being
         # invalidated.
         self.destroyed()
         return 0
 
     def invalidatecaches(self):
         def delcache(name):
             try:
                 delattr(self, name)
             except AttributeError:
                 pass
 
         delcache('_tagscache')
 
         self._branchcache = None # in UTF-8
         self._branchcachetip = None
+        obsolete.clearobscaches(self)
 
     def invalidatedirstate(self):
         '''Invalidates the dirstate, causing the next call to dirstate
         to check if it was modified since the last time it was read,
         rereading it if it has.
 
         This is different to dirstate.invalidate() that it doesn't always
         rereads the dirstate. Use dirstate.invalidate() if you want to
         explicitly read the dirstate again (i.e. restoring it to a previous
         known good state).'''
         if 'dirstate' in self.__dict__:
             for k in self.dirstate._filecache:
                 try:
                     delattr(self.dirstate, k)
                 except AttributeError:
                     pass
             delattr(self, 'dirstate')
 
     def invalidate(self):
         for k in self._filecache:
             # dirstate is invalidated separately in invalidatedirstate()
             if k == 'dirstate':
                 continue
 
             try:
                 delattr(self, k)
             except AttributeError:
                 pass
         self.invalidatecaches()
 
     def _lock(self, lockname, wait, releasefn, acquirefn, desc):
         try:
             l = lock.lock(lockname, 0, releasefn, desc=desc)
         except error.LockHeld, inst:
             if not wait:
                 raise
             self.ui.warn(_("waiting for lock on %s held by %r\n") %
                          (desc, inst.locker))
             # default to 600 seconds timeout
             l = lock.lock(lockname, int(self.ui.config("ui", "timeout", "600")),
                           releasefn, desc=desc)
         if acquirefn:
             acquirefn()
         return l
 
     def _afterlock(self, callback):
         """add a callback to the current repository lock.
 
         The callback will be executed on lock release."""
         l = self._lockref and self._lockref()
         if l:
             l.postrelease.append(callback)
         else:
             callback()
 
     def lock(self, wait=True):
         '''Lock the repository store (.hg/store) and return a weak reference
         to the lock. Use this before modifying the store (e.g. committing or
         stripping). If you are opening a transaction, get a lock as well.)'''
         l = self._lockref and self._lockref()
         if l is not None and l.held:
             l.lock()
             return l
 
         def unlock():
             self.store.write()
             if '_phasecache' in vars(self):
                 self._phasecache.write()
             for k, ce in self._filecache.items():
                 if k == 'dirstate':
                     continue
                 ce.refresh()
 
         l = self._lock(self.sjoin("lock"), wait, unlock,
                        self.invalidate, _('repository %s') % self.origroot)
         self._lockref = weakref.ref(l)
         return l
 
     def wlock(self, wait=True):
         '''Lock the non-store parts of the repository (everything under
         .hg except .hg/store) and return a weak reference to the lock.
         Use this before modifying files in .hg.'''
         l = self._wlockref and self._wlockref()
         if l is not None and l.held:
             l.lock()
             return l
 
         def unlock():
             self.dirstate.write()
             ce = self._filecache.get('dirstate')
             if ce:
                 ce.refresh()
 
         l = self._lock(self.join("wlock"), wait, unlock,
                        self.invalidatedirstate, _('working directory of %s') %
                        self.origroot)
         self._wlockref = weakref.ref(l)
         return l
 
     def _filecommit(self, fctx, manifest1, manifest2, linkrev, tr, changelist):
         """
         commit an individual file as part of a larger transaction
         """
 
         fname = fctx.path()
         text = fctx.data()
         flog = self.file(fname)
         fparent1 = manifest1.get(fname, nullid)
         fparent2 = fparent2o = manifest2.get(fname, nullid)
 
         meta = {}
         copy = fctx.renamed()
         if copy and copy[0] != fname:
             # Mark the new revision of this file as a copy of another
             # file.  This copy data will effectively act as a parent
             # of this new revision.  If this is a merge, the first
             # parent will be the nullid (meaning "look up the copy data")
             # and the second one will be the other parent.  For example:
             #
             # 0 --- 1 --- 3   rev1 changes file foo
             #   \       /     rev2 renames foo to bar and changes it
             #    \- 2 -/      rev3 should have bar with all changes and
             #                      should record that bar descends from
             #                      bar in rev2 and foo in rev1
             #
             # this allows this merge to succeed:
             #
             # 0 --- 1 --- 3   rev4 reverts the content change from rev2
             #   \       /     merging rev3 and rev4 should use bar@rev2
             #    \- 2 --- 4        as the merge base
             #
 
             cfname = copy[0]
             crev = manifest1.get(cfname)
             newfparent = fparent2
 
             if manifest2: # branch merge
                 if fparent2 == nullid or crev is None: # copied on remote side
                     if cfname in manifest2:
                         crev = manifest2[cfname]
                         newfparent = fparent1
 
             # find source in nearest ancestor if we've lost track
             if not crev:
                 self.ui.debug(" %s: searching for copy revision for %s\n" %
                               (fname, cfname))
                 for ancestor in self[None].ancestors():
                     if cfname in ancestor:
                         crev = ancestor[cfname].filenode()
                         break
 
             if crev:
                 self.ui.debug(" %s: copy %s:%s\n" % (fname, cfname, hex(crev)))
                 meta["copy"] = cfname
                 meta["copyrev"] = hex(crev)
                 fparent1, fparent2 = nullid, newfparent
             else:
                 self.ui.warn(_("warning: can't find ancestor for '%s' "
                                "copied from '%s'!\n") % (fname, cfname))
 
         elif fparent2 != nullid:
             # is one parent an ancestor of the other?
             fparentancestor = flog.ancestor(fparent1, fparent2)
             if fparentancestor == fparent1:
                 fparent1, fparent2 = fparent2, nullid
             elif fparentancestor == fparent2:
                 fparent2 = nullid
 
         # is the file changed?
         if fparent2 != nullid or flog.cmp(fparent1, text) or meta:
             changelist.append(fname)
             return flog.add(text, meta, tr, linkrev, fparent1, fparent2)
 
         # are just the flags changed during merge?
         if fparent1 != fparent2o and manifest1.flags(fname) != fctx.flags():
             changelist.append(fname)
 
         return fparent1
 
     def commit(self, text="", user=None, date=None, match=None, force=False,
                editor=False, extra={}):
         """Add a new revision to current repository.
 
         Revision information is gathered from the working directory,
         match can be used to filter the committed files. If editor is
         supplied, it is called to get a commit message.
         """
 
         def fail(f, msg):
             raise util.Abort('%s: %s' % (f, msg))
 
         if not match:
             match = matchmod.always(self.root, '')
 
         if not force:
             vdirs = []
             match.dir = vdirs.append
             match.bad = fail
 
         wlock = self.wlock()
         try:
             wctx = self[None]
             merge = len(wctx.parents()) > 1
 
             if (not force and merge and match and
                 (match.files() or match.anypats())):
                 raise util.Abort(_('cannot partially commit a merge '
                                    '(do not specify files or patterns)'))
 
             changes = self.status(match=match, clean=force)
             if force:
                 changes[0].extend(changes[6]) # mq may commit unchanged files
 
             # check subrepos
             subs = []
             commitsubs = set()
             newstate = wctx.substate.copy()
             # only manage subrepos and .hgsubstate if .hgsub is present
             if '.hgsub' in wctx:
                 # we'll decide whether to track this ourselves, thanks
                 if '.hgsubstate' in changes[0]:
                     changes[0].remove('.hgsubstate')
                 if '.hgsubstate' in changes[2]:
                     changes[2].remove('.hgsubstate')
 
                 # compare current state to last committed state
                 # build new substate based on last committed state
                 oldstate = wctx.p1().substate
                 for s in sorted(newstate.keys()):
                     if not match(s):
                         # ignore working copy, use old state if present
                         if s in oldstate:
                             newstate[s] = oldstate[s]
                             continue
                         if not force:
                             raise util.Abort(
                                 _("commit with new subrepo %s excluded") % s)
                     if wctx.sub(s).dirty(True):
                         if not self.ui.configbool('ui', 'commitsubrepos'):
                             raise util.Abort(
                                 _("uncommitted changes in subrepo %s") % s,
                                 hint=_("use --subrepos for recursive commit"))
                         subs.append(s)
                         commitsubs.add(s)
                     else:
                         bs = wctx.sub(s).basestate()
                         newstate[s] = (newstate[s][0], bs, newstate[s][2])
                         if oldstate.get(s, (None, None, None))[1] != bs:
                             subs.append(s)
 
                 # check for removed subrepos
                 for p in wctx.parents():
                     r = [s for s in p.substate if s not in newstate]
                     subs += [s for s in r if match(s)]
                 if subs:
                     if (not match('.hgsub') and
                         '.hgsub' in (wctx.modified() + wctx.added())):
                         raise util.Abort(
                             _("can't commit subrepos without .hgsub"))
                     changes[0].insert(0, '.hgsubstate')
 
             elif '.hgsub' in changes[2]:
                 # clean up .hgsubstate when .hgsub is removed
                 if ('.hgsubstate' in wctx and
                     '.hgsubstate' not in changes[0] + changes[1] + changes[2]):
                     changes[2].insert(0, '.hgsubstate')
 
             # make sure all explicit patterns are matched
             if not force and match.files():
                 matched = set(changes[0] + changes[1] + changes[2])
 
                 for f in match.files():
                     f = self.dirstate.normalize(f)
                     if f == '.' or f in matched or f in wctx.substate:
                         continue
                     if f in changes[3]: # missing
                         fail(f, _('file not found!'))
                     if f in vdirs: # visited directory
                         d = f + '/'
                         for mf in matched:
                             if mf.startswith(d):
                                 break
                         else:
                             fail(f, _("no match under directory!"))
                     elif f not in self.dirstate:
                         fail(f, _("file not tracked!"))
 
             if (not force and not extra.get("close") and not merge
                 and not (changes[0] or changes[1] or changes[2])
                 and wctx.branch() == wctx.p1().branch()):
                 return None
 
             if merge and changes[3]:
                 raise util.Abort(_("cannot commit merge with missing files"))
 
             ms = mergemod.mergestate(self)
             for f in changes[0]:
                 if f in ms and ms[f] == 'u':
                     raise util.Abort(_("unresolved merge conflicts "
                                        "(see hg help resolve)"))
 
             cctx = context.workingctx(self, text, user, date, extra, changes)
             if editor:
                 cctx._text = editor(self, cctx, subs)
             edited = (text != cctx._text)
 
             # commit subs and write new state
             if subs:
                 for s in sorted(commitsubs):
                     sub = wctx.sub(s)
                     self.ui.status(_('committing subrepository %s\n') %
                         subrepo.subrelpath(sub))
                     sr = sub.commit(cctx._text, user, date)
                     newstate[s] = (newstate[s][0], sr)
                 subrepo.writestate(self, newstate)
 
             # Save commit message in case this transaction gets rolled back
             # (e.g. by a pretxncommit hook).  Leave the content alone on
             # the assumption that the user will use the same editor again.
             msgfn = self.savecommitmessage(cctx._text)
 
             p1, p2 = self.dirstate.parents()
             hookp1, hookp2 = hex(p1), (p2 != nullid and hex(p2) or '')
             try:
                 self.hook("precommit", throw=True, parent1=hookp1,
                           parent2=hookp2)
                 ret = self.commitctx(cctx, True)
             except: # re-raises
                 if edited:
                     self.ui.write(
                         _('note: commit message saved in %s\n') % msgfn)
                 raise
 
             # update bookmarks, dirstate and mergestate
             bookmarks.update(self, [p1, p2], ret)
             for f in changes[0] + changes[1]:
                 self.dirstate.normal(f)
             for f in changes[2]:
                 self.dirstate.drop(f)
             self.dirstate.setparents(ret)
             ms.reset()
         finally:
             wlock.release()
 
         def commithook(node=hex(ret), parent1=hookp1, parent2=hookp2):
             self.hook("commit", node=node, parent1=parent1, parent2=parent2)
         self._afterlock(commithook)
         return ret
 
     def commitctx(self, ctx, error=False):
         """Add a new revision to current repository.
         Revision information is passed via the context argument.
         """
 
         tr = lock = None
         removed = list(ctx.removed())
         p1, p2 = ctx.p1(), ctx.p2()
         user = ctx.user()
 
         lock = self.lock()
         try:
             tr = self.transaction("commit")
             trp = weakref.proxy(tr)
 
             if ctx.files():
                 m1 = p1.manifest().copy()
                 m2 = p2.manifest()
 
                 # check in files
                 new = {}
                 changed = []
                 linkrev = len(self)
                 for f in sorted(ctx.modified() + ctx.added()):
                     self.ui.note(f + "\n")
                     try:
                         fctx = ctx[f]
                         new[f] = self._filecommit(fctx, m1, m2, linkrev, trp,
                                                   changed)
                         m1.set(f, fctx.flags())
                     except OSError, inst:
                         self.ui.warn(_("trouble committing %s!\n") % f)
                         raise
                     except IOError, inst:
                         errcode = getattr(inst, 'errno', errno.ENOENT)
                         if error or errcode and errcode != errno.ENOENT:
                             self.ui.warn(_("trouble committing %s!\n") % f)
                             raise
                         else:
                             removed.append(f)
 
                 # update manifest
                 m1.update(new)
                 removed = [f for f in sorted(removed) if f in m1 or f in m2]
                 drop = [f for f in removed if f in m1]
                 for f in drop:
                     del m1[f]
                 mn = self.manifest.add(m1, trp, linkrev, p1.manifestnode(),
                                        p2.manifestnode(), (new, drop))
                 files = changed + removed
             else:
                 mn = p1.manifestnode()
                 files = []
 
             # update changelog
             self.changelog.delayupdate()
             n = self.changelog.add(mn, files, ctx.description(),
                                    trp, p1.node(), p2.node(),
                                    user, ctx.date(), ctx.extra().copy())
             p = lambda: self.changelog.writepending() and self.root or ""
             xp1, xp2 = p1.hex(), p2 and p2.hex() or ''
             self.hook('pretxncommit', throw=True, node=hex(n), parent1=xp1,
                       parent2=xp2, pending=p)
             self.changelog.finalize(trp)
             # set the new commit is proper phase
             targetphase = phases.newcommitphase(self.ui)
             if targetphase:
                 # retract boundary do not alter parent changeset.
                 # if a parent have higher the resulting phase will
                 # be compliant anyway
                 #
                 # if minimal phase was 0 we don't need to retract anything
                 phases.retractboundary(self, targetphase, [n])
             tr.close()
             self.updatebranchcache()
             return n
         finally:
             if tr:
                 tr.release()
             lock.release()
 
     def destroyed(self, newheadnodes=None):
         '''Inform the repository that nodes have been destroyed.
         Intended for use by strip and rollback, so there's a common
         place for anything that has to be done after destroying history.
 
         If you know the branchheadcache was uptodate before nodes were removed
         and you also know the set of candidate new heads that may have resulted
         from the destruction, you can set newheadnodes.  This will enable the
         code to update the branchheads cache, rather than having future code
         decide it's invalid and regenerating it from scratch.
         '''
         # If we have info, newheadnodes, on how to update the branch cache, do
         # it, Otherwise, since nodes were destroyed, the cache is stale and this
         # will be caught the next time it is read.
         if newheadnodes:
             tiprev = len(self) - 1
             ctxgen = (self[node] for node in newheadnodes
                       if self.changelog.hasnode(node))
             self._updatebranchcache(self._branchcache, ctxgen)
             self._writebranchcache(self._branchcache, self.changelog.tip(),
                                    tiprev)
 
         # Ensure the persistent tag cache is updated.  Doing it now
         # means that the tag cache only has to worry about destroyed
         # heads immediately after a strip/rollback.  That in turn
         # guarantees that "cachetip == currenttip" (comparing both rev
         # and node) always means no nodes have been added or destroyed.
 
         # XXX this is suboptimal when qrefresh'ing: we strip the current
         # head, refresh the tag cache, then immediately add a new head.
         # But I think doing it this way is necessary for the "instant
         # tag cache retrieval" case to work.
         self.invalidatecaches()
 
         # Discard all cache entries to force reloading everything.
         self._filecache.clear()
 
     def walk(self, match, node=None):
         '''
         walk recursively through the directory tree or a given
         changeset, finding all files matched by the match
         function
         '''
         return self[node].walk(match)
 
     def status(self, node1='.', node2=None, match=None,
                ignored=False, clean=False, unknown=False,
                listsubrepos=False):
         """return status of files between two nodes or node and working
         directory.
 
         If node1 is None, use the first dirstate parent instead.
         If node2 is None, compare node1 with working directory.
         """
 
         def mfmatches(ctx):
             mf = ctx.manifest().copy()
             if match.always():
                 return mf
             for fn in mf.keys():
                 if not match(fn):
                     del mf[fn]
             return mf
 
         if isinstance(node1, context.changectx):
             ctx1 = node1
         else:
             ctx1 = self[node1]
         if isinstance(node2, context.changectx):
             ctx2 = node2
         else:
             ctx2 = self[node2]
 
         working = ctx2.rev() is None
         parentworking = working and ctx1 == self['.']
         match = match or matchmod.always(self.root, self.getcwd())
         listignored, listclean, listunknown = ignored, clean, unknown
 
         # load earliest manifest first for caching reasons
         if not working and ctx2.rev() < ctx1.rev():
             ctx2.manifest()
 
         if not parentworking:
             def bad(f, msg):
                 # 'f' may be a directory pattern from 'match.files()',
                 # so 'f not in ctx1' is not enough
                 if f not in ctx1 and f not in ctx1.dirs():
                     self.ui.warn('%s: %s\n' % (self.dirstate.pathto(f), msg))
             match.bad = bad
 
         if working: # we need to scan the working dir
             subrepos = []
             if '.hgsub' in self.dirstate:
                 subrepos = ctx2.substate.keys()
             s = self.dirstate.status(match, subrepos, listignored,
                                      listclean, listunknown)
             cmp, modified, added, removed, deleted, unknown, ignored, clean = s
 
             # check for any possibly clean files
             if parentworking and cmp:
                 fixup = []
                 # do a full compare of any files that might have changed
                 for f in sorted(cmp):
                     if (f not in ctx1 or ctx2.flags(f) != ctx1.flags(f)
                         or ctx1[f].cmp(ctx2[f])):
                         modified.append(f)
                     else:
                         fixup.append(f)
 
                 # update dirstate for files that are actually clean
                 if fixup:
                     if listclean:
                         clean += fixup
 
                     try:
                         # updating the dirstate is optional
                         # so we don't wait on the lock
                         wlock = self.wlock(False)
                         try:
                             for f in fixup:
                                 self.dirstate.normal(f)
                         finally:
                             wlock.release()
                     except error.LockError:
                         pass
 
         if not parentworking:
             mf1 = mfmatches(ctx1)
             if working:
                 # we are comparing working dir against non-parent
                 # generate a pseudo-manifest for the working dir
                 mf2 = mfmatches(self['.'])
                 for f in cmp + modified + added:
                     mf2[f] = None
                     mf2.set(f, ctx2.flags(f))
                 for f in removed:
                     if f in mf2:
                         del mf2[f]
             else:
                 # we are comparing two revisions
                 deleted, unknown, ignored = [], [], []
                 mf2 = mfmatches(ctx2)
 
             modified, added, clean = [], [], []
             withflags = mf1.withflags() | mf2.withflags()
             for fn in mf2:
                 if fn in mf1:
                     if (fn not in deleted and
                         ((fn in withflags and mf1.flags(fn) != mf2.flags(fn)) or
                          (mf1[fn] != mf2[fn] and
                           (mf2[fn] or ctx1[fn].cmp(ctx2[fn]))))):
                         modified.append(fn)
                     elif listclean:
                         clean.append(fn)
                     del mf1[fn]
                 elif fn not in deleted:
                     added.append(fn)
             removed = mf1.keys()
 
         if working and modified and not self.dirstate._checklink:
             # Symlink placeholders may get non-symlink-like contents
             # via user error or dereferencing by NFS or Samba servers,
             # so we filter out any placeholders that don't look like a
             # symlink
             sane = []
             for f in modified:
                 if ctx2.flags(f) == 'l':
                     d = ctx2[f].data()
                     if len(d) >= 1024 or '\n' in d or util.binary(d):
                         self.ui.debug('ignoring suspect symlink placeholder'
                                       ' "%s"\n' % f)
                         continue
                 sane.append(f)
             modified = sane
 
         r = modified, added, removed, deleted, unknown, ignored, clean
 
         if listsubrepos:
             for subpath, sub in subrepo.itersubrepos(ctx1, ctx2):
                 if working:
                     rev2 = None
                 else:
                     rev2 = ctx2.substate[subpath][1]
                 try:
                     submatch = matchmod.narrowmatcher(subpath, match)
                     s = sub.status(rev2, match=submatch, ignored=listignored,
                                    clean=listclean, unknown=listunknown,
                                    listsubrepos=True)
                     for rfiles, sfiles in zip(r, s):
                         rfiles.extend("%s/%s" % (subpath, f) for f in sfiles)
                 except error.LookupError:
                     self.ui.status(_("skipping missing subrepository: %s\n")
                                    % subpath)
 
         for l in r:
             l.sort()
         return r
 
     def heads(self, start=None):
         heads = self.changelog.heads(start)
         # sort the output in rev descending order
         return sorted(heads, key=self.changelog.rev, reverse=True)
 
     def branchheads(self, branch=None, start=None, closed=False):
         '''return a (possibly filtered) list of heads for the given branch
 
         Heads are returned in topological order, from newest to oldest.
         If branch is None, use the dirstate branch.
         If start is not None, return only heads reachable from start.
         If closed is True, return heads that are marked as closed as well.
         '''
         if branch is None:
             branch = self[None].branch()
         branches = self.branchmap()
         if branch not in branches:
             return []
         # the cache returns heads ordered lowest to highest
         bheads = list(reversed(branches[branch]))
         if start is not None:
             # filter out the heads that cannot be reached from startrev
             fbheads = set(self.changelog.nodesbetween([start], bheads)[2])
             bheads = [h for h in bheads if h in fbheads]
         if not closed:
             bheads = [h for h in bheads if not self[h].closesbranch()]
         return bheads
 
     def branches(self, nodes):
         if not nodes:
             nodes = [self.changelog.tip()]
         b = []
         for n in nodes:
             t = n
             while True:
                 p = self.changelog.parents(n)
                 if p[1] != nullid or p[0] == nullid:
                     b.append((t, n, p[0], p[1]))
                     break
                 n = p[0]
         return b
 
     def between(self, pairs):
         r = []
 
         for top, bottom in pairs:
             n, l, i = top, [], 0
             f = 1
 
             while n != bottom and n != nullid:
                 p = self.changelog.parents(n)[0]
                 if i == f:
                     l.append(n)
                     f = f * 2
                 n = p
                 i += 1
 
             r.append(l)
 
         return r
 
     def pull(self, remote, heads=None, force=False):
         # don't open transaction for nothing or you break future useful
         # rollback call
         tr = None
         trname = 'pull\n' + util.hidepassword(remote.url())
         lock = self.lock()
         try:
             tmp = discovery.findcommonincoming(self, remote, heads=heads,
                                                force=force)
             common, fetch, rheads = tmp
             if not fetch:
                 self.ui.status(_("no changes found\n"))
                 added = []
                 result = 0
             else:
                 tr = self.transaction(trname)
                 if heads is None and list(common) == [nullid]:
                     self.ui.status(_("requesting all changes\n"))
                 elif heads is None and remote.capable('changegroupsubset'):
                     # issue1320, avoid a race if remote changed after discovery
                     heads = rheads
 
                 if remote.capable('getbundle'):
                     cg = remote.getbundle('pull', common=common,
                                           heads=heads or rheads)
                 elif heads is None:
                     cg = remote.changegroup(fetch, 'pull')
                 elif not remote.capable('changegroupsubset'):
                     raise util.Abort(_("partial pull cannot be done because "
                                            "other repository doesn't support "
                                            "changegroupsubset."))
                 else:
                     cg = remote.changegroupsubset(fetch, heads, 'pull')
                 clstart = len(self.changelog)
                 result = self.addchangegroup(cg, 'pull', remote.url())
                 clend = len(self.changelog)
                 added = [self.changelog.node(r) for r in xrange(clstart, clend)]
 
             # compute target subset
             if heads is None:
                 # We pulled every thing possible
                 # sync on everything common
                 subset = common + added
             else:
                 # We pulled a specific subset
                 # sync on this subset
                 subset = heads
 
             # Get remote phases data from remote
             remotephases = remote.listkeys('phases')
             publishing = bool(remotephases.get('publishing', False))
             if remotephases and not publishing:
                 # remote is new and unpublishing
                 pheads, _dr = phases.analyzeremotephases(self, subset,
                                                          remotephases)
                 phases.advanceboundary(self, phases.public, pheads)
                 phases.advanceboundary(self, phases.draft, subset)
             else:
                 # Remote is old or publishing all common changesets
                 # should be seen as public
                 phases.advanceboundary(self, phases.public, subset)
 
             if obsolete._enabled:
                 self.ui.debug('fetching remote obsolete markers')
                 remoteobs = remote.listkeys('obsolete')
                 if 'dump0' in remoteobs:
                     if tr is None:
                         tr = self.transaction(trname)
                     for key in sorted(remoteobs, reverse=True):
                         if key.startswith('dump'):
                             data = base85.b85decode(remoteobs[key])
                             self.obsstore.mergemarkers(tr, data)
             if tr is not None:
                 tr.close()
         finally:
             if tr is not None:
                 tr.release()
             lock.release()
 
         return result
 
     def checkpush(self, force, revs):
         """Extensions can override this function if additional checks have
         to be performed before pushing, or call it if they override push
         command.
         """
         pass
 
     def push(self, remote, force=False, revs=None, newbranch=False):
         '''Push outgoing changesets (limited by revs) from the current
         repository to remote. Return an integer:
           - None means nothing to push
           - 0 means HTTP error
           - 1 means we pushed and remote head count is unchanged *or*
             we have outgoing changesets but refused to push
           - other values as described by addchangegroup()
         '''
         # there are two ways to push to remote repo:
         #
         # addchangegroup assumes local user can lock remote
         # repo (local filesystem, old ssh servers).
         #
         # unbundle assumes local user cannot lock remote repo (new ssh
         # servers, http servers).
 
         if not remote.canpush():
             raise util.Abort(_("destination does not support push"))
         # get local lock as we might write phase data
         locallock = self.lock()
         try:
             self.checkpush(force, revs)
             lock = None
             unbundle = remote.capable('unbundle')
             if not unbundle:
                 lock = remote.lock()
             try:
                 # discovery
                 fci = discovery.findcommonincoming
                 commoninc = fci(self, remote, force=force)
                 common, inc, remoteheads = commoninc
                 fco = discovery.findcommonoutgoing
                 outgoing = fco(self, remote, onlyheads=revs,
                                commoninc=commoninc, force=force)
 
 
                 if not outgoing.missing:
                     # nothing to push
                     scmutil.nochangesfound(self.ui, self, outgoing.excluded)
                     ret = None
                 else:
                     # something to push
                     if not force:
                         # if self.obsstore == False --> no obsolete
                         # then, save the iteration
                         if self.obsstore:
                             # this message are here for 80 char limit reason
                             mso = _("push includes an obsolete changeset: %s!")
                             msu = _("push includes an unstable changeset: %s!")
                             # If we are to push if there is at least one
                             # obsolete or unstable changeset in missing, at
                             # least one of the missinghead will be obsolete or
                             # unstable. So checking heads only is ok
                             for node in outgoing.missingheads:
                                 ctx = self[node]
                                 if ctx.obsolete():
                                     raise util.Abort(_(mso) % ctx)
                                 elif ctx.unstable():
                                     raise util.Abort(_(msu) % ctx)
                         discovery.checkheads(self, remote, outgoing,
                                              remoteheads, newbranch,
                                              bool(inc))
 
                     # create a changegroup from local
                     if revs is None and not outgoing.excluded:
                         # push everything,
                         # use the fast path, no race possible on push
                         cg = self._changegroup(outgoing.missing, 'push')
                     else:
                         cg = self.getlocalbundle('push', outgoing)
 
                     # apply changegroup to remote
                     if unbundle:
                         # local repo finds heads on server, finds out what
                         # revs it must push. once revs transferred, if server
                         # finds it has different heads (someone else won
                         # commit/push race), server aborts.
                         if force:
                             remoteheads = ['force']
                         # ssh: return remote's addchangegroup()
                         # http: return remote's addchangegroup() or 0 for error
                         ret = remote.unbundle(cg, remoteheads, 'push')
                     else:
                         # we return an integer indicating remote head count
                         # change
                         ret = remote.addchangegroup(cg, 'push', self.url())
 
                 if ret:
                     # push succeed, synchronize target of the push
                     cheads = outgoing.missingheads
                 elif revs is None:
                     # All out push fails. synchronize all common
                     cheads = outgoing.commonheads
                 else:
                     # I want cheads = heads(::missingheads and ::commonheads)
                     # (missingheads is revs with secret changeset filtered out)
                     #
                     # This can be expressed as:
                     #     cheads = ( (missingheads and ::commonheads)
                     #              + (commonheads and ::missingheads))"
                     #              )
                     #
                     # while trying to push we already computed the following:
                     #     common = (::commonheads)
                     #     missing = ((commonheads::missingheads) - commonheads)
                     #
                     # We can pick:
                     # * missingheads part of common (::commonheads)
                     common = set(outgoing.common)
                     cheads = [node for node in revs if node in common]
                     # and
                     # * commonheads parents on missing
                     revset = self.set('%ln and parents(roots(%ln))',
                                      outgoing.commonheads,
                                      outgoing.missing)
                     cheads.extend(c.node() for c in revset)
                 # even when we don't push, exchanging phase data is useful
                 remotephases = remote.listkeys('phases')
                 if not remotephases: # old server or public only repo
                     phases.advanceboundary(self, phases.public, cheads)
                     # don't push any phase data as there is nothing to push
                 else:
                     ana = phases.analyzeremotephases(self, cheads, remotephases)
                     pheads, droots = ana
                     ### Apply remote phase on local
                     if remotephases.get('publishing', False):
                         phases.advanceboundary(self, phases.public, cheads)
                     else: # publish = False
                         phases.advanceboundary(self, phases.public, pheads)
                         phases.advanceboundary(self, phases.draft, cheads)
                     ### Apply local phase on remote
 
                     # Get the list of all revs draft on remote by public here.
                     # XXX Beware that revset break if droots is not strictly
                     # XXX root we may want to ensure it is but it is costly
                     outdated =  self.set('heads((%ln::%ln) and public())',
                                          droots, cheads)
                     for newremotehead in outdated:
                         r = remote.pushkey('phases',
                                            newremotehead.hex(),
                                            str(phases.draft),
                                            str(phases.public))
                         if not r:
                             self.ui.warn(_('updating %s to public failed!\n')
                                             % newremotehead)
                 self.ui.debug('try to push obsolete markers to remote\n')
                 if (obsolete._enabled and self.obsstore and
                     'obsolete' in remote.listkeys('namespaces')):
                     rslts = []
                     remotedata = self.listkeys('obsolete')
                     for key in sorted(remotedata, reverse=True):
                         # reverse sort to ensure we end with dump0
                         data = remotedata[key]
                         rslts.append(remote.pushkey('obsolete', key, '', data))
                     if [r for r in rslts if not r]:
                         msg = _('failed to push some obsolete markers!\n')
                         self.ui.warn(msg)
             finally:
                 if lock is not None:
                     lock.release()
         finally:
             locallock.release()
 
         self.ui.debug("checking for updated bookmarks\n")
         rb = remote.listkeys('bookmarks')
         for k in rb.keys():
             if k in self._bookmarks:
                 nr, nl = rb[k], hex(self._bookmarks[k])
                 if nr in self:
                     cr = self[nr]
                     cl = self[nl]
                     if cl in cr.descendants():
                         r = remote.pushkey('bookmarks', k, nr, nl)
                         if r:
                             self.ui.status(_("updating bookmark %s\n") % k)
                         else:
                             self.ui.warn(_('updating bookmark %s'
                                            ' failed!\n') % k)
 
         return ret
 
     def changegroupinfo(self, nodes, source):
         if self.ui.verbose or source == 'bundle':
             self.ui.status(_("%d changesets found\n") % len(nodes))
         if self.ui.debugflag:
             self.ui.debug("list of changesets:\n")
             for node in nodes:
                 self.ui.debug("%s\n" % hex(node))
 
     def changegroupsubset(self, bases, heads, source):
         """Compute a changegroup consisting of all the nodes that are
         descendants of any of the bases and ancestors of any of the heads.
         Return a chunkbuffer object whose read() method will return
         successive changegroup chunks.
 
         It is fairly complex as determining which filenodes and which
         manifest nodes need to be included for the changeset to be complete
         is non-trivial.
 
         Another wrinkle is doing the reverse, figuring out which changeset in
         the changegroup a particular filenode or manifestnode belongs to.
         """
         cl = self.changelog
         if not bases:
             bases = [nullid]
         csets, bases, heads = cl.nodesbetween(bases, heads)
         # We assume that all ancestors of bases are known
         common = set(cl.ancestors([cl.rev(n) for n in bases]))
         return self._changegroupsubset(common, csets, heads, source)
 
     def getlocalbundle(self, source, outgoing):
         """Like getbundle, but taking a discovery.outgoing as an argument.
 
         This is only implemented for local repos and reuses potentially
         precomputed sets in outgoing."""
         if not outgoing.missing:
             return None
         return self._changegroupsubset(outgoing.common,
                                        outgoing.missing,
                                        outgoing.missingheads,
                                        source)
 
     def getbundle(self, source, heads=None, common=None):
         """Like changegroupsubset, but returns the set difference between the
         ancestors of heads and the ancestors common.
 
         If heads is None, use the local heads. If common is None, use [nullid].
 
         The nodes in common might not all be known locally due to the way the
         current discovery protocol works.
         """
         cl = self.changelog
         if common:
             nm = cl.nodemap
             common = [n for n in common if n in nm]
         else:
             common = [nullid]
         if not heads:
             heads = cl.heads()
         return self.getlocalbundle(source,
                                    discovery.outgoing(cl, common, heads))
 
     def _changegroupsubset(self, commonrevs, csets, heads, source):
 
         cl = self.changelog
         mf = self.manifest
         mfs = {} # needed manifests
         fnodes = {} # needed file nodes
         changedfiles = set()
         fstate = ['', {}]
         count = [0, 0]
 
         # can we go through the fast path ?
         heads.sort()
         if heads == sorted(self.heads()):
             return self._changegroup(csets, source)
 
         # slow path
         self.hook('preoutgoing', throw=True, source=source)
         self.changegroupinfo(csets, source)
 
         # filter any nodes that claim to be part of the known set
         def prune(revlog, missing):
             rr, rl = revlog.rev, revlog.linkrev
             return [n for n in missing
                     if rl(rr(n)) not in commonrevs]
 
         progress = self.ui.progress
         _bundling = _('bundling')
         _changesets = _('changesets')
         _manifests = _('manifests')
         _files = _('files')
 
         def lookup(revlog, x):
             if revlog == cl:
                 c = cl.read(x)
                 changedfiles.update(c[3])
                 mfs.setdefault(c[0], x)
                 count[0] += 1
                 progress(_bundling, count[0],
                          unit=_changesets, total=count[1])
                 return x
             elif revlog == mf:
                 clnode = mfs[x]
                 mdata = mf.readfast(x)
                 for f, n in mdata.iteritems():
                     if f in changedfiles:
                         fnodes[f].setdefault(n, clnode)
                 count[0] += 1
                 progress(_bundling, count[0],
                          unit=_manifests, total=count[1])
                 return clnode
             else:
                 progress(_bundling, count[0], item=fstate[0],
                          unit=_files, total=count[1])
                 return fstate[1][x]
 
         bundler = changegroup.bundle10(lookup)
         reorder = self.ui.config('bundle', 'reorder', 'auto')
         if reorder == 'auto':
             reorder = None
         else:
             reorder = util.parsebool(reorder)
 
         def gengroup():
             # Create a changenode group generator that will call our functions
             # back to lookup the owning changenode and collect information.
             count[:] = [0, len(csets)]
             for chunk in cl.group(csets, bundler, reorder=reorder):
                 yield chunk
             progress(_bundling, None)
 
             # Create a generator for the manifestnodes that calls our lookup
             # and data collection functions back.
             for f in changedfiles:
                 fnodes[f] = {}
             count[:] = [0, len(mfs)]
             for chunk in mf.group(prune(mf, mfs), bundler, reorder=reorder):
                 yield chunk
             progress(_bundling, None)
 
             mfs.clear()
 
             # Go through all our files in order sorted by name.
             count[:] = [0, len(changedfiles)]
             for fname in sorted(changedfiles):
                 filerevlog = self.file(fname)
                 if not len(filerevlog):
                     raise util.Abort(_("empty or missing revlog for %s")
                                      % fname)
                 fstate[0] = fname
                 fstate[1] = fnodes.pop(fname, {})
 
                 nodelist = prune(filerevlog, fstate[1])
                 if nodelist:
                     count[0] += 1
                     yield bundler.fileheader(fname)
                     for chunk in filerevlog.group(nodelist, bundler, reorder):
                         yield chunk
 
             # Signal that no more groups are left.
             yield bundler.close()
             progress(_bundling, None)
 
             if csets:
                 self.hook('outgoing', node=hex(csets[0]), source=source)
 
         return changegroup.unbundle10(util.chunkbuffer(gengroup()), 'UN')
 
     def changegroup(self, basenodes, source):
         # to avoid a race we use changegroupsubset() (issue1320)
         return self.changegroupsubset(basenodes, self.heads(), source)
 
     def _changegroup(self, nodes, source):
         """Compute the changegroup of all nodes that we have that a recipient
         doesn't.  Return a chunkbuffer object whose read() method will return
         successive changegroup chunks.
 
         This is much easier than the previous function as we can assume that
         the recipient has any changenode we aren't sending them.
 
         nodes is the set of nodes to send"""
 
         cl = self.changelog
         mf = self.manifest
         mfs = {}
         changedfiles = set()
         fstate = ['']
         count = [0, 0]
 
         self.hook('preoutgoing', throw=True, source=source)
         self.changegroupinfo(nodes, source)
 
         revset = set([cl.rev(n) for n in nodes])
 
         def gennodelst(log):
             ln, llr = log.node, log.linkrev
             return [ln(r) for r in log if llr(r) in revset]
 
         progress = self.ui.progress
         _bundling = _('bundling')
         _changesets = _('changesets')
         _manifests = _('manifests')
         _files = _('files')
 
         def lookup(revlog, x):
             if revlog == cl:
                 c = cl.read(x)
                 changedfiles.update(c[3])
                 mfs.setdefault(c[0], x)
                 count[0] += 1
                 progress(_bundling, count[0],
                          unit=_changesets, total=count[1])
                 return x
             elif revlog == mf:
                 count[0] += 1
                 progress(_bundling, count[0],
                          unit=_manifests, total=count[1])
                 return cl.node(revlog.linkrev(revlog.rev(x)))
             else:
                 progress(_bundling, count[0], item=fstate[0],
                     total=count[1], unit=_files)
                 return cl.node(revlog.linkrev(revlog.rev(x)))
 
         bundler = changegroup.bundle10(lookup)
         reorder = self.ui.config('bundle', 'reorder', 'auto')
         if reorder == 'auto':
             reorder = None
         else:
             reorder = util.parsebool(reorder)
 
         def gengroup():
             '''yield a sequence of changegroup chunks (strings)'''
             # construct a list of all changed files
 
             count[:] = [0, len(nodes)]
             for chunk in cl.group(nodes, bundler, reorder=reorder):
                 yield chunk
             progress(_bundling, None)
 
             count[:] = [0, len(mfs)]
             for chunk in mf.group(gennodelst(mf), bundler, reorder=reorder):
                 yield chunk
             progress(_bundling, None)
 
             count[:] = [0, len(changedfiles)]
             for fname in sorted(changedfiles):
                 filerevlog = self.file(fname)
                 if not len(filerevlog):
                     raise util.Abort(_("empty or missing revlog for %s")
                                      % fname)
                 fstate[0] = fname
                 nodelist = gennodelst(filerevlog)
                 if nodelist:
                     count[0] += 1
                     yield bundler.fileheader(fname)
                     for chunk in filerevlog.group(nodelist, bundler, reorder):
                         yield chunk
             yield bundler.close()
             progress(_bundling, None)
 
             if nodes:
                 self.hook('outgoing', node=hex(nodes[0]), source=source)
 
         return changegroup.unbundle10(util.chunkbuffer(gengroup()), 'UN')
 
     def addchangegroup(self, source, srctype, url, emptyok=False):
         """Add the changegroup returned by source.read() to this repo.
         srctype is a string like 'push', 'pull', or 'unbundle'.  url is
         the URL of the repo where this changegroup is coming from.
 
         Return an integer summarizing the change to this repo:
         - nothing changed or no source: 0
         - more heads than before: 1+added heads (2..n)
         - fewer heads than before: -1-removed heads (-2..-n)
         - number of heads stays the same: 1
         """
         def csmap(x):
             self.ui.debug("add changeset %s\n" % short(x))
             return len(cl)
 
         def revmap(x):
             return cl.rev(x)
 
         if not source:
             return 0
 
         self.hook('prechangegroup', throw=True, source=srctype, url=url)
 
         changesets = files = revisions = 0
         efiles = set()
 
         # write changelog data to temp files so concurrent readers will not see
         # inconsistent view
         cl = self.changelog
         cl.delayupdate()
         oldheads = cl.heads()
 
         tr = self.transaction("\n".join([srctype, util.hidepassword(url)]))
         try:
             trp = weakref.proxy(tr)
             # pull off the changeset group
             self.ui.status(_("adding changesets\n"))
             clstart = len(cl)
             class prog(object):
                 step = _('changesets')
                 count = 1
                 ui = self.ui
                 total = None
                 def __call__(self):
                     self.ui.progress(self.step, self.count, unit=_('chunks'),
                                      total=self.total)
                     self.count += 1
             pr = prog()
             source.callback = pr
 
             source.changelogheader()
             srccontent = cl.addgroup(source, csmap, trp)
             if not (srccontent or emptyok):
                 raise util.Abort(_("received changelog group is empty"))
             clend = len(cl)
             changesets = clend - clstart
             for c in xrange(clstart, clend):
                 efiles.update(self[c].files())
             efiles = len(efiles)
             self.ui.progress(_('changesets'), None)
 
             # pull off the manifest group
             self.ui.status(_("adding manifests\n"))
             pr.step = _('manifests')
             pr.count = 1
             pr.total = changesets # manifests <= changesets
             # no need to check for empty manifest group here:
             # if the result of the merge of 1 and 2 is the same in 3 and 4,
             # no new manifest will be created and the manifest group will
             # be empty during the pull
             source.manifestheader()
             self.manifest.addgroup(source, revmap, trp)
             self.ui.progress(_('manifests'), None)
 
             needfiles = {}
             if self.ui.configbool('server', 'validate', default=False):
                 # validate incoming csets have their manifests
                 for cset in xrange(clstart, clend):
                     mfest = self.changelog.read(self.changelog.node(cset))[0]
                     mfest = self.manifest.readdelta(mfest)
                     # store file nodes we must see
                     for f, n in mfest.iteritems():
                         needfiles.setdefault(f, set()).add(n)
 
             # process the files
             self.ui.status(_("adding file changes\n"))
             pr.step = _('files')
             pr.count = 1
             pr.total = efiles
             source.callback = None
 
             while True:
                 chunkdata = source.filelogheader()
                 if not chunkdata:
                     break
                 f = chunkdata["filename"]
                 self.ui.debug("adding %s revisions\n" % f)
                 pr()
                 fl = self.file(f)
                 o = len(fl)
                 if not fl.addgroup(source, revmap, trp):
                     raise util.Abort(_("received file revlog group is empty"))
                 revisions += len(fl) - o
                 files += 1
                 if f in needfiles:
                     needs = needfiles[f]
                     for new in xrange(o, len(fl)):
                         n = fl.node(new)
                         if n in needs:
                             needs.remove(n)
                     if not needs:
                         del needfiles[f]
             self.ui.progress(_('files'), None)
 
             for f, needs in needfiles.iteritems():
                 fl = self.file(f)
                 for n in needs:
                     try:
                         fl.rev(n)
                     except error.LookupError:
                         raise util.Abort(
                             _('missing file data for %s:%s - run hg verify') %
                             (f, hex(n)))
 
             dh = 0
             if oldheads:
                 heads = cl.heads()
                 dh = len(heads) - len(oldheads)
                 for h in heads:
                     if h not in oldheads and self[h].closesbranch():
                         dh -= 1
             htext = ""
             if dh:
                 htext = _(" (%+d heads)") % dh
 
             self.ui.status(_("added %d changesets"
                              " with %d changes to %d files%s\n")
                              % (changesets, revisions, files, htext))
+            obsolete.clearobscaches(self)
 
             if changesets > 0:
                 p = lambda: cl.writepending() and self.root or ""
                 self.hook('pretxnchangegroup', throw=True,
                           node=hex(cl.node(clstart)), source=srctype,
                           url=url, pending=p)
 
             added = [cl.node(r) for r in xrange(clstart, clend)]
             publishing = self.ui.configbool('phases', 'publish', True)
             if srctype == 'push':
                 # Old server can not push the boundary themself.
                 # New server won't push the boundary if changeset already
                 # existed locally as secrete
                 #
                 # We should not use added here but the list of all change in
                 # the bundle
                 if publishing:
                     phases.advanceboundary(self, phases.public, srccontent)
                 else:
                     phases.advanceboundary(self, phases.draft, srccontent)
                     phases.retractboundary(self, phases.draft, added)
             elif srctype != 'strip':
                 # publishing only alter behavior during push
                 #
                 # strip should not touch boundary at all
                 phases.retractboundary(self, phases.draft, added)
 
             # make changelog see real files again
             cl.finalize(trp)
 
             tr.close()
 
             if changesets > 0:
                 def runhooks():
                     # forcefully update the on-disk branch cache
                     self.ui.debug("updating the branch cache\n")
                     self.updatebranchcache()
                     self.hook("changegroup", node=hex(cl.node(clstart)),
                               source=srctype, url=url)
 
                     for n in added:
                         self.hook("incoming", node=hex(n), source=srctype,
                                   url=url)
                 self._afterlock(runhooks)
 
         finally:
             tr.release()
         # never return 0 here:
         if dh < 0:
             return dh - 1
         else:
             return dh + 1
 
     def stream_in(self, remote, requirements):
         lock = self.lock()
         try:
             fp = remote.stream_out()
             l = fp.readline()
             try:
                 resp = int(l)
             except ValueError:
                 raise error.ResponseError(
                     _('unexpected response from remote server:'), l)
             if resp == 1:
                 raise util.Abort(_('operation forbidden by server'))
             elif resp == 2:
                 raise util.Abort(_('locking the remote repository failed'))
             elif resp != 0:
                 raise util.Abort(_('the server sent an unknown error code'))
             self.ui.status(_('streaming all changes\n'))
             l = fp.readline()
             try:
                 total_files, total_bytes = map(int, l.split(' ', 1))
             except (ValueError, TypeError):
                 raise error.ResponseError(
                     _('unexpected response from remote server:'), l)
             self.ui.status(_('%d files to transfer, %s of data\n') %
                            (total_files, util.bytecount(total_bytes)))
             handled_bytes = 0
             self.ui.progress(_('clone'), 0, total=total_bytes)
             start = time.time()
             for i in xrange(total_files):
                 # XXX doesn't support '\n' or '\r' in filenames
                 l = fp.readline()
                 try:
                     name, size = l.split('\0', 1)
                     size = int(size)
                 except (ValueError, TypeError):
                     raise error.ResponseError(
                         _('unexpected response from remote server:'), l)
                 if self.ui.debugflag:
                     self.ui.debug('adding %s (%s)\n' %
                                   (name, util.bytecount(size)))
                 # for backwards compat, name was partially encoded
                 ofp = self.sopener(store.decodedir(name), 'w')
                 for chunk in util.filechunkiter(fp, limit=size):
                     handled_bytes += len(chunk)
                     self.ui.progress(_('clone'), handled_bytes,
                                      total=total_bytes)
                     ofp.write(chunk)
                 ofp.close()
             elapsed = time.time() - start
             if elapsed <= 0:
                 elapsed = 0.001
             self.ui.progress(_('clone'), None)
             self.ui.status(_('transferred %s in %.1f seconds (%s/sec)\n') %
                            (util.bytecount(total_bytes), elapsed,
                             util.bytecount(total_bytes / elapsed)))
 
             # new requirements = old non-format requirements +
             #                    new format-related
             # requirements from the streamed-in repository
             requirements.update(set(self.requirements) - self.supportedformats)
             self._applyrequirements(requirements)
             self._writerequirements()
 
             self.invalidate()
             return len(self.heads()) + 1
         finally:
             lock.release()
 
     def clone(self, remote, heads=[], stream=False):
         '''clone remote repository.
 
         keyword arguments:
         heads: list of revs to clone (forces use of pull)
         stream: use streaming clone if possible'''
 
         # now, all clients that can request uncompressed clones can
         # read repo formats supported by all servers that can serve
         # them.
 
         # if revlog format changes, client will have to check version
         # and format flags on "stream" capability, and use
         # uncompressed only if compatible.
 
         if not stream:
             # if the server explicitly prefers to stream (for fast LANs)
             stream = remote.capable('stream-preferred')
 
         if stream and not heads:
             # 'stream' means remote revlog format is revlogv1 only
             if remote.capable('stream'):
                 return self.stream_in(remote, set(('revlogv1',)))
             # otherwise, 'streamreqs' contains the remote revlog format
             streamreqs = remote.capable('streamreqs')
             if streamreqs:
                 streamreqs = set(streamreqs.split(','))
                 # if we support it, stream in and adjust our requirements
                 if not streamreqs - self.supportedformats:
                     return self.stream_in(remote, streamreqs)
         return self.pull(remote, heads)
 
     def pushkey(self, namespace, key, old, new):
         self.hook('prepushkey', throw=True, namespace=namespace, key=key,
                   old=old, new=new)
         self.ui.debug('pushing key for "%s:%s"\n' % (namespace, key))
         ret = pushkey.push(self, namespace, key, old, new)
         self.hook('pushkey', namespace=namespace, key=key, old=old, new=new,
                   ret=ret)
         return ret
 
     def listkeys(self, namespace):
         self.hook('prelistkeys', throw=True, namespace=namespace)
         self.ui.debug('listing keys for "%s"\n' % namespace)
         values = pushkey.list(self, namespace)
         self.hook('listkeys', namespace=namespace, values=values)
         return values
 
     def debugwireargs(self, one, two, three=None, four=None, five=None):
         '''used to test argument passing over the wire'''
         return "%s %s %s %s %s" % (one, two, three, four, five)
 
     def savecommitmessage(self, text):
         fp = self.opener('last-message.txt', 'wb')
         try:
             fp.write(text)
         finally:
             fp.close()
         return self.pathto(fp.name[len(self.root)+1:])
 
 # used to avoid circular references so destructors work
 def aftertrans(files):
     renamefiles = [tuple(t) for t in files]
     def a():
         for src, dest in renamefiles:
             try:
                 util.rename(src, dest)
             except OSError: # journal file does not yet exist
                 pass
     return a
 
 def undoname(fn):
     base, name = os.path.split(fn)
     assert name.startswith('journal')
     return os.path.join(base, name.replace('journal', 'undo', 1))
 
 def instance(ui, path, create):
     return localrepository(ui, util.urllocalpath(path), create)
 
 def islocal(path):
     return True

mercurial/obsolete.py

@@ -1,329 +1,397 b''
 # obsolete.py - obsolete markers handling
 #
 # Copyright 2012 Pierre-Yves David <[e-mail]>
 #                Logilab SA        <[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.
 
 """Obsolete markers handling
 
 An obsolete marker maps an old changeset to a list of new
 changesets. If the list of new changesets is empty, the old changeset
 is said to be "killed". Otherwise, the old changeset is being
 "replaced" by the new changesets.
 
 Obsolete markers can be used to record and distribute changeset graph
 transformations performed by history rewriting operations, and help
 building new tools to reconciliate conflicting rewriting actions. To
 facilitate conflicts resolution, markers include various annotations
 besides old and news changeset identifiers, such as creation date or
 author name.
 
 
 Format
 ------
 
 Markers are stored in an append-only file stored in
 '.hg/store/obsstore'.
 
 The file starts with a version header:
 
 - 1 unsigned byte: version number, starting at zero.
 
 
 The header is followed by the markers. Each marker is made of:
 
 - 1 unsigned byte: number of new changesets "R", could be zero.
 
 - 1 unsigned 32-bits integer: metadata size "M" in bytes.
 
 - 1 byte: a bit field. It is reserved for flags used in obsolete
   markers common operations, to avoid repeated decoding of metadata
   entries.
 
 - 20 bytes: obsoleted changeset identifier.
 
 - N*20 bytes: new changesets identifiers.
 
 - M bytes: metadata as a sequence of nul-terminated strings. Each
   string contains a key and a value, separated by a color ':', without
   additional encoding. Keys cannot contain '\0' or ':' and values
   cannot contain '\0'.
 """
 import struct
 import util, base85
 from i18n import _
 
 _pack = struct.pack
 _unpack = struct.unpack
 
 _SEEK_END = 2 # os.SEEK_END was introduced in Python 2.5
 
 # the obsolete feature is not mature enough to be enabled by default.
 # you have to rely on third party extension extension to enable this.
 _enabled = False
 
 # data used for parsing and writing
 _fmversion = 0
 _fmfixed   = '>BIB20s'
 _fmnode = '20s'
 _fmfsize = struct.calcsize(_fmfixed)
 _fnodesize = struct.calcsize(_fmnode)
 
 def _readmarkers(data):
     """Read and enumerate markers from raw data"""
     off = 0
     diskversion = _unpack('>B', data[off:off + 1])[0]
     off += 1
     if diskversion != _fmversion:
         raise util.Abort(_('parsing obsolete marker: unknown version %r')
                          % diskversion)
 
     # Loop on markers
     l = len(data)
     while off + _fmfsize <= l:
         # read fixed part
         cur = data[off:off + _fmfsize]
         off += _fmfsize
         nbsuc, mdsize, flags, pre = _unpack(_fmfixed, cur)
         # read replacement
         sucs = ()
         if nbsuc:
             s = (_fnodesize * nbsuc)
             cur = data[off:off + s]
             sucs = _unpack(_fmnode * nbsuc, cur)
             off += s
         # read metadata
         # (metadata will be decoded on demand)
         metadata = data[off:off + mdsize]
         if len(metadata) != mdsize:
             raise util.Abort(_('parsing obsolete marker: metadata is too '
                                'short, %d bytes expected, got %d')
                              % (mdsize, len(metadata)))
         off += mdsize
         yield (pre, sucs, flags, metadata)
 
 def encodemeta(meta):
     """Return encoded metadata string to string mapping.
 
     Assume no ':' in key and no '\0' in both key and value."""
     for key, value in meta.iteritems():
         if ':' in key or '\0' in key:
             raise ValueError("':' and '\0' are forbidden in metadata key'")
         if '\0' in value:
             raise ValueError("':' are forbidden in metadata value'")
     return '\0'.join(['%s:%s' % (k, meta[k]) for k in sorted(meta)])
 
 def decodemeta(data):
     """Return string to string dictionary from encoded version."""
     d = {}
     for l in data.split('\0'):
         if l:
             key, value = l.split(':')
             d[key] = value
     return d
 
 class marker(object):
     """Wrap obsolete marker raw data"""
 
     def __init__(self, repo, data):
         # the repo argument will be used to create changectx in later version
         self._repo = repo
         self._data = data
         self._decodedmeta = None
 
     def precnode(self):
         """Precursor changeset node identifier"""
         return self._data[0]
 
     def succnodes(self):
         """List of successor changesets node identifiers"""
         return self._data[1]
 
     def metadata(self):
         """Decoded metadata dictionary"""
         if self._decodedmeta is None:
             self._decodedmeta = decodemeta(self._data[3])
         return self._decodedmeta
 
     def date(self):
         """Creation date as (unixtime, offset)"""
         parts = self.metadata()['date'].split(' ')
         return (float(parts[0]), int(parts[1]))
 
 class obsstore(object):
     """Store obsolete markers
 
     Markers can be accessed with two mappings:
     - precursors: old -> set(new)
     - successors: new -> set(old)
     """
 
     def __init__(self, sopener):
+        # caches for various obsolescence related cache
+        self.caches = {}
         self._all = []
         # new markers to serialize
         self.precursors = {}
         self.successors = {}
         self.sopener = sopener
         data = sopener.tryread('obsstore')
         if data:
             self._load(_readmarkers(data))
 
     def __iter__(self):
         return iter(self._all)
 
     def __nonzero__(self):
         return bool(self._all)
 
     def create(self, transaction, prec, succs=(), flag=0, metadata=None):
         """obsolete: add a new obsolete marker
 
         * ensuring it is hashable
         * check mandatory metadata
         * encode metadata
         """
         if metadata is None:
             metadata = {}
         if len(prec) != 20:
             raise ValueError(prec)
         for succ in succs:
             if len(succ) != 20:
                 raise ValueError(succ)
         marker = (str(prec), tuple(succs), int(flag), encodemeta(metadata))
         self.add(transaction, [marker])
 
     def add(self, transaction, markers):
         """Add new markers to the store
 
         Take care of filtering duplicate.
         Return the number of new marker."""
         if not _enabled:
             raise util.Abort('obsolete feature is not enabled on this repo')
         new = [m for m in markers if m not in self._all]
         if new:
             f = self.sopener('obsstore', 'ab')
             try:
                 # Whether the file's current position is at the begin or at
                 # the end after opening a file for appending is implementation
                 # defined. So we must seek to the end before calling tell(),
                 # or we may get a zero offset for non-zero sized files on
                 # some platforms (issue3543).
                 f.seek(0, _SEEK_END)
                 offset = f.tell()
                 transaction.add('obsstore', offset)
                 # offset == 0: new file - add the version header
                 for bytes in _encodemarkers(new, offset == 0):
                     f.write(bytes)
             finally:
                 # XXX: f.close() == filecache invalidation == obsstore rebuilt.
                 # call 'filecacheentry.refresh()'  here
                 f.close()
             self._load(new)
+            # new marker *may* have changed several set. invalidate the cache.
+            self.caches.clear()
         return len(new)
 
     def mergemarkers(self, transation, data):
         markers = _readmarkers(data)
         self.add(transation, markers)
 
     def _load(self, markers):
         for mark in markers:
             self._all.append(mark)
             pre, sucs = mark[:2]
             self.precursors.setdefault(pre, set()).add(mark)
             for suc in sucs:
                 self.successors.setdefault(suc, set()).add(mark)
 
 def _encodemarkers(markers, addheader=False):
     # Kept separate from flushmarkers(), it will be reused for
     # markers exchange.
     if addheader:
         yield _pack('>B', _fmversion)
     for marker in markers:
         yield _encodeonemarker(marker)
 
 
 def _encodeonemarker(marker):
     pre, sucs, flags, metadata = marker
     nbsuc = len(sucs)
     format = _fmfixed + (_fmnode * nbsuc)
     data = [nbsuc, len(metadata), flags, pre]
     data.extend(sucs)
     return _pack(format, *data) + metadata
 
 # arbitrary picked to fit into 8K limit from HTTP server
 # you have to take in account:
 # - the version header
 # - the base85 encoding
 _maxpayload = 5300
 
 def listmarkers(repo):
     """List markers over pushkey"""
     if not repo.obsstore:
         return {}
     keys = {}
     parts = []
     currentlen = _maxpayload * 2  # ensure we create a new part
     for marker in  repo.obsstore:
         nextdata = _encodeonemarker(marker)
         if (len(nextdata) + currentlen > _maxpayload):
             currentpart = []
             currentlen = 0
             parts.append(currentpart)
         currentpart.append(nextdata)
         currentlen += len(nextdata)
     for idx, part in enumerate(reversed(parts)):
         data = ''.join([_pack('>B', _fmversion)] + part)
         keys['dump%i' % idx] = base85.b85encode(data)
     return keys
 
 def pushmarker(repo, key, old, new):
     """Push markers over pushkey"""
     if not key.startswith('dump'):
         repo.ui.warn(_('unknown key: %r') % key)
         return 0
     if old:
         repo.ui.warn(_('unexpected old value') % key)
         return 0
     data = base85.b85decode(new)
     lock = repo.lock()
     try:
         tr = repo.transaction('pushkey: obsolete markers')
         try:
             repo.obsstore.mergemarkers(tr, data)
             tr.close()
             return 1
         finally:
             tr.release()
     finally:
         lock.release()
 
 def allmarkers(repo):
     """all obsolete markers known in a repository"""
     for markerdata in repo.obsstore:
         yield marker(repo, markerdata)
 
 def precursormarkers(ctx):
     """obsolete marker making this changeset obsolete"""
     for data in ctx._repo.obsstore.precursors.get(ctx.node(), ()):
         yield marker(ctx._repo, data)
 
 def successormarkers(ctx):
     """obsolete marker marking this changeset as a successors"""
     for data in ctx._repo.obsstore.successors.get(ctx.node(), ()):
         yield marker(ctx._repo, data)
 
 def anysuccessors(obsstore, node):
     """Yield every successor of <node>
 
     This is a linear yield unsuitable to detect split changesets."""
     remaining = set([node])
     seen = set(remaining)
     while remaining:
         current = remaining.pop()
         yield current
         for mark in obsstore.precursors.get(current, ()):
             for suc in mark[1]:
                 if suc not in seen:
                     seen.add(suc)
                     remaining.add(suc)
+
+# mapping of 'set-name' -> <function to computer this set>
+cachefuncs = {}
+def cachefor(name):
+    """Decorator to register a function as computing the cache for a set"""
+    def decorator(func):
+        assert name not in cachefuncs
+        cachefuncs[name] = func
+        return func
+    return decorator
+
+def getobscache(repo, name):
+    """Return the set of revision that belong to the <name> set
+
+    Such access may compute the set and cache it for future use"""
+    if not repo.obsstore:
+        return ()
+    if name not in repo.obsstore.caches:
+        repo.obsstore.caches[name] = cachefuncs[name](repo)
+    return repo.obsstore.caches[name]
+
+# To be simple we need to invalidate obsolescence cache when:
+#
+# - new changeset is added:
+# - public phase is changed
+# - obsolescence marker are added
+# - strip is used a repo
+def clearobscaches(repo):
+    """Remove all obsolescence related cache from a repo
+
+    This remove all cache in obsstore is the obsstore already exist on the
+    repo.
+
+    (We could be smarter here given the exact event that trigger the cache
+    clearing)"""
+    # only clear cache is there is obsstore data in this repo
+    if 'obsstore' in repo._filecache:
+        repo.obsstore.caches.clear()
+
+@cachefor('obsolete')
+def _computeobsoleteset(repo):
+    """the set of obsolete revisions"""
+    obs = set()
+    nm = repo.changelog.nodemap
+    for prec in repo.obsstore.precursors:
+        rev = nm.get(prec)
+        if rev is not None:
+            obs.add(rev)
+    return set(repo.revs('%ld - public()', obs))
+
+@cachefor('unstable')
+def _computeunstableset(repo):
+    """the set of non obsolete revisions with obsolete parents"""
+    return set(repo.revs('(obsolete()::) - obsolete()'))
+
+@cachefor('suspended')
+def _computesuspendedset(repo):
+    """the set of obsolete parents with non obsolete descendants"""
+    return set(repo.revs('obsolete() and obsolete()::unstable()'))
+
+@cachefor('extinct')
+def _computeextinctset(repo):
+    """the set of obsolete parents without non obsolete descendants"""
+    return set(repo.revs('obsolete() - obsolete()::unstable()'))

mercurial/phases.py

@@ -1,387 +1,390 b''
 """ Mercurial phases support code
 
     ---
 
     Copyright 2011 Pierre-Yves David <[e-mail]>
                    Logilab SA        <[e-mail]>
                    Augie Fackler     <[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.
 
     ---
 
 This module implements most phase logic in mercurial.
 
 
 Basic Concept
 =============
 
 A 'changeset phase' is an indicator that tells us how a changeset is
 manipulated and communicated. The details of each phase is described
 below, here we describe the properties they have in common.
 
 Like bookmarks, phases are not stored in history and thus are not
 permanent and leave no audit trail.
 
 First, no changeset can be in two phases at once. Phases are ordered,
 so they can be considered from lowest to highest. The default, lowest
 phase is 'public' - this is the normal phase of existing changesets. A
 child changeset can not be in a lower phase than its parents.
 
 These phases share a hierarchy of traits:
 
             immutable shared
     public:     X        X
     draft:               X
     secret:
 
 Local commits are draft by default.
 
 Phase Movement and Exchange
 ===========================
 
 Phase data is exchanged by pushkey on pull and push. Some servers have
 a publish option set, we call such a server a "publishing server".
 Pushing a draft changeset to a publishing server changes the phase to
 public.
 
 A small list of fact/rules define the exchange of phase:
 
 * old client never changes server states
 * pull never changes server states
 * publish and old server changesets are seen as public by client
 * any secret changeset seen in another repository is lowered to at
   least draft
 
 Here is the final table summing up the 49 possible use cases of phase
 exchange:
 
                            server
                   old     publish      non-publish
                  N   X    N   D   P    N   D   P
     old client
     pull
      N           -   X/X  -   X/D X/P  -   X/D X/P
      X           -   X/X  -   X/D X/P  -   X/D X/P
     push
      X           X/X X/X  X/P X/P X/P  X/D X/D X/P
     new client
     pull
      N           -   P/X  -   P/D P/P  -   D/D P/P
      D           -   P/X  -   P/D P/P  -   D/D P/P
      P           -   P/X  -   P/D P/P  -   P/D P/P
     push
      D           P/X P/X  P/P P/P P/P  D/D D/D P/P
      P           P/X P/X  P/P P/P P/P  P/P P/P P/P
 
 Legend:
 
     A/B = final state on client / state on server
 
     * N = new/not present,
     * P = public,
     * D = draft,
     * X = not tracked (i.e., the old client or server has no internal
           way of recording the phase.)
 
     passive = only pushes
 
 
     A cell here can be read like this:
 
     "When a new client pushes a draft changeset (D) to a publishing
     server where it's not present (N), it's marked public on both
     sides (P/P)."
 
 Note: old client behave as a publishing server with draft only content
 - other people see it as public
 - content is pushed as draft
 
 """
 
 import errno
 from node import nullid, nullrev, bin, hex, short
 from i18n import _
 import util
+import obsolete
 
 allphases = public, draft, secret = range(3)
 trackedphases = allphases[1:]
 phasenames = ['public', 'draft', 'secret']
 
 def _filterunknown(ui, changelog, phaseroots):
     """remove unknown nodes from the phase boundary
 
     Nothing is lost as unknown nodes only hold data for their descendants.
     """
     updated = False
     nodemap = changelog.nodemap # to filter unknown nodes
     for phase, nodes in enumerate(phaseroots):
         missing = [node for node in nodes if node not in nodemap]
         if missing:
             for mnode in missing:
                 ui.debug(
                     'removing unknown node %s from %i-phase boundary\n'
                     % (short(mnode), phase))
             nodes.symmetric_difference_update(missing)
             updated = True
     return updated
 
 def _readroots(repo, phasedefaults=None):
     """Read phase roots from disk
 
     phasedefaults is a list of fn(repo, roots) callable, which are
     executed if the phase roots file does not exist. When phases are
     being initialized on an existing repository, this could be used to
     set selected changesets phase to something else than public.
 
     Return (roots, dirty) where dirty is true if roots differ from
     what is being stored.
     """
     dirty = False
     roots = [set() for i in allphases]
     try:
         f = repo.sopener('phaseroots')
         try:
             for line in f:
                 phase, nh = line.split()
                 roots[int(phase)].add(bin(nh))
         finally:
             f.close()
     except IOError, inst:
         if inst.errno != errno.ENOENT:
             raise
         if phasedefaults:
             for f in phasedefaults:
                 roots = f(repo, roots)
         dirty = True
     if _filterunknown(repo.ui, repo.changelog, roots):
         dirty = True
     return roots, dirty
 
 class phasecache(object):
     def __init__(self, repo, phasedefaults, _load=True):
         if _load:
             # Cheap trick to allow shallow-copy without copy module
             self.phaseroots, self.dirty = _readroots(repo, phasedefaults)
             self.opener = repo.sopener
             self._phaserevs = None
 
     def copy(self):
         # Shallow copy meant to ensure isolation in
         # advance/retractboundary(), nothing more.
         ph = phasecache(None, None, _load=False)
         ph.phaseroots = self.phaseroots[:]
         ph.dirty = self.dirty
         ph.opener = self.opener
         ph._phaserevs = self._phaserevs
         return ph
 
     def replace(self, phcache):
         for a in 'phaseroots dirty opener _phaserevs'.split():
             setattr(self, a, getattr(phcache, a))
 
     def getphaserevs(self, repo, rebuild=False):
         if rebuild or self._phaserevs is None:
             revs = [public] * len(repo.changelog)
             for phase in trackedphases:
                 roots = map(repo.changelog.rev, self.phaseroots[phase])
                 if roots:
                     for rev in roots:
                         revs[rev] = phase
                     for rev in repo.changelog.descendants(roots):
                         revs[rev] = phase
             self._phaserevs = revs
         return self._phaserevs
 
     def phase(self, repo, rev):
         # We need a repo argument here to be able to build _phaserevs
         # if necessary. The repository instance is not stored in
         # phasecache to avoid reference cycles. The changelog instance
         # is not stored because it is a filecache() property and can
         # be replaced without us being notified.
         if rev == nullrev:
             return public
         if self._phaserevs is None or rev >= len(self._phaserevs):
             self._phaserevs = self.getphaserevs(repo, rebuild=True)
         return self._phaserevs[rev]
 
     def write(self):
         if not self.dirty:
             return
         f = self.opener('phaseroots', 'w', atomictemp=True)
         try:
             for phase, roots in enumerate(self.phaseroots):
                 for h in roots:
                     f.write('%i %s\n' % (phase, hex(h)))
         finally:
             f.close()
         self.dirty = False
 
     def _updateroots(self, phase, newroots):
         self.phaseroots[phase] = newroots
         self._phaserevs = None
         self.dirty = True
 
     def advanceboundary(self, repo, targetphase, nodes):
         # Be careful to preserve shallow-copied values: do not update
         # phaseroots values, replace them.
 
         delroots = [] # set of root deleted by this path
         for phase in xrange(targetphase + 1, len(allphases)):
             # filter nodes that are not in a compatible phase already
             nodes = [n for n in nodes
                      if self.phase(repo, repo[n].rev()) >= phase]
             if not nodes:
                 break # no roots to move anymore
             olds = self.phaseroots[phase]
             roots = set(ctx.node() for ctx in repo.set(
                     'roots((%ln::) - (%ln::%ln))', olds, olds, nodes))
             if olds != roots:
                 self._updateroots(phase, roots)
                 # some roots may need to be declared for lower phases
                 delroots.extend(olds - roots)
             # declare deleted root in the target phase
             if targetphase != 0:
                 self.retractboundary(repo, targetphase, delroots)
+        obsolete.clearobscaches(repo)
 
     def retractboundary(self, repo, targetphase, nodes):
         # Be careful to preserve shallow-copied values: do not update
         # phaseroots values, replace them.
 
         currentroots = self.phaseroots[targetphase]
         newroots = [n for n in nodes
                     if self.phase(repo, repo[n].rev()) < targetphase]
         if newroots:
             if nullid in newroots:
                 raise util.Abort(_('cannot change null revision phase'))
             currentroots = currentroots.copy()
             currentroots.update(newroots)
             ctxs = repo.set('roots(%ln::)', currentroots)
             currentroots.intersection_update(ctx.node() for ctx in ctxs)
             self._updateroots(targetphase, currentroots)
+        obsolete.clearobscaches(repo)
 
 def advanceboundary(repo, targetphase, nodes):
     """Add nodes to a phase changing other nodes phases if necessary.
 
     This function move boundary *forward* this means that all nodes
     are set in the target phase or kept in a *lower* phase.
 
     Simplify boundary to contains phase roots only."""
     phcache = repo._phasecache.copy()
     phcache.advanceboundary(repo, targetphase, nodes)
     repo._phasecache.replace(phcache)
 
 def retractboundary(repo, targetphase, nodes):
     """Set nodes back to a phase changing other nodes phases if
     necessary.
 
     This function move boundary *backward* this means that all nodes
     are set in the target phase or kept in a *higher* phase.
 
     Simplify boundary to contains phase roots only."""
     phcache = repo._phasecache.copy()
     phcache.retractboundary(repo, targetphase, nodes)
     repo._phasecache.replace(phcache)
 
 def listphases(repo):
     """List phases root for serialization over pushkey"""
     keys = {}
     value = '%i' % draft
     for root in repo._phasecache.phaseroots[draft]:
         keys[hex(root)] = value
 
     if repo.ui.configbool('phases', 'publish', True):
         # Add an extra data to let remote know we are a publishing
         # repo. Publishing repo can't just pretend they are old repo.
         # When pushing to a publishing repo, the client still need to
         # push phase boundary
         #
         # Push do not only push changeset. It also push phase data.
         # New phase data may apply to common changeset which won't be
         # push (as they are common). Here is a very simple example:
         #
         # 1) repo A push changeset X as draft to repo B
         # 2) repo B make changeset X public
         # 3) repo B push to repo A. X is not pushed but the data that
         #    X as now public should
         #
         # The server can't handle it on it's own as it has no idea of
         # client phase data.
         keys['publishing'] = 'True'
     return keys
 
 def pushphase(repo, nhex, oldphasestr, newphasestr):
     """List phases root for serialisation over pushkey"""
     lock = repo.lock()
     try:
         currentphase = repo[nhex].phase()
         newphase = abs(int(newphasestr)) # let's avoid negative index surprise
         oldphase = abs(int(oldphasestr)) # let's avoid negative index surprise
         if currentphase == oldphase and newphase < oldphase:
             advanceboundary(repo, newphase, [bin(nhex)])
             return 1
         elif currentphase == newphase:
             # raced, but got correct result
             return 1
         else:
             return 0
     finally:
         lock.release()
 
 def analyzeremotephases(repo, subset, roots):
     """Compute phases heads and root in a subset of node from root dict
 
     * subset is heads of the subset
     * roots is {<nodeid> => phase} mapping. key and value are string.
 
     Accept unknown element input
     """
     # build list from dictionary
     draftroots = []
     nodemap = repo.changelog.nodemap # to filter unknown nodes
     for nhex, phase in roots.iteritems():
         if nhex == 'publishing': # ignore data related to publish option
             continue
         node = bin(nhex)
         phase = int(phase)
         if phase == 0:
             if node != nullid:
                 repo.ui.warn(_('ignoring inconsistent public root'
                                ' from remote: %s\n') % nhex)
         elif phase == 1:
             if node in nodemap:
                 draftroots.append(node)
         else:
             repo.ui.warn(_('ignoring unexpected root from remote: %i %s\n')
                          % (phase, nhex))
     # compute heads
     publicheads = newheads(repo, subset, draftroots)
     return publicheads, draftroots
 
 def newheads(repo, heads, roots):
     """compute new head of a subset minus another
 
     * `heads`: define the first subset
     * `roots`: define the second we subtract from the first"""
     revset = repo.set('heads((%ln + parents(%ln)) - (%ln::%ln))',
                       heads, roots, roots, heads)
     return [c.node() for c in revset]
 
 
 def newcommitphase(ui):
     """helper to get the target phase of new commit
 
     Handle all possible values for the phases.new-commit options.
 
     """
     v = ui.config('phases', 'new-commit', draft)
     try:
         return phasenames.index(v)
     except ValueError:
         try:
             return int(v)
         except ValueError:
             msg = _("phases.new-commit: not a valid phase name ('%s')")
             raise error.ConfigError(msg % v)
 

mercurial/revset.py

@@ -1,1865 +1,1867 b''
 # revset.py - revision set queries for mercurial
 #
 # Copyright 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 re
 import parser, util, error, discovery, hbisect, phases
 import node
 import bookmarks as bookmarksmod
 import match as matchmod
 from i18n import _
 import encoding
+import obsolete as obsmod
 
 def _revancestors(repo, revs, followfirst):
     """Like revlog.ancestors(), but supports followfirst."""
     cut = followfirst and 1 or None
     cl = repo.changelog
     visit = util.deque(revs)
     seen = set([node.nullrev])
     while visit:
         for parent in cl.parentrevs(visit.popleft())[:cut]:
             if parent not in seen:
                 visit.append(parent)
                 seen.add(parent)
                 yield parent
 
 def _revdescendants(repo, revs, followfirst):
     """Like revlog.descendants() but supports followfirst."""
     cut = followfirst and 1 or None
     cl = repo.changelog
     first = min(revs)
     nullrev = node.nullrev
     if first == nullrev:
         # Are there nodes with a null first parent and a non-null
         # second one? Maybe. Do we care? Probably not.
         for i in cl:
             yield i
         return
 
     seen = set(revs)
     for i in xrange(first + 1, len(cl)):
         for x in cl.parentrevs(i)[:cut]:
             if x != nullrev and x in seen:
                 seen.add(i)
                 yield i
                 break
 
 def _revsbetween(repo, roots, heads):
     """Return all paths between roots and heads, inclusive of both endpoint
     sets."""
     if not roots:
         return []
     parentrevs = repo.changelog.parentrevs
     visit = heads[:]
     reachable = set()
     seen = {}
     minroot = min(roots)
     roots = set(roots)
     # open-code the post-order traversal due to the tiny size of
     # sys.getrecursionlimit()
     while visit:
         rev = visit.pop()
         if rev in roots:
             reachable.add(rev)
         parents = parentrevs(rev)
         seen[rev] = parents
         for parent in parents:
             if parent >= minroot and parent not in seen:
                 visit.append(parent)
     if not reachable:
         return []
     for rev in sorted(seen):
         for parent in seen[rev]:
             if parent in reachable:
                 reachable.add(rev)
     return sorted(reachable)
 
 elements = {
     "(": (20, ("group", 1, ")"), ("func", 1, ")")),
     "~": (18, None, ("ancestor", 18)),
     "^": (18, None, ("parent", 18), ("parentpost", 18)),
     "-": (5, ("negate", 19), ("minus", 5)),
     "::": (17, ("dagrangepre", 17), ("dagrange", 17),
            ("dagrangepost", 17)),
     "..": (17, ("dagrangepre", 17), ("dagrange", 17),
            ("dagrangepost", 17)),
     ":": (15, ("rangepre", 15), ("range", 15), ("rangepost", 15)),
     "not": (10, ("not", 10)),
     "!": (10, ("not", 10)),
     "and": (5, None, ("and", 5)),
     "&": (5, None, ("and", 5)),
     "or": (4, None, ("or", 4)),
     "|": (4, None, ("or", 4)),
     "+": (4, None, ("or", 4)),
     ",": (2, None, ("list", 2)),
     ")": (0, None, None),
     "symbol": (0, ("symbol",), None),
     "string": (0, ("string",), None),
     "end": (0, None, None),
 }
 
 keywords = set(['and', 'or', 'not'])
 
 def tokenize(program):
     pos, l = 0, len(program)
     while pos < l:
         c = program[pos]
         if c.isspace(): # skip inter-token whitespace
             pass
         elif c == ':' and program[pos:pos + 2] == '::': # look ahead carefully
             yield ('::', None, pos)
             pos += 1 # skip ahead
         elif c == '.' and program[pos:pos + 2] == '..': # look ahead carefully
             yield ('..', None, pos)
             pos += 1 # skip ahead
         elif c in "():,-|&+!~^": # handle simple operators
             yield (c, None, pos)
         elif (c in '"\'' or c == 'r' and
               program[pos:pos + 2] in ("r'", 'r"')): # handle quoted strings
             if c == 'r':
                 pos += 1
                 c = program[pos]
                 decode = lambda x: x
             else:
                 decode = lambda x: x.decode('string-escape')
             pos += 1
             s = pos
             while pos < l: # find closing quote
                 d = program[pos]
                 if d == '\\': # skip over escaped characters
                     pos += 2
                     continue
                 if d == c:
                     yield ('string', decode(program[s:pos]), s)
                     break
                 pos += 1
             else:
                 raise error.ParseError(_("unterminated string"), s)
         # gather up a symbol/keyword
         elif c.isalnum() or c in '._' or ord(c) > 127:
             s = pos
             pos += 1
             while pos < l: # find end of symbol
                 d = program[pos]
                 if not (d.isalnum() or d in "._/" or ord(d) > 127):
                     break
                 if d == '.' and program[pos - 1] == '.': # special case for ..
                     pos -= 1
                     break
                 pos += 1
             sym = program[s:pos]
             if sym in keywords: # operator keywords
                 yield (sym, None, s)
             else:
                 yield ('symbol', sym, s)
             pos -= 1
         else:
             raise error.ParseError(_("syntax error"), pos)
         pos += 1
     yield ('end', None, pos)
 
 # helpers
 
 def getstring(x, err):
     if x and (x[0] == 'string' or x[0] == 'symbol'):
         return x[1]
     raise error.ParseError(err)
 
 def getlist(x):
     if not x:
         return []
     if x[0] == 'list':
         return getlist(x[1]) + [x[2]]
     return [x]
 
 def getargs(x, min, max, err):
     l = getlist(x)
     if len(l) < min or (max >= 0 and len(l) > max):
         raise error.ParseError(err)
     return l
 
 def getset(repo, subset, x):
     if not x:
         raise error.ParseError(_("missing argument"))
     return methods[x[0]](repo, subset, *x[1:])
 
 def _getrevsource(repo, r):
     extra = repo[r].extra()
     for label in ('source', 'transplant_source', 'rebase_source'):
         if label in extra:
             try:
                 return repo[extra[label]].rev()
             except error.RepoLookupError:
                 pass
     return None
 
 # operator methods
 
 def stringset(repo, subset, x):
     x = repo[x].rev()
     if x == -1 and len(subset) == len(repo):
         return [-1]
     if len(subset) == len(repo) or x in subset:
         return [x]
     return []
 
 def symbolset(repo, subset, x):
     if x in symbols:
         raise error.ParseError(_("can't use %s here") % x)
     return stringset(repo, subset, x)
 
 def rangeset(repo, subset, x, y):
     m = getset(repo, subset, x)
     if not m:
         m = getset(repo, range(len(repo)), x)
 
     n = getset(repo, subset, y)
     if not n:
         n = getset(repo, range(len(repo)), y)
 
     if not m or not n:
         return []
     m, n = m[0], n[-1]
 
     if m < n:
         r = range(m, n + 1)
     else:
         r = range(m, n - 1, -1)
     s = set(subset)
     return [x for x in r if x in s]
 
 def dagrange(repo, subset, x, y):
     if subset:
         r = range(len(repo))
         xs = _revsbetween(repo, getset(repo, r, x), getset(repo, r, y))
         s = set(subset)
         return [r for r in xs if r in s]
     return []
 
 def andset(repo, subset, x, y):
     return getset(repo, getset(repo, subset, x), y)
 
 def orset(repo, subset, x, y):
     xl = getset(repo, subset, x)
     s = set(xl)
     yl = getset(repo, [r for r in subset if r not in s], y)
     return xl + yl
 
 def notset(repo, subset, x):
     s = set(getset(repo, subset, x))
     return [r for r in subset if r not in s]
 
 def listset(repo, subset, a, b):
     raise error.ParseError(_("can't use a list in this context"))
 
 def func(repo, subset, a, b):
     if a[0] == 'symbol' and a[1] in symbols:
         return symbols[a[1]](repo, subset, b)
     raise error.ParseError(_("not a function: %s") % a[1])
 
 # functions
 
 def adds(repo, subset, x):
     """``adds(pattern)``
     Changesets that add a file matching pattern.
     """
     # i18n: "adds" is a keyword
     pat = getstring(x, _("adds requires a pattern"))
     return checkstatus(repo, subset, pat, 1)
 
 def ancestor(repo, subset, x):
     """``ancestor(single, single)``
     Greatest common ancestor of the two changesets.
     """
     # i18n: "ancestor" is a keyword
     l = getargs(x, 2, 2, _("ancestor requires two arguments"))
     r = range(len(repo))
     a = getset(repo, r, l[0])
     b = getset(repo, r, l[1])
     if len(a) != 1 or len(b) != 1:
         # i18n: "ancestor" is a keyword
         raise error.ParseError(_("ancestor arguments must be single revisions"))
     an = [repo[a[0]].ancestor(repo[b[0]]).rev()]
 
     return [r for r in an if r in subset]
 
 def _ancestors(repo, subset, x, followfirst=False):
     args = getset(repo, range(len(repo)), x)
     if not args:
         return []
     s = set(_revancestors(repo, args, followfirst)) | set(args)
     return [r for r in subset if r in s]
 
 def ancestors(repo, subset, x):
     """``ancestors(set)``
     Changesets that are ancestors of a changeset in set.
     """
     return _ancestors(repo, subset, x)
 
 def _firstancestors(repo, subset, x):
     # ``_firstancestors(set)``
     # Like ``ancestors(set)`` but follows only the first parents.
     return _ancestors(repo, subset, x, followfirst=True)
 
 def ancestorspec(repo, subset, x, n):
     """``set~n``
     Changesets that are the Nth ancestor (first parents only) of a changeset
     in set.
     """
     try:
         n = int(n[1])
     except (TypeError, ValueError):
         raise error.ParseError(_("~ expects a number"))
     ps = set()
     cl = repo.changelog
     for r in getset(repo, subset, x):
         for i in range(n):
             r = cl.parentrevs(r)[0]
         ps.add(r)
     return [r for r in subset if r in ps]
 
 def author(repo, subset, x):
     """``author(string)``
     Alias for ``user(string)``.
     """
     # i18n: "author" is a keyword
     n = encoding.lower(getstring(x, _("author requires a string")))
     kind, pattern, matcher = _substringmatcher(n)
     return [r for r in subset if matcher(encoding.lower(repo[r].user()))]
 
 def bisect(repo, subset, x):
     """``bisect(string)``
     Changesets marked in the specified bisect status:
 
     - ``good``, ``bad``, ``skip``: csets explicitly marked as good/bad/skip
     - ``goods``, ``bads``      : csets topologically good/bad
     - ``range``              : csets taking part in the bisection
     - ``pruned``             : csets that are goods, bads or skipped
     - ``untested``           : csets whose fate is yet unknown
     - ``ignored``            : csets ignored due to DAG topology
     - ``current``            : the cset currently being bisected
     """
     # i18n: "bisect" is a keyword
     status = getstring(x, _("bisect requires a string")).lower()
     state = set(hbisect.get(repo, status))
     return [r for r in subset if r in state]
 
 # Backward-compatibility
 # - no help entry so that we do not advertise it any more
 def bisected(repo, subset, x):
     return bisect(repo, subset, x)
 
 def bookmark(repo, subset, x):
     """``bookmark([name])``
     The named bookmark or all bookmarks.
 
     If `name` starts with `re:`, the remainder of the name is treated as
     a regular expression. To match a bookmark that actually starts with `re:`,
     use the prefix `literal:`.
     """
     # i18n: "bookmark" is a keyword
     args = getargs(x, 0, 1, _('bookmark takes one or no arguments'))
     if args:
         bm = getstring(args[0],
                        # i18n: "bookmark" is a keyword
                        _('the argument to bookmark must be a string'))
         kind, pattern, matcher = _stringmatcher(bm)
         if kind == 'literal':
             bmrev = bookmarksmod.listbookmarks(repo).get(bm, None)
             if not bmrev:
                 raise util.Abort(_("bookmark '%s' does not exist") % bm)
             bmrev = repo[bmrev].rev()
             return [r for r in subset if r == bmrev]
         else:
             matchrevs = set()
             for name, bmrev in bookmarksmod.listbookmarks(repo).iteritems():
                 if matcher(name):
                     matchrevs.add(bmrev)
             if not matchrevs:
                 raise util.Abort(_("no bookmarks exist that match '%s'")
                                  % pattern)
             bmrevs = set()
             for bmrev in matchrevs:
                 bmrevs.add(repo[bmrev].rev())
             return [r for r in subset if r in bmrevs]
 
     bms = set([repo[r].rev()
                for r in bookmarksmod.listbookmarks(repo).values()])
     return [r for r in subset if r in bms]
 
 def branch(repo, subset, x):
     """``branch(string or set)``
     All changesets belonging to the given branch or the branches of the given
     changesets.
 
     If `string` starts with `re:`, the remainder of the name is treated as
     a regular expression. To match a branch that actually starts with `re:`,
     use the prefix `literal:`.
     """
     try:
         b = getstring(x, '')
     except error.ParseError:
         # not a string, but another revspec, e.g. tip()
         pass
     else:
         kind, pattern, matcher = _stringmatcher(b)
         if kind == 'literal':
             # note: falls through to the revspec case if no branch with
             # this name exists
             if pattern in repo.branchmap():
                 return [r for r in subset if matcher(repo[r].branch())]
         else:
             return [r for r in subset if matcher(repo[r].branch())]
 
     s = getset(repo, range(len(repo)), x)
     b = set()
     for r in s:
         b.add(repo[r].branch())
     s = set(s)
     return [r for r in subset if r in s or repo[r].branch() in b]
 
 def checkstatus(repo, subset, pat, field):
     m = None
     s = []
     hasset = matchmod.patkind(pat) == 'set'
     fname = None
     for r in subset:
         c = repo[r]
         if not m or hasset:
             m = matchmod.match(repo.root, repo.getcwd(), [pat], ctx=c)
             if not m.anypats() and len(m.files()) == 1:
                 fname = m.files()[0]
         if fname is not None:
             if fname not in c.files():
                 continue
         else:
             for f in c.files():
                 if m(f):
                     break
             else:
                 continue
         files = repo.status(c.p1().node(), c.node())[field]
         if fname is not None:
             if fname in files:
                 s.append(r)
         else:
             for f in files:
                 if m(f):
                     s.append(r)
                     break
     return s
 
 def _children(repo, narrow, parentset):
     cs = set()
     pr = repo.changelog.parentrevs
     for r in narrow:
         for p in pr(r):
             if p in parentset:
                 cs.add(r)
     return cs
 
 def children(repo, subset, x):
     """``children(set)``
     Child changesets of changesets in set.
     """
     s = set(getset(repo, range(len(repo)), x))
     cs = _children(repo, subset, s)
     return [r for r in subset if r in cs]
 
 def closed(repo, subset, x):
     """``closed()``
     Changeset is closed.
     """
     # i18n: "closed" is a keyword
     getargs(x, 0, 0, _("closed takes no arguments"))
     return [r for r in subset if repo[r].closesbranch()]
 
 def contains(repo, subset, x):
     """``contains(pattern)``
     Revision contains a file matching pattern. See :hg:`help patterns`
     for information about file patterns.
     """
     # i18n: "contains" is a keyword
     pat = getstring(x, _("contains requires a pattern"))
     m = None
     s = []
     if not matchmod.patkind(pat):
         for r in subset:
             if pat in repo[r]:
                 s.append(r)
     else:
         for r in subset:
             c = repo[r]
             if not m or matchmod.patkind(pat) == 'set':
                 m = matchmod.match(repo.root, repo.getcwd(), [pat], ctx=c)
             for f in c.manifest():
                 if m(f):
                     s.append(r)
                     break
     return s
 
 def converted(repo, subset, x):
     """``converted([id])``
     Changesets converted from the given identifier in the old repository if
     present, or all converted changesets if no identifier is specified.
     """
 
     # There is exactly no chance of resolving the revision, so do a simple
     # string compare and hope for the best
 
     rev = None
     # i18n: "converted" is a keyword
     l = getargs(x, 0, 1, _('converted takes one or no arguments'))
     if l:
         # i18n: "converted" is a keyword
         rev = getstring(l[0], _('converted requires a revision'))
 
     def _matchvalue(r):
         source = repo[r].extra().get('convert_revision', None)
         return source is not None and (rev is None or source.startswith(rev))
 
     return [r for r in subset if _matchvalue(r)]
 
 def date(repo, subset, x):
     """``date(interval)``
     Changesets within the interval, see :hg:`help dates`.
     """
     # i18n: "date" is a keyword
     ds = getstring(x, _("date requires a string"))
     dm = util.matchdate(ds)
     return [r for r in subset if dm(repo[r].date()[0])]
 
 def desc(repo, subset, x):
     """``desc(string)``
     Search commit message for string. The match is case-insensitive.
     """
     # i18n: "desc" is a keyword
     ds = encoding.lower(getstring(x, _("desc requires a string")))
     l = []
     for r in subset:
         c = repo[r]
         if ds in encoding.lower(c.description()):
             l.append(r)
     return l
 
 def _descendants(repo, subset, x, followfirst=False):
     args = getset(repo, range(len(repo)), x)
     if not args:
         return []
     s = set(_revdescendants(repo, args, followfirst)) | set(args)
     return [r for r in subset if r in s]
 
 def descendants(repo, subset, x):
     """``descendants(set)``
     Changesets which are descendants of changesets in set.
     """
     return _descendants(repo, subset, x)
 
 def _firstdescendants(repo, subset, x):
     # ``_firstdescendants(set)``
     # Like ``descendants(set)`` but follows only the first parents.
     return _descendants(repo, subset, x, followfirst=True)
 
 def destination(repo, subset, x):
     """``destination([set])``
     Changesets that were created by a graft, transplant or rebase operation,
     with the given revisions specified as the source.  Omitting the optional set
     is the same as passing all().
     """
     if x is not None:
         args = set(getset(repo, range(len(repo)), x))
     else:
         args = set(getall(repo, range(len(repo)), x))
 
     dests = set()
 
     # subset contains all of the possible destinations that can be returned, so
     # iterate over them and see if their source(s) were provided in the args.
     # Even if the immediate src of r is not in the args, src's source (or
     # further back) may be.  Scanning back further than the immediate src allows
     # transitive transplants and rebases to yield the same results as transitive
     # grafts.
     for r in subset:
         src = _getrevsource(repo, r)
         lineage = None
 
         while src is not None:
             if lineage is None:
                 lineage = list()
 
             lineage.append(r)
 
             # The visited lineage is a match if the current source is in the arg
             # set.  Since every candidate dest is visited by way of iterating
             # subset, any dests further back in the lineage will be tested by a
             # different iteration over subset.  Likewise, if the src was already
             # selected, the current lineage can be selected without going back
             # further.
             if src in args or src in dests:
                 dests.update(lineage)
                 break
 
             r = src
             src = _getrevsource(repo, r)
 
     return [r for r in subset if r in dests]
 
 def draft(repo, subset, x):
     """``draft()``
     Changeset in draft phase."""
     # i18n: "draft" is a keyword
     getargs(x, 0, 0, _("draft takes no arguments"))
     pc = repo._phasecache
     return [r for r in subset if pc.phase(repo, r) == phases.draft]
 
 def extinct(repo, subset, x):
     """``extinct()``
     Obsolete changesets with obsolete descendants only.
     """
     # i18n: "extinct" is a keyword
     getargs(x, 0, 0, _("extinct takes no arguments"))
-    extinctset = set(repo.revs('(obsolete()::) - (::(not obsolete()))'))
-    return [r for r in subset if r in extinctset]
+    extincts = obsmod.getobscache(repo, 'extinct')
+    return [r for r in subset if r in extincts]
 
 def extra(repo, subset, x):
     """``extra(label, [value])``
     Changesets with the given label in the extra metadata, with the given
     optional value.
 
     If `value` starts with `re:`, the remainder of the value is treated as
     a regular expression. To match a value that actually starts with `re:`,
     use the prefix `literal:`.
     """
 
     # i18n: "extra" is a keyword
     l = getargs(x, 1, 2, _('extra takes at least 1 and at most 2 arguments'))
     # i18n: "extra" is a keyword
     label = getstring(l[0], _('first argument to extra must be a string'))
     value = None
 
     if len(l) > 1:
         # i18n: "extra" is a keyword
         value = getstring(l[1], _('second argument to extra must be a string'))
         kind, value, matcher = _stringmatcher(value)
 
     def _matchvalue(r):
         extra = repo[r].extra()
         return label in extra and (value is None or matcher(extra[label]))
 
     return [r for r in subset if _matchvalue(r)]
 
 def filelog(repo, subset, x):
     """``filelog(pattern)``
     Changesets connected to the specified filelog.
 
     For performance reasons, ``filelog()`` does not show every changeset
     that affects the requested file(s). See :hg:`help log` for details. For
     a slower, more accurate result, use ``file()``.
     """
 
     # i18n: "filelog" is a keyword
     pat = getstring(x, _("filelog requires a pattern"))
     m = matchmod.match(repo.root, repo.getcwd(), [pat], default='relpath',
                        ctx=repo[None])
     s = set()
 
     if not matchmod.patkind(pat):
         for f in m.files():
             fl = repo.file(f)
             for fr in fl:
                 s.add(fl.linkrev(fr))
     else:
         for f in repo[None]:
             if m(f):
                 fl = repo.file(f)
                 for fr in fl:
                     s.add(fl.linkrev(fr))
 
     return [r for r in subset if r in s]
 
 def first(repo, subset, x):
     """``first(set, [n])``
     An alias for limit().
     """
     return limit(repo, subset, x)
 
 def _follow(repo, subset, x, name, followfirst=False):
     l = getargs(x, 0, 1, _("%s takes no arguments or a filename") % name)
     c = repo['.']
     if l:
         x = getstring(l[0], _("%s expected a filename") % name)
         if x in c:
             cx = c[x]
             s = set(ctx.rev() for ctx in cx.ancestors(followfirst=followfirst))
             # include the revision responsible for the most recent version
             s.add(cx.linkrev())
         else:
             return []
     else:
         s = set(_revancestors(repo, [c.rev()], followfirst)) | set([c.rev()])
 
     return [r for r in subset if r in s]
 
 def follow(repo, subset, x):
     """``follow([file])``
     An alias for ``::.`` (ancestors of the working copy's first parent).
     If a filename is specified, the history of the given file is followed,
     including copies.
     """
     return _follow(repo, subset, x, 'follow')
 
 def _followfirst(repo, subset, x):
     # ``followfirst([file])``
     # Like ``follow([file])`` but follows only the first parent of
     # every revision or file revision.
     return _follow(repo, subset, x, '_followfirst', followfirst=True)
 
 def getall(repo, subset, x):
     """``all()``
     All changesets, the same as ``0:tip``.
     """
     # i18n: "all" is a keyword
     getargs(x, 0, 0, _("all takes no arguments"))
     return subset
 
 def grep(repo, subset, x):
     """``grep(regex)``
     Like ``keyword(string)`` but accepts a regex. Use ``grep(r'...')``
     to ensure special escape characters are handled correctly. Unlike
     ``keyword(string)``, the match is case-sensitive.
     """
     try:
         # i18n: "grep" is a keyword
         gr = re.compile(getstring(x, _("grep requires a string")))
     except re.error, e:
         raise error.ParseError(_('invalid match pattern: %s') % e)
     l = []
     for r in subset:
         c = repo[r]
         for e in c.files() + [c.user(), c.description()]:
             if gr.search(e):
                 l.append(r)
                 break
     return l
 
 def _matchfiles(repo, subset, x):
     # _matchfiles takes a revset list of prefixed arguments:
     #
     #   [p:foo, i:bar, x:baz]
     #
     # builds a match object from them and filters subset. Allowed
     # prefixes are 'p:' for regular patterns, 'i:' for include
     # patterns and 'x:' for exclude patterns. Use 'r:' prefix to pass
     # a revision identifier, or the empty string to reference the
     # working directory, from which the match object is
     # initialized. Use 'd:' to set the default matching mode, default
     # to 'glob'. At most one 'r:' and 'd:' argument can be passed.
 
     # i18n: "_matchfiles" is a keyword
     l = getargs(x, 1, -1, _("_matchfiles requires at least one argument"))
     pats, inc, exc = [], [], []
     hasset = False
     rev, default = None, None
     for arg in l:
         # i18n: "_matchfiles" is a keyword
         s = getstring(arg, _("_matchfiles requires string arguments"))
         prefix, value = s[:2], s[2:]
         if prefix == 'p:':
             pats.append(value)
         elif prefix == 'i:':
             inc.append(value)
         elif prefix == 'x:':
             exc.append(value)
         elif prefix == 'r:':
             if rev is not None:
                 # i18n: "_matchfiles" is a keyword
                 raise error.ParseError(_('_matchfiles expected at most one '
                                          'revision'))
             rev = value
         elif prefix == 'd:':
             if default is not None:
                 # i18n: "_matchfiles" is a keyword
                 raise error.ParseError(_('_matchfiles expected at most one '
                                          'default mode'))
             default = value
         else:
             # i18n: "_matchfiles" is a keyword
             raise error.ParseError(_('invalid _matchfiles prefix: %s') % prefix)
         if not hasset and matchmod.patkind(value) == 'set':
             hasset = True
     if not default:
         default = 'glob'
     m = None
     s = []
     for r in subset:
         c = repo[r]
         if not m or (hasset and rev is None):
             ctx = c
             if rev is not None:
                 ctx = repo[rev or None]
             m = matchmod.match(repo.root, repo.getcwd(), pats, include=inc,
                                exclude=exc, ctx=ctx, default=default)
         for f in c.files():
             if m(f):
                 s.append(r)
                 break
     return s
 
 def hasfile(repo, subset, x):
     """``file(pattern)``
     Changesets affecting files matched by pattern.
 
     For a faster but less accurate result, consider using ``filelog()``
     instead.
     """
     # i18n: "file" is a keyword
     pat = getstring(x, _("file requires a pattern"))
     return _matchfiles(repo, subset, ('string', 'p:' + pat))
 
 def head(repo, subset, x):
     """``head()``
     Changeset is a named branch head.
     """
     # i18n: "head" is a keyword
     getargs(x, 0, 0, _("head takes no arguments"))
     hs = set()
     for b, ls in repo.branchmap().iteritems():
         hs.update(repo[h].rev() for h in ls)
     return [r for r in subset if r in hs]
 
 def heads(repo, subset, x):
     """``heads(set)``
     Members of set with no children in set.
     """
     s = getset(repo, subset, x)
     ps = set(parents(repo, subset, x))
     return [r for r in s if r not in ps]
 
 def hidden(repo, subset, x):
     """``hidden()``
     Hidden changesets.
     """
     # i18n: "hidden" is a keyword
     getargs(x, 0, 0, _("hidden takes no arguments"))
     return [r for r in subset if r in repo.hiddenrevs]
 
 def keyword(repo, subset, x):
     """``keyword(string)``
     Search commit message, user name, and names of changed files for
     string. The match is case-insensitive.
     """
     # i18n: "keyword" is a keyword
     kw = encoding.lower(getstring(x, _("keyword requires a string")))
     l = []
     for r in subset:
         c = repo[r]
         t = " ".join(c.files() + [c.user(), c.description()])
         if kw in encoding.lower(t):
             l.append(r)
     return l
 
 def limit(repo, subset, x):
     """``limit(set, [n])``
     First n members of set, defaulting to 1.
     """
     # i18n: "limit" is a keyword
     l = getargs(x, 1, 2, _("limit requires one or two arguments"))
     try:
         lim = 1
         if len(l) == 2:
             # i18n: "limit" is a keyword
             lim = int(getstring(l[1], _("limit requires a number")))
     except (TypeError, ValueError):
         # i18n: "limit" is a keyword
         raise error.ParseError(_("limit expects a number"))
     ss = set(subset)
     os = getset(repo, range(len(repo)), l[0])[:lim]
     return [r for r in os if r in ss]
 
 def last(repo, subset, x):
     """``last(set, [n])``
     Last n members of set, defaulting to 1.
     """
     # i18n: "last" is a keyword
     l = getargs(x, 1, 2, _("last requires one or two arguments"))
     try:
         lim = 1
         if len(l) == 2:
             # i18n: "last" is a keyword
             lim = int(getstring(l[1], _("last requires a number")))
     except (TypeError, ValueError):
         # i18n: "last" is a keyword
         raise error.ParseError(_("last expects a number"))
     ss = set(subset)
     os = getset(repo, range(len(repo)), l[0])[-lim:]
     return [r for r in os if r in ss]
 
 def maxrev(repo, subset, x):
     """``max(set)``
     Changeset with highest revision number in set.
     """
     os = getset(repo, range(len(repo)), x)
     if os:
         m = max(os)
         if m in subset:
             return [m]
     return []
 
 def merge(repo, subset, x):
     """``merge()``
     Changeset is a merge changeset.
     """
     # i18n: "merge" is a keyword
     getargs(x, 0, 0, _("merge takes no arguments"))
     cl = repo.changelog
     return [r for r in subset if cl.parentrevs(r)[1] != -1]
 
 def minrev(repo, subset, x):
     """``min(set)``
     Changeset with lowest revision number in set.
     """
     os = getset(repo, range(len(repo)), x)
     if os:
         m = min(os)
         if m in subset:
             return [m]
     return []
 
 def modifies(repo, subset, x):
     """``modifies(pattern)``
     Changesets modifying files matched by pattern.
     """
     # i18n: "modifies" is a keyword
     pat = getstring(x, _("modifies requires a pattern"))
     return checkstatus(repo, subset, pat, 0)
 
 def node_(repo, subset, x):
     """``id(string)``
     Revision non-ambiguously specified by the given hex string prefix.
     """
     # i18n: "id" is a keyword
     l = getargs(x, 1, 1, _("id requires one argument"))
     # i18n: "id" is a keyword
     n = getstring(l[0], _("id requires a string"))
     if len(n) == 40:
         rn = repo[n].rev()
     else:
         rn = None
         pm = repo.changelog._partialmatch(n)
         if pm is not None:
             rn = repo.changelog.rev(pm)
 
     return [r for r in subset if r == rn]
 
 def obsolete(repo, subset, x):
     """``obsolete()``
     Mutable changeset with a newer version."""
     # i18n: "obsolete" is a keyword
     getargs(x, 0, 0, _("obsolete takes no arguments"))
-    return [r for r in subset if repo[r].obsolete()]
+    obsoletes = obsmod.getobscache(repo, 'obsolete')
+    return [r for r in subset if r in obsoletes]
 
 def origin(repo, subset, x):
     """``origin([set])``
     Changesets that were specified as a source for the grafts, transplants or
     rebases that created the given revisions.  Omitting the optional set is the
     same as passing all().  If a changeset created by these operations is itself
     specified as a source for one of these operations, only the source changeset
     for the first operation is selected.
     """
     if x is not None:
         args = set(getset(repo, range(len(repo)), x))
     else:
         args = set(getall(repo, range(len(repo)), x))
 
     def _firstsrc(rev):
         src = _getrevsource(repo, rev)
         if src is None:
             return None
 
         while True:
             prev = _getrevsource(repo, src)
 
             if prev is None:
                 return src
             src = prev
 
     o = set([_firstsrc(r) for r in args])
     return [r for r in subset if r in o]
 
 def outgoing(repo, subset, x):
     """``outgoing([path])``
     Changesets not found in the specified destination repository, or the
     default push location.
     """
     import hg # avoid start-up nasties
     # i18n: "outgoing" is a keyword
     l = getargs(x, 0, 1, _("outgoing takes one or no arguments"))
     # i18n: "outgoing" is a keyword
     dest = l and getstring(l[0], _("outgoing requires a repository path")) or ''
     dest = repo.ui.expandpath(dest or 'default-push', dest or 'default')
     dest, branches = hg.parseurl(dest)
     revs, checkout = hg.addbranchrevs(repo, repo, branches, [])
     if revs:
         revs = [repo.lookup(rev) for rev in revs]
     other = hg.peer(repo, {}, dest)
     repo.ui.pushbuffer()
     outgoing = discovery.findcommonoutgoing(repo, other, onlyheads=revs)
     repo.ui.popbuffer()
     cl = repo.changelog
     o = set([cl.rev(r) for r in outgoing.missing])
     return [r for r in subset if r in o]
 
 def p1(repo, subset, x):
     """``p1([set])``
     First parent of changesets in set, or the working directory.
     """
     if x is None:
         p = repo[x].p1().rev()
         return [r for r in subset if r == p]
 
     ps = set()
     cl = repo.changelog
     for r in getset(repo, range(len(repo)), x):
         ps.add(cl.parentrevs(r)[0])
     return [r for r in subset if r in ps]
 
 def p2(repo, subset, x):
     """``p2([set])``
     Second parent of changesets in set, or the working directory.
     """
     if x is None:
         ps = repo[x].parents()
         try:
             p = ps[1].rev()
             return [r for r in subset if r == p]
         except IndexError:
             return []
 
     ps = set()
     cl = repo.changelog
     for r in getset(repo, range(len(repo)), x):
         ps.add(cl.parentrevs(r)[1])
     return [r for r in subset if r in ps]
 
 def parents(repo, subset, x):
     """``parents([set])``
     The set of all parents for all changesets in set, or the working directory.
     """
     if x is None:
         ps = tuple(p.rev() for p in repo[x].parents())
         return [r for r in subset if r in ps]
 
     ps = set()
     cl = repo.changelog
     for r in getset(repo, range(len(repo)), x):
         ps.update(cl.parentrevs(r))
     return [r for r in subset if r in ps]
 
 def parentspec(repo, subset, x, n):
     """``set^0``
     The set.
     ``set^1`` (or ``set^``), ``set^2``
     First or second parent, respectively, of all changesets in set.
     """
     try:
         n = int(n[1])
         if n not in (0, 1, 2):
             raise ValueError
     except (TypeError, ValueError):
         raise error.ParseError(_("^ expects a number 0, 1, or 2"))
     ps = set()
     cl = repo.changelog
     for r in getset(repo, subset, x):
         if n == 0:
             ps.add(r)
         elif n == 1:
             ps.add(cl.parentrevs(r)[0])
         elif n == 2:
             parents = cl.parentrevs(r)
             if len(parents) > 1:
                 ps.add(parents[1])
     return [r for r in subset if r in ps]
 
 def present(repo, subset, x):
     """``present(set)``
     An empty set, if any revision in set isn't found; otherwise,
     all revisions in set.
 
     If any of specified revisions is not present in the local repository,
     the query is normally aborted. But this predicate allows the query
     to continue even in such cases.
     """
     try:
         return getset(repo, subset, x)
     except error.RepoLookupError:
         return []
 
 def public(repo, subset, x):
     """``public()``
     Changeset in public phase."""
     # i18n: "public" is a keyword
     getargs(x, 0, 0, _("public takes no arguments"))
     pc = repo._phasecache
     return [r for r in subset if pc.phase(repo, r) == phases.public]
 
 def remote(repo, subset, x):
     """``remote([id [,path]])``
     Local revision that corresponds to the given identifier in a
     remote repository, if present. Here, the '.' identifier is a
     synonym for the current local branch.
     """
 
     import hg # avoid start-up nasties
     # i18n: "remote" is a keyword
     l = getargs(x, 0, 2, _("remote takes one, two or no arguments"))
 
     q = '.'
     if len(l) > 0:
     # i18n: "remote" is a keyword
         q = getstring(l[0], _("remote requires a string id"))
     if q == '.':
         q = repo['.'].branch()
 
     dest = ''
     if len(l) > 1:
         # i18n: "remote" is a keyword
         dest = getstring(l[1], _("remote requires a repository path"))
     dest = repo.ui.expandpath(dest or 'default')
     dest, branches = hg.parseurl(dest)
     revs, checkout = hg.addbranchrevs(repo, repo, branches, [])
     if revs:
         revs = [repo.lookup(rev) for rev in revs]
     other = hg.peer(repo, {}, dest)
     n = other.lookup(q)
     if n in repo:
         r = repo[n].rev()
         if r in subset:
             return [r]
     return []
 
 def removes(repo, subset, x):
     """``removes(pattern)``
     Changesets which remove files matching pattern.
     """
     # i18n: "removes" is a keyword
     pat = getstring(x, _("removes requires a pattern"))
     return checkstatus(repo, subset, pat, 2)
 
 def rev(repo, subset, x):
     """``rev(number)``
     Revision with the given numeric identifier.
     """
     # i18n: "rev" is a keyword
     l = getargs(x, 1, 1, _("rev requires one argument"))
     try:
         # i18n: "rev" is a keyword
         l = int(getstring(l[0], _("rev requires a number")))
     except (TypeError, ValueError):
         # i18n: "rev" is a keyword
         raise error.ParseError(_("rev expects a number"))
     return [r for r in subset if r == l]
 
 def matching(repo, subset, x):
     """``matching(revision [, field])``
     Changesets in which a given set of fields match the set of fields in the
     selected revision or set.
 
     To match more than one field pass the list of fields to match separated
     by spaces (e.g. ``author description``).
 
     Valid fields are most regular revision fields and some special fields.
 
     Regular revision fields are ``description``, ``author``, ``branch``,
     ``date``, ``files``, ``phase``, ``parents``, ``substate``, ``user``
     and ``diff``.
     Note that ``author`` and ``user`` are synonyms. ``diff`` refers to the
     contents of the revision. Two revisions matching their ``diff`` will
     also match their ``files``.
 
     Special fields are ``summary`` and ``metadata``:
     ``summary`` matches the first line of the description.
     ``metadata`` is equivalent to matching ``description user date``
     (i.e. it matches the main metadata fields).
 
     ``metadata`` is the default field which is used when no fields are
     specified. You can match more than one field at a time.
     """
     # i18n: "matching" is a keyword
     l = getargs(x, 1, 2, _("matching takes 1 or 2 arguments"))
 
     revs = getset(repo, xrange(len(repo)), l[0])
 
     fieldlist = ['metadata']
     if len(l) > 1:
             fieldlist = getstring(l[1],
                 # i18n: "matching" is a keyword
                 _("matching requires a string "
                 "as its second argument")).split()
 
     # Make sure that there are no repeated fields,
     # expand the 'special' 'metadata' field type
     # and check the 'files' whenever we check the 'diff'
     fields = []
     for field in fieldlist:
         if field == 'metadata':
             fields += ['user', 'description', 'date']
         elif field == 'diff':
             # a revision matching the diff must also match the files
             # since matching the diff is very costly, make sure to
             # also match the files first
             fields += ['files', 'diff']
         else:
             if field == 'author':
                 field = 'user'
             fields.append(field)
     fields = set(fields)
     if 'summary' in fields and 'description' in fields:
         # If a revision matches its description it also matches its summary
         fields.discard('summary')
 
     # We may want to match more than one field
     # Not all fields take the same amount of time to be matched
     # Sort the selected fields in order of increasing matching cost
     fieldorder = ['phase', 'parents', 'user', 'date', 'branch', 'summary',
         'files', 'description', 'substate', 'diff']
     def fieldkeyfunc(f):
         try:
             return fieldorder.index(f)
         except ValueError:
             # assume an unknown field is very costly
             return len(fieldorder)
     fields = list(fields)
     fields.sort(key=fieldkeyfunc)
 
     # Each field will be matched with its own "getfield" function
     # which will be added to the getfieldfuncs array of functions
     getfieldfuncs = []
     _funcs = {
         'user': lambda r: repo[r].user(),
         'branch': lambda r: repo[r].branch(),
         'date': lambda r: repo[r].date(),
         'description': lambda r: repo[r].description(),
         'files': lambda r: repo[r].files(),
         'parents': lambda r: repo[r].parents(),
         'phase': lambda r: repo[r].phase(),
         'substate': lambda r: repo[r].substate,
         'summary': lambda r: repo[r].description().splitlines()[0],
         'diff': lambda r: list(repo[r].diff(git=True),)
     }
     for info in fields:
         getfield = _funcs.get(info, None)
         if getfield is None:
             raise error.ParseError(
                 # i18n: "matching" is a keyword
                 _("unexpected field name passed to matching: %s") % info)
         getfieldfuncs.append(getfield)
     # convert the getfield array of functions into a "getinfo" function
     # which returns an array of field values (or a single value if there
     # is only one field to match)
     getinfo = lambda r: [f(r) for f in getfieldfuncs]
 
     matches = set()
     for rev in revs:
         target = getinfo(rev)
         for r in subset:
             match = True
             for n, f in enumerate(getfieldfuncs):
                 if target[n] != f(r):
                     match = False
                     break
             if match:
                 matches.add(r)
     return [r for r in subset if r in matches]
 
 def reverse(repo, subset, x):
     """``reverse(set)``
     Reverse order of set.
     """
     l = getset(repo, subset, x)
     if not isinstance(l, list):
         l = list(l)
     l.reverse()
     return l
 
 def roots(repo, subset, x):
     """``roots(set)``
     Changesets in set with no parent changeset in set.
     """
     s = set(getset(repo, xrange(len(repo)), x))
     subset = [r for r in subset if r in s]
     cs = _children(repo, subset, s)
     return [r for r in subset if r not in cs]
 
 def secret(repo, subset, x):
     """``secret()``
     Changeset in secret phase."""
     # i18n: "secret" is a keyword
     getargs(x, 0, 0, _("secret takes no arguments"))
     pc = repo._phasecache
     return [r for r in subset if pc.phase(repo, r) == phases.secret]
 
 def sort(repo, subset, x):
     """``sort(set[, [-]key...])``
     Sort set by keys. The default sort order is ascending, specify a key
     as ``-key`` to sort in descending order.
 
     The keys can be:
 
     - ``rev`` for the revision number,
     - ``branch`` for the branch name,
     - ``desc`` for the commit message (description),
     - ``user`` for user name (``author`` can be used as an alias),
     - ``date`` for the commit date
     """
     # i18n: "sort" is a keyword
     l = getargs(x, 1, 2, _("sort requires one or two arguments"))
     keys = "rev"
     if len(l) == 2:
         # i18n: "sort" is a keyword
         keys = getstring(l[1], _("sort spec must be a string"))
 
     s = l[0]
     keys = keys.split()
     l = []
     def invert(s):
         return "".join(chr(255 - ord(c)) for c in s)
     for r in getset(repo, subset, s):
         c = repo[r]
         e = []
         for k in keys:
             if k == 'rev':
                 e.append(r)
             elif k == '-rev':
                 e.append(-r)
             elif k == 'branch':
                 e.append(c.branch())
             elif k == '-branch':
                 e.append(invert(c.branch()))
             elif k == 'desc':
                 e.append(c.description())
             elif k == '-desc':
                 e.append(invert(c.description()))
             elif k in 'user author':
                 e.append(c.user())
             elif k in '-user -author':
                 e.append(invert(c.user()))
             elif k == 'date':
                 e.append(c.date()[0])
             elif k == '-date':
                 e.append(-c.date()[0])
             else:
                 raise error.ParseError(_("unknown sort key %r") % k)
         e.append(r)
         l.append(e)
     l.sort()
     return [e[-1] for e in l]
 
 def _stringmatcher(pattern):
     """
     accepts a string, possibly starting with 're:' or 'literal:' prefix.
     returns the matcher name, pattern, and matcher function.
     missing or unknown prefixes are treated as literal matches.
 
     helper for tests:
     >>> def test(pattern, *tests):
     ...     kind, pattern, matcher = _stringmatcher(pattern)
     ...     return (kind, pattern, [bool(matcher(t)) for t in tests])
 
     exact matching (no prefix):
     >>> test('abcdefg', 'abc', 'def', 'abcdefg')
     ('literal', 'abcdefg', [False, False, True])
 
     regex matching ('re:' prefix)
     >>> test('re:a.+b', 'nomatch', 'fooadef', 'fooadefbar')
     ('re', 'a.+b', [False, False, True])
 
     force exact matches ('literal:' prefix)
     >>> test('literal:re:foobar', 'foobar', 're:foobar')
     ('literal', 're:foobar', [False, True])
 
     unknown prefixes are ignored and treated as literals
     >>> test('foo:bar', 'foo', 'bar', 'foo:bar')
     ('literal', 'foo:bar', [False, False, True])
     """
     if pattern.startswith('re:'):
         pattern = pattern[3:]
         try:
             regex = re.compile(pattern)
         except re.error, e:
             raise error.ParseError(_('invalid regular expression: %s')
                                    % e)
         return 're', pattern, regex.search
     elif pattern.startswith('literal:'):
         pattern = pattern[8:]
     return 'literal', pattern, pattern.__eq__
 
 def _substringmatcher(pattern):
     kind, pattern, matcher = _stringmatcher(pattern)
     if kind == 'literal':
         matcher = lambda s: pattern in s
     return kind, pattern, matcher
 
 def tag(repo, subset, x):
     """``tag([name])``
     The specified tag by name, or all tagged revisions if no name is given.
     """
     # i18n: "tag" is a keyword
     args = getargs(x, 0, 1, _("tag takes one or no arguments"))
     cl = repo.changelog
     if args:
         pattern = getstring(args[0],
                             # i18n: "tag" is a keyword
                             _('the argument to tag must be a string'))
         kind, pattern, matcher = _stringmatcher(pattern)
         if kind == 'literal':
             # avoid resolving all tags
             tn = repo._tagscache.tags.get(pattern, None)
             if tn is None:
                 raise util.Abort(_("tag '%s' does not exist") % pattern)
             s = set([repo[tn].rev()])
         else:
             s = set([cl.rev(n) for t, n in repo.tagslist() if matcher(t)])
             if not s:
                 raise util.Abort(_("no tags exist that match '%s'") % pattern)
     else:
         s = set([cl.rev(n) for t, n in repo.tagslist() if t != 'tip'])
     return [r for r in subset if r in s]
 
 def tagged(repo, subset, x):
     return tag(repo, subset, x)
 
 def unstable(repo, subset, x):
     """``unstable()``
     Non-obsolete changesets with obsolete ancestors.
     """
     # i18n: "unstable" is a keyword
     getargs(x, 0, 0, _("unstable takes no arguments"))
-    unstableset = set(repo.revs('(obsolete()::) - obsolete()'))
-    return [r for r in subset if r in unstableset]
+    unstables = obsmod.getobscache(repo, 'unstable')
+    return [r for r in subset if r in unstables]
 
 
 def user(repo, subset, x):
     """``user(string)``
     User name contains string. The match is case-insensitive.
 
     If `string` starts with `re:`, the remainder of the string is treated as
     a regular expression. To match a user that actually contains `re:`, use
     the prefix `literal:`.
     """
     return author(repo, subset, x)
 
 # for internal use
 def _list(repo, subset, x):
     s = getstring(x, "internal error")
     if not s:
         return []
     if not isinstance(subset, set):
         subset = set(subset)
     ls = [repo[r].rev() for r in s.split('\0')]
     return [r for r in ls if r in subset]
 
 symbols = {
     "adds": adds,
     "all": getall,
     "ancestor": ancestor,
     "ancestors": ancestors,
     "_firstancestors": _firstancestors,
     "author": author,
     "bisect": bisect,
     "bisected": bisected,
     "bookmark": bookmark,
     "branch": branch,
     "children": children,
     "closed": closed,
     "contains": contains,
     "converted": converted,
     "date": date,
     "desc": desc,
     "descendants": descendants,
     "_firstdescendants": _firstdescendants,
     "destination": destination,
     "draft": draft,
     "extinct": extinct,
     "extra": extra,
     "file": hasfile,
     "filelog": filelog,
     "first": first,
     "follow": follow,
     "_followfirst": _followfirst,
     "grep": grep,
     "head": head,
     "heads": heads,
     "hidden": hidden,
     "id": node_,
     "keyword": keyword,
     "last": last,
     "limit": limit,
     "_matchfiles": _matchfiles,
     "max": maxrev,
     "merge": merge,
     "min": minrev,
     "modifies": modifies,
     "obsolete": obsolete,
     "origin": origin,
     "outgoing": outgoing,
     "p1": p1,
     "p2": p2,
     "parents": parents,
     "present": present,
     "public": public,
     "remote": remote,
     "removes": removes,
     "rev": rev,
     "reverse": reverse,
     "roots": roots,
     "sort": sort,
     "secret": secret,
     "matching": matching,
     "tag": tag,
     "tagged": tagged,
     "user": user,
     "unstable": unstable,
     "_list": _list,
 }
 
 methods = {
     "range": rangeset,
     "dagrange": dagrange,
     "string": stringset,
     "symbol": symbolset,
     "and": andset,
     "or": orset,
     "not": notset,
     "list": listset,
     "func": func,
     "ancestor": ancestorspec,
     "parent": parentspec,
     "parentpost": p1,
 }
 
 def optimize(x, small):
     if x is None:
         return 0, x
 
     smallbonus = 1
     if small:
         smallbonus = .5
 
     op = x[0]
     if op == 'minus':
         return optimize(('and', x[1], ('not', x[2])), small)
     elif op == 'dagrangepre':
         return optimize(('func', ('symbol', 'ancestors'), x[1]), small)
     elif op == 'dagrangepost':
         return optimize(('func', ('symbol', 'descendants'), x[1]), small)
     elif op == 'rangepre':
         return optimize(('range', ('string', '0'), x[1]), small)
     elif op == 'rangepost':
         return optimize(('range', x[1], ('string', 'tip')), small)
     elif op == 'negate':
         return optimize(('string',
                          '-' + getstring(x[1], _("can't negate that"))), small)
     elif op in 'string symbol negate':
         return smallbonus, x # single revisions are small
     elif op == 'and':
         wa, ta = optimize(x[1], True)
         wb, tb = optimize(x[2], True)
         w = min(wa, wb)
         if wa > wb:
             return w, (op, tb, ta)
         return w, (op, ta, tb)
     elif op == 'or':
         wa, ta = optimize(x[1], False)
         wb, tb = optimize(x[2], False)
         if wb < wa:
             wb, wa = wa, wb
         return max(wa, wb), (op, ta, tb)
     elif op == 'not':
         o = optimize(x[1], not small)
         return o[0], (op, o[1])
     elif op == 'parentpost':
         o = optimize(x[1], small)
         return o[0], (op, o[1])
     elif op == 'group':
         return optimize(x[1], small)
     elif op in 'dagrange range list parent ancestorspec':
         if op == 'parent':
             # x^:y means (x^) : y, not x ^ (:y)
             post = ('parentpost', x[1])
             if x[2][0] == 'dagrangepre':
                 return optimize(('dagrange', post, x[2][1]), small)
             elif x[2][0] == 'rangepre':
                 return optimize(('range', post, x[2][1]), small)
 
         wa, ta = optimize(x[1], small)
         wb, tb = optimize(x[2], small)
         return wa + wb, (op, ta, tb)
     elif op == 'func':
         f = getstring(x[1], _("not a symbol"))
         wa, ta = optimize(x[2], small)
         if f in ("author branch closed date desc file grep keyword "
                  "outgoing user"):
             w = 10 # slow
         elif f in "modifies adds removes":
             w = 30 # slower
         elif f == "contains":
             w = 100 # very slow
         elif f == "ancestor":
             w = 1 * smallbonus
         elif f in "reverse limit first":
             w = 0
         elif f in "sort":
             w = 10 # assume most sorts look at changelog
         else:
             w = 1
         return w + wa, (op, x[1], ta)
     return 1, x
 
 _aliasarg = ('func', ('symbol', '_aliasarg'))
 def _getaliasarg(tree):
     """If tree matches ('func', ('symbol', '_aliasarg'), ('string', X))
     return X, None otherwise.
     """
     if (len(tree) == 3 and tree[:2] == _aliasarg
         and tree[2][0] == 'string'):
         return tree[2][1]
     return None
 
 def _checkaliasarg(tree, known=None):
     """Check tree contains no _aliasarg construct or only ones which
     value is in known. Used to avoid alias placeholders injection.
     """
     if isinstance(tree, tuple):
         arg = _getaliasarg(tree)
         if arg is not None and (not known or arg not in known):
             raise error.ParseError(_("not a function: %s") % '_aliasarg')
         for t in tree:
             _checkaliasarg(t, known)
 
 class revsetalias(object):
     funcre = re.compile('^([^(]+)\(([^)]+)\)$')
     args = None
 
     def __init__(self, name, value):
         '''Aliases like:
 
         h = heads(default)
         b($1) = ancestors($1) - ancestors(default)
         '''
         m = self.funcre.search(name)
         if m:
             self.name = m.group(1)
             self.tree = ('func', ('symbol', m.group(1)))
             self.args = [x.strip() for x in m.group(2).split(',')]
             for arg in self.args:
                 # _aliasarg() is an unknown symbol only used separate
                 # alias argument placeholders from regular strings.
                 value = value.replace(arg, '_aliasarg(%r)' % (arg,))
         else:
             self.name = name
             self.tree = ('symbol', name)
 
         self.replacement, pos = parse(value)
         if pos != len(value):
             raise error.ParseError(_('invalid token'), pos)
         # Check for placeholder injection
         _checkaliasarg(self.replacement, self.args)
 
 def _getalias(aliases, tree):
     """If tree looks like an unexpanded alias, return it. Return None
     otherwise.
     """
     if isinstance(tree, tuple) and tree:
         if tree[0] == 'symbol' and len(tree) == 2:
             name = tree[1]
             alias = aliases.get(name)
             if alias and alias.args is None and alias.tree == tree:
                 return alias
         if tree[0] == 'func' and len(tree) > 1:
             if tree[1][0] == 'symbol' and len(tree[1]) == 2:
                 name = tree[1][1]
                 alias = aliases.get(name)
                 if alias and alias.args is not None and alias.tree == tree[:2]:
                     return alias
     return None
 
 def _expandargs(tree, args):
     """Replace _aliasarg instances with the substitution value of the
     same name in args, recursively.
     """
     if not tree or not isinstance(tree, tuple):
         return tree
     arg = _getaliasarg(tree)
     if arg is not None:
         return args[arg]
     return tuple(_expandargs(t, args) for t in tree)
 
 def _expandaliases(aliases, tree, expanding, cache):
     """Expand aliases in tree, recursively.
 
     'aliases' is a dictionary mapping user defined aliases to
     revsetalias objects.
     """
     if not isinstance(tree, tuple):
         # Do not expand raw strings
         return tree
     alias = _getalias(aliases, tree)
     if alias is not None:
         if alias in expanding:
             raise error.ParseError(_('infinite expansion of revset alias "%s" '
                                      'detected') % alias.name)
         expanding.append(alias)
         if alias.name not in cache:
             cache[alias.name] = _expandaliases(aliases, alias.replacement,
                                                expanding, cache)
         result = cache[alias.name]
         expanding.pop()
         if alias.args is not None:
             l = getlist(tree[2])
             if len(l) != len(alias.args):
                 raise error.ParseError(
                     _('invalid number of arguments: %s') % len(l))
             l = [_expandaliases(aliases, a, [], cache) for a in l]
             result = _expandargs(result, dict(zip(alias.args, l)))
     else:
         result = tuple(_expandaliases(aliases, t, expanding, cache)
                        for t in tree)
     return result
 
 def findaliases(ui, tree):
     _checkaliasarg(tree)
     aliases = {}
     for k, v in ui.configitems('revsetalias'):
         alias = revsetalias(k, v)
         aliases[alias.name] = alias
     return _expandaliases(aliases, tree, [], {})
 
 parse = parser.parser(tokenize, elements).parse
 
 def match(ui, spec):
     if not spec:
         raise error.ParseError(_("empty query"))
     tree, pos = parse(spec)
     if (pos != len(spec)):
         raise error.ParseError(_("invalid token"), pos)
     if ui:
         tree = findaliases(ui, tree)
     weight, tree = optimize(tree, True)
     def mfunc(repo, subset):
         return getset(repo, subset, tree)
     return mfunc
 
 def formatspec(expr, *args):
     '''
     This is a convenience function for using revsets internally, and
     escapes arguments appropriately. Aliases are intentionally ignored
     so that intended expression behavior isn't accidentally subverted.
 
     Supported arguments:
 
     %r = revset expression, parenthesized
     %d = int(arg), no quoting
     %s = string(arg), escaped and single-quoted
     %b = arg.branch(), escaped and single-quoted
     %n = hex(arg), single-quoted
     %% = a literal '%'
 
     Prefixing the type with 'l' specifies a parenthesized list of that type.
 
     >>> formatspec('%r:: and %lr', '10 or 11', ("this()", "that()"))
     '(10 or 11):: and ((this()) or (that()))'
     >>> formatspec('%d:: and not %d::', 10, 20)
     '10:: and not 20::'
     >>> formatspec('%ld or %ld', [], [1])
     "_list('') or 1"
     >>> formatspec('keyword(%s)', 'foo\\xe9')
     "keyword('foo\\\\xe9')"
     >>> b = lambda: 'default'
     >>> b.branch = b
     >>> formatspec('branch(%b)', b)
     "branch('default')"
     >>> formatspec('root(%ls)', ['a', 'b', 'c', 'd'])
     "root(_list('a\\x00b\\x00c\\x00d'))"
     '''
 
     def quote(s):
         return repr(str(s))
 
     def argtype(c, arg):
         if c == 'd':
             return str(int(arg))
         elif c == 's':
             return quote(arg)
         elif c == 'r':
             parse(arg) # make sure syntax errors are confined
             return '(%s)' % arg
         elif c == 'n':
             return quote(node.hex(arg))
         elif c == 'b':
             return quote(arg.branch())
 
     def listexp(s, t):
         l = len(s)
         if l == 0:
             return "_list('')"
         elif l == 1:
             return argtype(t, s[0])
         elif t == 'd':
             return "_list('%s')" % "\0".join(str(int(a)) for a in s)
         elif t == 's':
             return "_list('%s')" % "\0".join(s)
         elif t == 'n':
             return "_list('%s')" % "\0".join(node.hex(a) for a in s)
         elif t == 'b':
             return "_list('%s')" % "\0".join(a.branch() for a in s)
 
         m = l // 2
         return '(%s or %s)' % (listexp(s[:m], t), listexp(s[m:], t))
 
     ret = ''
     pos = 0
     arg = 0
     while pos < len(expr):
         c = expr[pos]
         if c == '%':
             pos += 1
             d = expr[pos]
             if d == '%':
                 ret += d
             elif d in 'dsnbr':
                 ret += argtype(d, args[arg])
                 arg += 1
             elif d == 'l':
                 # a list of some type
                 pos += 1
                 d = expr[pos]
                 ret += listexp(list(args[arg]), d)
                 arg += 1
             else:
                 raise util.Abort('unexpected revspec format character %s' % d)
         else:
             ret += c
         pos += 1
 
     return ret
 
 def prettyformat(tree):
     def _prettyformat(tree, level, lines):
         if not isinstance(tree, tuple) or tree[0] in ('string', 'symbol'):
             lines.append((level, str(tree)))
         else:
             lines.append((level, '(%s' % tree[0]))
             for s in tree[1:]:
                 _prettyformat(s, level + 1, lines)
             lines[-1:] = [(lines[-1][0], lines[-1][1] + ')')]
 
     lines = []
     _prettyformat(tree, 0, lines)
     output = '\n'.join(('  '*l + s) for l, s in lines)
     return output
 
 # tell hggettext to extract docstrings from these functions:
 i18nfunctions = symbols.values()