merge with stable

mercurial/hook.py

@@ -1,188 +1,189 b''
 # hook.py - hook support for mercurial
 #
 # Copyright 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 i18n import _
 import os, sys
 import extensions, util
 
 def _pythonhook(ui, repo, name, hname, funcname, args, throw):
     '''call python hook. hook is callable object, looked up as
     name in python module. if callable returns "true", hook
     fails, else passes. if hook raises exception, treated as
     hook failure. exception propagates if throw is "true".
 
     reason for "true" meaning "hook failed" is so that
     unmodified commands (e.g. mercurial.commands.update) can
     be run as hooks without wrappers to convert return values.'''
 
     ui.note(_("calling hook %s: %s\n") % (hname, funcname))
     obj = funcname
     if not util.safehasattr(obj, '__call__'):
         d = funcname.rfind('.')
         if d == -1:
             raise util.Abort(_('%s hook is invalid ("%s" not in '
                                'a module)') % (hname, funcname))
         modname = funcname[:d]
         oldpaths = sys.path
         if util.mainfrozen():
             # binary installs require sys.path manipulation
             modpath, modfile = os.path.split(modname)
             if modpath and modfile:
                 sys.path = sys.path[:] + [modpath]
                 modname = modfile
         try:
             obj = __import__(modname)
         except ImportError:
             e1 = sys.exc_type, sys.exc_value, sys.exc_traceback
             try:
                 # extensions are loaded with hgext_ prefix
                 obj = __import__("hgext_%s" % modname)
             except ImportError:
                 e2 = sys.exc_type, sys.exc_value, sys.exc_traceback
                 if ui.tracebackflag:
                     ui.warn(_('exception from first failed import attempt:\n'))
                 ui.traceback(e1)
                 if ui.tracebackflag:
                     ui.warn(_('exception from second failed import attempt:\n'))
                 ui.traceback(e2)
                 raise util.Abort(_('%s hook is invalid '
                                    '(import of "%s" failed)') %
                                  (hname, modname))
         sys.path = oldpaths
         try:
             for p in funcname.split('.')[1:]:
                 obj = getattr(obj, p)
         except AttributeError:
             raise util.Abort(_('%s hook is invalid '
                                '("%s" is not defined)') %
                              (hname, funcname))
         if not util.safehasattr(obj, '__call__'):
             raise util.Abort(_('%s hook is invalid '
                                '("%s" is not callable)') %
                              (hname, funcname))
     try:
         try:
             # redirect IO descriptors to the ui descriptors so hooks
             # that write directly to these don't mess up the command
             # protocol when running through the command server
             old = sys.stdout, sys.stderr, sys.stdin
             sys.stdout, sys.stderr, sys.stdin = ui.fout, ui.ferr, ui.fin
 
             r = obj(ui=ui, repo=repo, hooktype=name, **args)
         except KeyboardInterrupt:
             raise
         except Exception, exc:
             if isinstance(exc, util.Abort):
                 ui.warn(_('error: %s hook failed: %s\n') %
                              (hname, exc.args[0]))
             else:
                 ui.warn(_('error: %s hook raised an exception: '
                                '%s\n') % (hname, exc))
             if throw:
                 raise
             ui.traceback()
             return True
     finally:
         sys.stdout, sys.stderr, sys.stdin = old
     if r:
         if throw:
             raise util.Abort(_('%s hook failed') % hname)
         ui.warn(_('warning: %s hook failed\n') % hname)
     return r
 
 def _exthook(ui, repo, name, cmd, args, throw):
     ui.note(_("running hook %s: %s\n") % (name, cmd))
 
     env = {}
     for k, v in args.iteritems():
         if util.safehasattr(v, '__call__'):
             v = v()
         if isinstance(v, dict):
             # make the dictionary element order stable across Python
             # implementations
             v = ('{' +
                  ', '.join('%r: %r' % i for i in sorted(v.iteritems())) +
                  '}')
         env['HG_' + k.upper()] = v
 
     if repo:
         cwd = repo.root
     else:
         cwd = os.getcwd()
     if 'HG_URL' in env and env['HG_URL'].startswith('remote:http'):
         r = util.system(cmd, environ=env, cwd=cwd, out=ui)
     else:
         r = util.system(cmd, environ=env, cwd=cwd, out=ui.fout)
     if r:
         desc, r = util.explainexit(r)
         if throw:
             raise util.Abort(_('%s hook %s') % (name, desc))
         ui.warn(_('warning: %s hook %s\n') % (name, desc))
     return r
 
 def _allhooks(ui):
     hooks = []
     for name, cmd in ui.configitems('hooks'):
         if not name.startswith('priority'):
             priority = ui.configint('hooks', 'priority.%s' % name, 0)
             hooks.append((-priority, len(hooks), name, cmd))
     return [(k, v) for p, o, k, v in sorted(hooks)]
 
 _redirect = False
 def redirect(state):
     global _redirect
     _redirect = state
 
 def hook(ui, repo, name, throw=False, **args):
     if not ui.callhooks:
         return False
 
     r = False
+    oldstdout = -1
 
-    oldstdout = -1
-    if _redirect:
+    try:
+        for hname, cmd in _allhooks(ui):
+            if hname.split('.')[0] != name or not cmd:
+                continue
+
+            if oldstdout == -1 and _redirect:
         try:
             stdoutno = sys.__stdout__.fileno()
             stderrno = sys.__stderr__.fileno()
             # temporarily redirect stdout to stderr, if possible
             if stdoutno >= 0 and stderrno >= 0:
                 sys.__stdout__.flush()
                 oldstdout = os.dup(stdoutno)
                 os.dup2(stderrno, stdoutno)
-        except AttributeError:
-            # __stdout__/__stderr__ doesn't have fileno(), it's not a real file
+                except (OSError, AttributeError):
+                    # files seem to be bogus, give up on redirecting (WSGI, etc)
             pass
 
-    try:
-        for hname, cmd in _allhooks(ui):
-            if hname.split('.')[0] != name or not cmd:
-                continue
             if util.safehasattr(cmd, '__call__'):
                 r = _pythonhook(ui, repo, name, hname, cmd, args, throw) or r
             elif cmd.startswith('python:'):
                 if cmd.count(':') >= 2:
                     path, cmd = cmd[7:].rsplit(':', 1)
                     path = util.expandpath(path)
                     if repo:
                         path = os.path.join(repo.root, path)
                     try:
                         mod = extensions.loadpath(path, 'hghook.%s' % hname)
                     except Exception:
                         ui.write(_("loading %s hook failed:\n") % hname)
                         raise
                     hookfn = getattr(mod, cmd)
                 else:
                     hookfn = cmd[7:].strip()
                 r = _pythonhook(ui, repo, name, hname, hookfn, args, throw) or r
             else:
                 r = _exthook(ui, repo, hname, cmd, args, throw) or r
     finally:
         if _redirect and oldstdout >= 0:
             os.dup2(oldstdout, stdoutno)
             os.close(oldstdout)
 
     return r

mercurial/phases.py

@@ -1,393 +1,393 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 util, error
 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 serialization 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)
 
 def hassecret(repo):
     """utility function that check if a repo have any secret changeset."""
     return bool(repo._phasecache.phaseroots[2])

mercurial/revset.py

@@ -1,1931 +1,1920 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 cl.revs(first + 1):
         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):
     '''
     Parse a revset statement into a stream of tokens
 
     Check that @ is a valid unquoted token character (issue3686):
     >>> list(tokenize("@::"))
     [('symbol', '@', 0), ('::', None, 1), ('end', None, 3)]
 
     '''
 
     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, list(repo), x)
 
     n = getset(repo, subset, y)
     if not n:
         n = getset(repo, list(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 = list(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 = list(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, list(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, list(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 bumped(repo, subset, x):
     """``bumped()``
     Mutable changesets marked as successors of public changesets.
 
     Only non-public and non-obsolete changesets can be `bumped`.
     """
     # i18n: "bumped" is a keyword
     getargs(x, 0, 0, _("bumped takes no arguments"))
     bumped = obsmod.getrevs(repo, 'bumped')
     return [r for r in subset if r in bumped]
 
 def bundle(repo, subset, x):
     """``bundle()``
     Changesets in the bundle.
 
     Bundle must be specified by the -R option."""
 
     try:
         bundlenodes = repo.changelog.bundlenodes
     except AttributeError:
         raise util.Abort(_("no bundle provided - specify with -R"))
     revs = set(repo[n].rev() for n in bundlenodes)
     return [r for r in subset if r in revs]
 
 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, list(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, list(repo), x)
     if not args:
         return []
     s = set(_revdescendants(repo, args, followfirst)) | set(args)
-
-    if len(subset) == len(repo):
-        # the passed in revisions may not exist, -1 for example
-        for arg in args:
-            if arg not in subset:
-                s.remove(arg)
-        return list(s)
-
     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, list(repo), x))
     else:
         args = set(getall(repo, list(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"))
     extincts = obsmod.getrevs(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, list(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, list(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, list(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 branchpoint(repo, subset, x):
     """``branchpoint()``
     Changesets with more than one child.
     """
     # i18n: "branchpoint" is a keyword
     getargs(x, 0, 0, _("branchpoint takes no arguments"))
     cl = repo.changelog
     if not subset:
         return []
     baserev = min(subset)
     parentscount = [0]*(len(repo) - baserev)
     for r in cl.revs(start=baserev + 1):
         for p in cl.parentrevs(r):
             if p >= baserev:
                 parentscount[p - baserev] += 1
     return [r for r in subset if (parentscount[r - baserev] > 1)]
 
 def minrev(repo, subset, x):
     """``min(set)``
     Changeset with lowest revision number in set.
     """
     os = getset(repo, list(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"))
     obsoletes = obsmod.getrevs(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, list(repo), x))
     else:
         args = set(getall(repo, list(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, list(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, list(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, list(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, repo.changelog, 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, repo.changelog, x))
-    if len(subset) == len(repo):
-        subset = s
-    else:
         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"))
     unstables = obsmod.getrevs(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,
     "branchpoint": branchpoint,
     "bumped": bumped,
     "bundle": bundle,
     "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()

mercurial/util.py

@@ -1,1805 +1,1805 b''
 # util.py - Mercurial utility functions and platform specific implementations
 #
 #  Copyright 2005 K. Thananchayan <[e-mail]>
 #  Copyright 2005-2007 Matt Mackall <[e-mail]>
 #  Copyright 2006 Vadim Gelfer <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 """Mercurial utility functions and platform specific implementations.
 
 This contains helper routines that are independent of the SCM core and
 hide platform-specific details from the core.
 """
 
 from i18n import _
 import error, osutil, encoding, collections
 import errno, re, shutil, sys, tempfile, traceback
 import os, time, datetime, calendar, textwrap, signal
 import imp, socket, urllib
 
 if os.name == 'nt':
     import windows as platform
 else:
     import posix as platform
 
 cachestat = platform.cachestat
 checkexec = platform.checkexec
 checklink = platform.checklink
 copymode = platform.copymode
 executablepath = platform.executablepath
 expandglobs = platform.expandglobs
 explainexit = platform.explainexit
 findexe = platform.findexe
 gethgcmd = platform.gethgcmd
 getuser = platform.getuser
 groupmembers = platform.groupmembers
 groupname = platform.groupname
 hidewindow = platform.hidewindow
 isexec = platform.isexec
 isowner = platform.isowner
 localpath = platform.localpath
 lookupreg = platform.lookupreg
 makedir = platform.makedir
 nlinks = platform.nlinks
 normpath = platform.normpath
 normcase = platform.normcase
 openhardlinks = platform.openhardlinks
 oslink = platform.oslink
 parsepatchoutput = platform.parsepatchoutput
 pconvert = platform.pconvert
 popen = platform.popen
 posixfile = platform.posixfile
 quotecommand = platform.quotecommand
 realpath = platform.realpath
 rename = platform.rename
 samedevice = platform.samedevice
 samefile = platform.samefile
 samestat = platform.samestat
 setbinary = platform.setbinary
 setflags = platform.setflags
 setsignalhandler = platform.setsignalhandler
 shellquote = platform.shellquote
 spawndetached = platform.spawndetached
 split = platform.split
 sshargs = platform.sshargs
 statfiles = platform.statfiles
 termwidth = platform.termwidth
 testpid = platform.testpid
 umask = platform.umask
 unlink = platform.unlink
 unlinkpath = platform.unlinkpath
 username = platform.username
 
 # Python compatibility
 
 _notset = object()
 
 def safehasattr(thing, attr):
     return getattr(thing, attr, _notset) is not _notset
 
 def sha1(s=''):
     '''
     Low-overhead wrapper around Python's SHA support
 
     >>> f = _fastsha1
     >>> a = sha1()
     >>> a = f()
     >>> a.hexdigest()
     'da39a3ee5e6b4b0d3255bfef95601890afd80709'
     '''
 
     return _fastsha1(s)
 
 def _fastsha1(s=''):
     # This function will import sha1 from hashlib or sha (whichever is
     # available) and overwrite itself with it on the first call.
     # Subsequent calls will go directly to the imported function.
     if sys.version_info >= (2, 5):
         from hashlib import sha1 as _sha1
     else:
         from sha import sha as _sha1
     global _fastsha1, sha1
     _fastsha1 = sha1 = _sha1
     return _sha1(s)
 
 try:
     buffer = buffer
 except NameError:
     if sys.version_info[0] < 3:
         def buffer(sliceable, offset=0):
             return sliceable[offset:]
     else:
         def buffer(sliceable, offset=0):
             return memoryview(sliceable)[offset:]
 
 import subprocess
 closefds = os.name == 'posix'
 
 def popen2(cmd, env=None, newlines=False):
     # Setting bufsize to -1 lets the system decide the buffer size.
     # The default for bufsize is 0, meaning unbuffered. This leads to
     # poor performance on Mac OS X: http://bugs.python.org/issue4194
     p = subprocess.Popen(cmd, shell=True, bufsize=-1,
                          close_fds=closefds,
                          stdin=subprocess.PIPE, stdout=subprocess.PIPE,
                          universal_newlines=newlines,
                          env=env)
     return p.stdin, p.stdout
 
 def popen3(cmd, env=None, newlines=False):
     p = subprocess.Popen(cmd, shell=True, bufsize=-1,
                          close_fds=closefds,
                          stdin=subprocess.PIPE, stdout=subprocess.PIPE,
                          stderr=subprocess.PIPE,
                          universal_newlines=newlines,
                          env=env)
     return p.stdin, p.stdout, p.stderr
 
 def version():
     """Return version information if available."""
     try:
         import __version__
         return __version__.version
     except ImportError:
         return 'unknown'
 
 # used by parsedate
 defaultdateformats = (
     '%Y-%m-%d %H:%M:%S',
     '%Y-%m-%d %I:%M:%S%p',
     '%Y-%m-%d %H:%M',
     '%Y-%m-%d %I:%M%p',
     '%Y-%m-%d',
     '%m-%d',
     '%m/%d',
     '%m/%d/%y',
     '%m/%d/%Y',
     '%a %b %d %H:%M:%S %Y',
     '%a %b %d %I:%M:%S%p %Y',
     '%a, %d %b %Y %H:%M:%S',        #  GNU coreutils "/bin/date --rfc-2822"
     '%b %d %H:%M:%S %Y',
     '%b %d %I:%M:%S%p %Y',
     '%b %d %H:%M:%S',
     '%b %d %I:%M:%S%p',
     '%b %d %H:%M',
     '%b %d %I:%M%p',
     '%b %d %Y',
     '%b %d',
     '%H:%M:%S',
     '%I:%M:%S%p',
     '%H:%M',
     '%I:%M%p',
 )
 
 extendeddateformats = defaultdateformats + (
     "%Y",
     "%Y-%m",
     "%b",
     "%b %Y",
     )
 
 def cachefunc(func):
     '''cache the result of function calls'''
     # XXX doesn't handle keywords args
     cache = {}
     if func.func_code.co_argcount == 1:
         # we gain a small amount of time because
         # we don't need to pack/unpack the list
         def f(arg):
             if arg not in cache:
                 cache[arg] = func(arg)
             return cache[arg]
     else:
         def f(*args):
             if args not in cache:
                 cache[args] = func(*args)
             return cache[args]
 
     return f
 
 try:
     collections.deque.remove
     deque = collections.deque
 except AttributeError:
     # python 2.4 lacks deque.remove
     class deque(collections.deque):
         def remove(self, val):
             for i, v in enumerate(self):
                 if v == val:
                     del self[i]
                     break
 
 def lrucachefunc(func):
     '''cache most recent results of function calls'''
     cache = {}
     order = deque()
     if func.func_code.co_argcount == 1:
         def f(arg):
             if arg not in cache:
                 if len(cache) > 20:
                     del cache[order.popleft()]
                 cache[arg] = func(arg)
             else:
                 order.remove(arg)
             order.append(arg)
             return cache[arg]
     else:
         def f(*args):
             if args not in cache:
                 if len(cache) > 20:
                     del cache[order.popleft()]
                 cache[args] = func(*args)
             else:
                 order.remove(args)
             order.append(args)
             return cache[args]
 
     return f
 
 class propertycache(object):
     def __init__(self, func):
         self.func = func
         self.name = func.__name__
     def __get__(self, obj, type=None):
         result = self.func(obj)
         setattr(obj, self.name, result)
         return result
 
 def pipefilter(s, cmd):
     '''filter string S through command CMD, returning its output'''
     p = subprocess.Popen(cmd, shell=True, close_fds=closefds,
                          stdin=subprocess.PIPE, stdout=subprocess.PIPE)
     pout, perr = p.communicate(s)
     return pout
 
 def tempfilter(s, cmd):
     '''filter string S through a pair of temporary files with CMD.
     CMD is used as a template to create the real command to be run,
     with the strings INFILE and OUTFILE replaced by the real names of
     the temporary files generated.'''
     inname, outname = None, None
     try:
         infd, inname = tempfile.mkstemp(prefix='hg-filter-in-')
         fp = os.fdopen(infd, 'wb')
         fp.write(s)
         fp.close()
         outfd, outname = tempfile.mkstemp(prefix='hg-filter-out-')
         os.close(outfd)
         cmd = cmd.replace('INFILE', inname)
         cmd = cmd.replace('OUTFILE', outname)
         code = os.system(cmd)
         if sys.platform == 'OpenVMS' and code & 1:
             code = 0
         if code:
             raise Abort(_("command '%s' failed: %s") %
                         (cmd, explainexit(code)))
         fp = open(outname, 'rb')
         r = fp.read()
         fp.close()
         return r
     finally:
         try:
             if inname:
                 os.unlink(inname)
         except OSError:
             pass
         try:
             if outname:
                 os.unlink(outname)
         except OSError:
             pass
 
 filtertable = {
     'tempfile:': tempfilter,
     'pipe:': pipefilter,
     }
 
 def filter(s, cmd):
     "filter a string through a command that transforms its input to its output"
     for name, fn in filtertable.iteritems():
         if cmd.startswith(name):
             return fn(s, cmd[len(name):].lstrip())
     return pipefilter(s, cmd)
 
 def binary(s):
     """return true if a string is binary data"""
     return bool(s and '\0' in s)
 
 def increasingchunks(source, min=1024, max=65536):
     '''return no less than min bytes per chunk while data remains,
     doubling min after each chunk until it reaches max'''
     def log2(x):
         if not x:
             return 0
         i = 0
         while x:
             x >>= 1
             i += 1
         return i - 1
 
     buf = []
     blen = 0
     for chunk in source:
         buf.append(chunk)
         blen += len(chunk)
         if blen >= min:
             if min < max:
                 min = min << 1
                 nmin = 1 << log2(blen)
                 if nmin > min:
                     min = nmin
                 if min > max:
                     min = max
             yield ''.join(buf)
             blen = 0
             buf = []
     if buf:
         yield ''.join(buf)
 
 Abort = error.Abort
 
 def always(fn):
     return True
 
 def never(fn):
     return False
 
 def pathto(root, n1, n2):
     '''return the relative path from one place to another.
     root should use os.sep to separate directories
     n1 should use os.sep to separate directories
     n2 should use "/" to separate directories
     returns an os.sep-separated path.
 
     If n1 is a relative path, it's assumed it's
     relative to root.
     n2 should always be relative to root.
     '''
     if not n1:
         return localpath(n2)
     if os.path.isabs(n1):
         if os.path.splitdrive(root)[0] != os.path.splitdrive(n1)[0]:
             return os.path.join(root, localpath(n2))
         n2 = '/'.join((pconvert(root), n2))
     a, b = splitpath(n1), n2.split('/')
     a.reverse()
     b.reverse()
     while a and b and a[-1] == b[-1]:
         a.pop()
         b.pop()
     b.reverse()
     return os.sep.join((['..'] * len(a)) + b) or '.'
 
 _hgexecutable = None
 
 def mainfrozen():
     """return True if we are a frozen executable.
 
     The code supports py2exe (most common, Windows only) and tools/freeze
     (portable, not much used).
     """
     return (safehasattr(sys, "frozen") or # new py2exe
             safehasattr(sys, "importers") or # old py2exe
             imp.is_frozen("__main__")) # tools/freeze
 
 def hgexecutable():
     """return location of the 'hg' executable.
 
     Defaults to $HG or 'hg' in the search path.
     """
     if _hgexecutable is None:
         hg = os.environ.get('HG')
         mainmod = sys.modules['__main__']
         if hg:
             _sethgexecutable(hg)
         elif mainfrozen():
             _sethgexecutable(sys.executable)
         elif os.path.basename(getattr(mainmod, '__file__', '')) == 'hg':
             _sethgexecutable(mainmod.__file__)
         else:
             exe = findexe('hg') or os.path.basename(sys.argv[0])
             _sethgexecutable(exe)
     return _hgexecutable
 
 def _sethgexecutable(path):
     """set location of the 'hg' executable"""
     global _hgexecutable
     _hgexecutable = path
 
 def system(cmd, environ={}, cwd=None, onerr=None, errprefix=None, out=None):
     '''enhanced shell command execution.
     run with environment maybe modified, maybe in different dir.
 
     if command fails and onerr is None, return status.  if ui object,
     print error message and return status, else raise onerr object as
     exception.
 
     if out is specified, it is assumed to be a file-like object that has a
     write() method. stdout and stderr will be redirected to out.'''
     try:
         sys.stdout.flush()
     except Exception:
         pass
     def py2shell(val):
         'convert python object into string that is useful to shell'
         if val is None or val is False:
             return '0'
         if val is True:
             return '1'
         return str(val)
     origcmd = cmd
     cmd = quotecommand(cmd)
     if sys.platform == 'plan9':
         # subprocess kludge to work around issues in half-baked Python
         # ports, notably bichued/python:
         if not cwd is None:
             os.chdir(cwd)
         rc = os.system(cmd)
     else:
         env = dict(os.environ)
         env.update((k, py2shell(v)) for k, v in environ.iteritems())
         env['HG'] = hgexecutable()
         if out is None or out == sys.__stdout__:
             rc = subprocess.call(cmd, shell=True, close_fds=closefds,
                                  env=env, cwd=cwd)
         else:
             proc = subprocess.Popen(cmd, shell=True, close_fds=closefds,
                                     env=env, cwd=cwd, stdout=subprocess.PIPE,
                                     stderr=subprocess.STDOUT)
             for line in proc.stdout:
                 out.write(line)
             proc.wait()
             rc = proc.returncode
         if sys.platform == 'OpenVMS' and rc & 1:
             rc = 0
     if rc and onerr:
         errmsg = '%s %s' % (os.path.basename(origcmd.split(None, 1)[0]),
                             explainexit(rc)[0])
         if errprefix:
             errmsg = '%s: %s' % (errprefix, errmsg)
         try:
             onerr.warn(errmsg + '\n')
         except AttributeError:
             raise onerr(errmsg)
     return rc
 
 def checksignature(func):
     '''wrap a function with code to check for calling errors'''
     def check(*args, **kwargs):
         try:
             return func(*args, **kwargs)
         except TypeError:
             if len(traceback.extract_tb(sys.exc_info()[2])) == 1:
                 raise error.SignatureError
             raise
 
     return check
 
 def copyfile(src, dest):
     "copy a file, preserving mode and atime/mtime"
     if os.path.islink(src):
         try:
             os.unlink(dest)
         except OSError:
             pass
         os.symlink(os.readlink(src), dest)
     else:
         try:
             shutil.copyfile(src, dest)
             shutil.copymode(src, dest)
         except shutil.Error, inst:
             raise Abort(str(inst))
 
 def copyfiles(src, dst, hardlink=None):
     """Copy a directory tree using hardlinks if possible"""
 
     if hardlink is None:
         hardlink = (os.stat(src).st_dev ==
                     os.stat(os.path.dirname(dst)).st_dev)
 
     num = 0
     if os.path.isdir(src):
         os.mkdir(dst)
         for name, kind in osutil.listdir(src):
             srcname = os.path.join(src, name)
             dstname = os.path.join(dst, name)
             hardlink, n = copyfiles(srcname, dstname, hardlink)
             num += n
     else:
         if hardlink:
             try:
                 oslink(src, dst)
             except (IOError, OSError):
                 hardlink = False
                 shutil.copy(src, dst)
         else:
             shutil.copy(src, dst)
         num += 1
 
     return hardlink, num
 
 _winreservednames = '''con prn aux nul
     com1 com2 com3 com4 com5 com6 com7 com8 com9
     lpt1 lpt2 lpt3 lpt4 lpt5 lpt6 lpt7 lpt8 lpt9'''.split()
 _winreservedchars = ':*?"<>|'
 def checkwinfilename(path):
     '''Check that the base-relative path is a valid filename on Windows.
     Returns None if the path is ok, or a UI string describing the problem.
 
     >>> checkwinfilename("just/a/normal/path")
     >>> checkwinfilename("foo/bar/con.xml")
     "filename contains 'con', which is reserved on Windows"
     >>> checkwinfilename("foo/con.xml/bar")
     "filename contains 'con', which is reserved on Windows"
     >>> checkwinfilename("foo/bar/xml.con")
     >>> checkwinfilename("foo/bar/AUX/bla.txt")
     "filename contains 'AUX', which is reserved on Windows"
     >>> checkwinfilename("foo/bar/bla:.txt")
     "filename contains ':', which is reserved on Windows"
     >>> checkwinfilename("foo/bar/b\07la.txt")
     "filename contains '\\\\x07', which is invalid on Windows"
     >>> checkwinfilename("foo/bar/bla ")
     "filename ends with ' ', which is not allowed on Windows"
     >>> checkwinfilename("../bar")
     '''
     for n in path.replace('\\', '/').split('/'):
         if not n:
             continue
         for c in n:
             if c in _winreservedchars:
                 return _("filename contains '%s', which is reserved "
                          "on Windows") % c
             if ord(c) <= 31:
                 return _("filename contains %r, which is invalid "
                          "on Windows") % c
         base = n.split('.')[0]
         if base and base.lower() in _winreservednames:
             return _("filename contains '%s', which is reserved "
                      "on Windows") % base
         t = n[-1]
         if t in '. ' and n not in '..':
             return _("filename ends with '%s', which is not allowed "
                      "on Windows") % t
 
 if os.name == 'nt':
     checkosfilename = checkwinfilename
 else:
     checkosfilename = platform.checkosfilename
 
 def makelock(info, pathname):
     try:
         return os.symlink(info, pathname)
     except OSError, why:
         if why.errno == errno.EEXIST:
             raise
     except AttributeError: # no symlink in os
         pass
 
     ld = os.open(pathname, os.O_CREAT | os.O_WRONLY | os.O_EXCL)
     os.write(ld, info)
     os.close(ld)
 
 def readlock(pathname):
     try:
         return os.readlink(pathname)
     except OSError, why:
         if why.errno not in (errno.EINVAL, errno.ENOSYS):
             raise
     except AttributeError: # no symlink in os
         pass
     fp = posixfile(pathname)
     r = fp.read()
     fp.close()
     return r
 
 def fstat(fp):
     '''stat file object that may not have fileno method.'''
     try:
         return os.fstat(fp.fileno())
     except AttributeError:
         return os.stat(fp.name)
 
 # File system features
 
 def checkcase(path):
     """
     Check whether the given path is on a case-sensitive filesystem
 
     Requires a path (like /foo/.hg) ending with a foldable final
     directory component.
     """
     s1 = os.stat(path)
     d, b = os.path.split(path)
     b2 = b.upper()
     if b == b2:
         b2 = b.lower()
         if b == b2:
             return True # no evidence against case sensitivity
     p2 = os.path.join(d, b2)
     try:
         s2 = os.stat(p2)
         if s2 == s1:
             return False
         return True
     except OSError:
         return True
 
 try:
     import re2
     _re2 = None
 except ImportError:
     _re2 = False
 
 def compilere(pat):
     '''Compile a regular expression, using re2 if possible
 
     For best performance, use only re2-compatible regexp features.'''
     global _re2
     if _re2 is None:
         try:
             re2.compile
             _re2 = True
         except ImportError:
             _re2 = False
     if _re2:
         try:
             return re2.compile(pat)
         except re2.error:
             pass
     return re.compile(pat)
 
 _fspathcache = {}
 def fspath(name, root):
     '''Get name in the case stored in the filesystem
 
     The name should be relative to root, and be normcase-ed for efficiency.
 
     Note that this function is unnecessary, and should not be
     called, for case-sensitive filesystems (simply because it's expensive).
 
     The root should be normcase-ed, too.
     '''
     def find(p, contents):
         for n in contents:
             if normcase(n) == p:
                 return n
         return None
 
     seps = os.sep
     if os.altsep:
         seps = seps + os.altsep
     # Protect backslashes. This gets silly very quickly.
     seps.replace('\\','\\\\')
     pattern = re.compile(r'([^%s]+)|([%s]+)' % (seps, seps))
     dir = os.path.normpath(root)
     result = []
     for part, sep in pattern.findall(name):
         if sep:
             result.append(sep)
             continue
 
         if dir not in _fspathcache:
             _fspathcache[dir] = os.listdir(dir)
         contents = _fspathcache[dir]
 
         found = find(part, contents)
         if not found:
             # retry "once per directory" per "dirstate.walk" which
             # may take place for each patches of "hg qpush", for example
             contents = os.listdir(dir)
             _fspathcache[dir] = contents
             found = find(part, contents)
 
         result.append(found or part)
         dir = os.path.join(dir, part)
 
     return ''.join(result)
 
 def checknlink(testfile):
     '''check whether hardlink count reporting works properly'''
 
     # testfile may be open, so we need a separate file for checking to
     # work around issue2543 (or testfile may get lost on Samba shares)
     f1 = testfile + ".hgtmp1"
     if os.path.lexists(f1):
         return False
     try:
         posixfile(f1, 'w').close()
     except IOError:
         return False
 
     f2 = testfile + ".hgtmp2"
     fd = None
     try:
         try:
             oslink(f1, f2)
         except OSError:
             return False
 
         # nlinks() may behave differently for files on Windows shares if
         # the file is open.
         fd = posixfile(f2)
         return nlinks(f2) > 1
     finally:
         if fd is not None:
             fd.close()
         for f in (f1, f2):
             try:
                 os.unlink(f)
             except OSError:
                 pass
 
     return False
 
 def endswithsep(path):
     '''Check path ends with os.sep or os.altsep.'''
     return path.endswith(os.sep) or os.altsep and path.endswith(os.altsep)
 
 def splitpath(path):
     '''Split path by os.sep.
     Note that this function does not use os.altsep because this is
     an alternative of simple "xxx.split(os.sep)".
     It is recommended to use os.path.normpath() before using this
     function if need.'''
     return path.split(os.sep)
 
 def gui():
     '''Are we running in a GUI?'''
     if sys.platform == 'darwin':
         if 'SSH_CONNECTION' in os.environ:
             # handle SSH access to a box where the user is logged in
             return False
         elif getattr(osutil, 'isgui', None):
             # check if a CoreGraphics session is available
             return osutil.isgui()
         else:
             # pure build; use a safe default
             return True
     else:
         return os.name == "nt" or os.environ.get("DISPLAY")
 
 def mktempcopy(name, emptyok=False, createmode=None):
     """Create a temporary file with the same contents from name
 
     The permission bits are copied from the original file.
 
     If the temporary file is going to be truncated immediately, you
     can use emptyok=True as an optimization.
 
     Returns the name of the temporary file.
     """
     d, fn = os.path.split(name)
     fd, temp = tempfile.mkstemp(prefix='.%s-' % fn, dir=d)
     os.close(fd)
     # Temporary files are created with mode 0600, which is usually not
     # what we want.  If the original file already exists, just copy
     # its mode.  Otherwise, manually obey umask.
     copymode(name, temp, createmode)
     if emptyok:
         return temp
     try:
         try:
             ifp = posixfile(name, "rb")
         except IOError, inst:
             if inst.errno == errno.ENOENT:
                 return temp
             if not getattr(inst, 'filename', None):
                 inst.filename = name
             raise
         ofp = posixfile(temp, "wb")
         for chunk in filechunkiter(ifp):
             ofp.write(chunk)
         ifp.close()
         ofp.close()
     except: # re-raises
         try: os.unlink(temp)
         except OSError: pass
         raise
     return temp
 
 class atomictempfile(object):
     '''writable file object that atomically updates a file
 
     All writes will go to a temporary copy of the original file. Call
     close() when you are done writing, and atomictempfile will rename
     the temporary copy to the original name, making the changes
     visible. If the object is destroyed without being closed, all your
     writes are discarded.
     '''
     def __init__(self, name, mode='w+b', createmode=None):
         self.__name = name      # permanent name
         self._tempname = mktempcopy(name, emptyok=('w' in mode),
                                     createmode=createmode)
         self._fp = posixfile(self._tempname, mode)
 
         # delegated methods
         self.write = self._fp.write
         self.seek = self._fp.seek
         self.tell = self._fp.tell
         self.fileno = self._fp.fileno
 
     def close(self):
         if not self._fp.closed:
             self._fp.close()
             rename(self._tempname, localpath(self.__name))
 
     def discard(self):
         if not self._fp.closed:
             try:
                 os.unlink(self._tempname)
             except OSError:
                 pass
             self._fp.close()
 
     def __del__(self):
         if safehasattr(self, '_fp'): # constructor actually did something
             self.discard()
 
 def makedirs(name, mode=None):
     """recursive directory creation with parent mode inheritance"""
     try:
         os.mkdir(name)
     except OSError, err:
         if err.errno == errno.EEXIST:
             return
         if err.errno != errno.ENOENT or not name:
             raise
         parent = os.path.dirname(os.path.abspath(name))
         if parent == name:
             raise
         makedirs(parent, mode)
         os.mkdir(name)
     if mode is not None:
         os.chmod(name, mode)
 
 def readfile(path):
     fp = open(path, 'rb')
     try:
         return fp.read()
     finally:
         fp.close()
 
 def writefile(path, text):
     fp = open(path, 'wb')
     try:
         fp.write(text)
     finally:
         fp.close()
 
 def appendfile(path, text):
     fp = open(path, 'ab')
     try:
         fp.write(text)
     finally:
         fp.close()
 
 class chunkbuffer(object):
     """Allow arbitrary sized chunks of data to be efficiently read from an
     iterator over chunks of arbitrary size."""
 
     def __init__(self, in_iter):
         """in_iter is the iterator that's iterating over the input chunks.
         targetsize is how big a buffer to try to maintain."""
         def splitbig(chunks):
             for chunk in chunks:
                 if len(chunk) > 2**20:
                     pos = 0
                     while pos < len(chunk):
                         end = pos + 2 ** 18
                         yield chunk[pos:end]
                         pos = end
                 else:
                     yield chunk
         self.iter = splitbig(in_iter)
         self._queue = deque()
 
     def read(self, l):
         """Read L bytes of data from the iterator of chunks of data.
         Returns less than L bytes if the iterator runs dry."""
         left = l
-        buf = ''
+        buf = []
         queue = self._queue
         while left > 0:
             # refill the queue
             if not queue:
                 target = 2**18
                 for chunk in self.iter:
                     queue.append(chunk)
                     target -= len(chunk)
                     if target <= 0:
                         break
                 if not queue:
                     break
 
             chunk = queue.popleft()
             left -= len(chunk)
             if left < 0:
                 queue.appendleft(chunk[left:])
-                buf += chunk[:left]
+                buf.append(chunk[:left])
             else:
-                buf += chunk
+                buf.append(chunk)
 
-        return buf
+        return ''.join(buf)
 
 def filechunkiter(f, size=65536, limit=None):
     """Create a generator that produces the data in the file size
     (default 65536) bytes at a time, up to optional limit (default is
     to read all data).  Chunks may be less than size bytes if the
     chunk is the last chunk in the file, or the file is a socket or
     some other type of file that sometimes reads less data than is
     requested."""
     assert size >= 0
     assert limit is None or limit >= 0
     while True:
         if limit is None:
             nbytes = size
         else:
             nbytes = min(limit, size)
         s = nbytes and f.read(nbytes)
         if not s:
             break
         if limit:
             limit -= len(s)
         yield s
 
 def makedate():
     ct = time.time()
     if ct < 0:
         hint = _("check your clock")
         raise Abort(_("negative timestamp: %d") % ct, hint=hint)
     delta = (datetime.datetime.utcfromtimestamp(ct) -
              datetime.datetime.fromtimestamp(ct))
     tz = delta.days * 86400 + delta.seconds
     return ct, tz
 
 def datestr(date=None, format='%a %b %d %H:%M:%S %Y %1%2'):
     """represent a (unixtime, offset) tuple as a localized time.
     unixtime is seconds since the epoch, and offset is the time zone's
     number of seconds away from UTC. if timezone is false, do not
     append time zone to string."""
     t, tz = date or makedate()
     if t < 0:
         t = 0   # time.gmtime(lt) fails on Windows for lt < -43200
         tz = 0
     if "%1" in format or "%2" in format:
         sign = (tz > 0) and "-" or "+"
         minutes = abs(tz) // 60
         format = format.replace("%1", "%c%02d" % (sign, minutes // 60))
         format = format.replace("%2", "%02d" % (minutes % 60))
     try:
         t = time.gmtime(float(t) - tz)
     except ValueError:
         # time was out of range
         t = time.gmtime(sys.maxint)
     s = time.strftime(format, t)
     return s
 
 def shortdate(date=None):
     """turn (timestamp, tzoff) tuple into iso 8631 date."""
     return datestr(date, format='%Y-%m-%d')
 
 def strdate(string, format, defaults=[]):
     """parse a localized time string and return a (unixtime, offset) tuple.
     if the string cannot be parsed, ValueError is raised."""
     def timezone(string):
         tz = string.split()[-1]
         if tz[0] in "+-" and len(tz) == 5 and tz[1:].isdigit():
             sign = (tz[0] == "+") and 1 or -1
             hours = int(tz[1:3])
             minutes = int(tz[3:5])
             return -sign * (hours * 60 + minutes) * 60
         if tz == "GMT" or tz == "UTC":
             return 0
         return None
 
     # NOTE: unixtime = localunixtime + offset
     offset, date = timezone(string), string
     if offset is not None:
         date = " ".join(string.split()[:-1])
 
     # add missing elements from defaults
     usenow = False # default to using biased defaults
     for part in ("S", "M", "HI", "d", "mb", "yY"): # decreasing specificity
         found = [True for p in part if ("%"+p) in format]
         if not found:
             date += "@" + defaults[part][usenow]
             format += "@%" + part[0]
         else:
             # We've found a specific time element, less specific time
             # elements are relative to today
             usenow = True
 
     timetuple = time.strptime(date, format)
     localunixtime = int(calendar.timegm(timetuple))
     if offset is None:
         # local timezone
         unixtime = int(time.mktime(timetuple))
         offset = unixtime - localunixtime
     else:
         unixtime = localunixtime + offset
     return unixtime, offset
 
 def parsedate(date, formats=None, bias={}):
     """parse a localized date/time and return a (unixtime, offset) tuple.
 
     The date may be a "unixtime offset" string or in one of the specified
     formats. If the date already is a (unixtime, offset) tuple, it is returned.
     """
     if not date:
         return 0, 0
     if isinstance(date, tuple) and len(date) == 2:
         return date
     if not formats:
         formats = defaultdateformats
     date = date.strip()
     try:
         when, offset = map(int, date.split(' '))
     except ValueError:
         # fill out defaults
         now = makedate()
         defaults = {}
         for part in ("d", "mb", "yY", "HI", "M", "S"):
             # this piece is for rounding the specific end of unknowns
             b = bias.get(part)
             if b is None:
                 if part[0] in "HMS":
                     b = "00"
                 else:
                     b = "0"
 
             # this piece is for matching the generic end to today's date
             n = datestr(now, "%" + part[0])
 
             defaults[part] = (b, n)
 
         for format in formats:
             try:
                 when, offset = strdate(date, format, defaults)
             except (ValueError, OverflowError):
                 pass
             else:
                 break
         else:
             raise Abort(_('invalid date: %r') % date)
     # validate explicit (probably user-specified) date and
     # time zone offset. values must fit in signed 32 bits for
     # current 32-bit linux runtimes. timezones go from UTC-12
     # to UTC+14
     if abs(when) > 0x7fffffff:
         raise Abort(_('date exceeds 32 bits: %d') % when)
     if when < 0:
         raise Abort(_('negative date value: %d') % when)
     if offset < -50400 or offset > 43200:
         raise Abort(_('impossible time zone offset: %d') % offset)
     return when, offset
 
 def matchdate(date):
     """Return a function that matches a given date match specifier
 
     Formats include:
 
     '{date}' match a given date to the accuracy provided
 
     '<{date}' on or before a given date
 
     '>{date}' on or after a given date
 
     >>> p1 = parsedate("10:29:59")
     >>> p2 = parsedate("10:30:00")
     >>> p3 = parsedate("10:30:59")
     >>> p4 = parsedate("10:31:00")
     >>> p5 = parsedate("Sep 15 10:30:00 1999")
     >>> f = matchdate("10:30")
     >>> f(p1[0])
     False
     >>> f(p2[0])
     True
     >>> f(p3[0])
     True
     >>> f(p4[0])
     False
     >>> f(p5[0])
     False
     """
 
     def lower(date):
         d = dict(mb="1", d="1")
         return parsedate(date, extendeddateformats, d)[0]
 
     def upper(date):
         d = dict(mb="12", HI="23", M="59", S="59")
         for days in ("31", "30", "29"):
             try:
                 d["d"] = days
                 return parsedate(date, extendeddateformats, d)[0]
             except Abort:
                 pass
         d["d"] = "28"
         return parsedate(date, extendeddateformats, d)[0]
 
     date = date.strip()
 
     if not date:
         raise Abort(_("dates cannot consist entirely of whitespace"))
     elif date[0] == "<":
         if not date[1:]:
             raise Abort(_("invalid day spec, use '<DATE'"))
         when = upper(date[1:])
         return lambda x: x <= when
     elif date[0] == ">":
         if not date[1:]:
             raise Abort(_("invalid day spec, use '>DATE'"))
         when = lower(date[1:])
         return lambda x: x >= when
     elif date[0] == "-":
         try:
             days = int(date[1:])
         except ValueError:
             raise Abort(_("invalid day spec: %s") % date[1:])
         if days < 0:
             raise Abort(_("%s must be nonnegative (see 'hg help dates')")
                 % date[1:])
         when = makedate()[0] - days * 3600 * 24
         return lambda x: x >= when
     elif " to " in date:
         a, b = date.split(" to ")
         start, stop = lower(a), upper(b)
         return lambda x: x >= start and x <= stop
     else:
         start, stop = lower(date), upper(date)
         return lambda x: x >= start and x <= stop
 
 def shortuser(user):
     """Return a short representation of a user name or email address."""
     f = user.find('@')
     if f >= 0:
         user = user[:f]
     f = user.find('<')
     if f >= 0:
         user = user[f + 1:]
     f = user.find(' ')
     if f >= 0:
         user = user[:f]
     f = user.find('.')
     if f >= 0:
         user = user[:f]
     return user
 
 def emailuser(user):
     """Return the user portion of an email address."""
     f = user.find('@')
     if f >= 0:
         user = user[:f]
     f = user.find('<')
     if f >= 0:
         user = user[f + 1:]
     return user
 
 def email(author):
     '''get email of author.'''
     r = author.find('>')
     if r == -1:
         r = None
     return author[author.find('<') + 1:r]
 
 def _ellipsis(text, maxlength):
     if len(text) <= maxlength:
         return text, False
     else:
         return "%s..." % (text[:maxlength - 3]), True
 
 def ellipsis(text, maxlength=400):
     """Trim string to at most maxlength (default: 400) characters."""
     try:
         # use unicode not to split at intermediate multi-byte sequence
         utext, truncated = _ellipsis(text.decode(encoding.encoding),
                                      maxlength)
         if not truncated:
             return text
         return utext.encode(encoding.encoding)
     except (UnicodeDecodeError, UnicodeEncodeError):
         return _ellipsis(text, maxlength)[0]
 
 _byteunits = (
     (100, 1 << 30, _('%.0f GB')),
     (10, 1 << 30, _('%.1f GB')),
     (1, 1 << 30, _('%.2f GB')),
     (100, 1 << 20, _('%.0f MB')),
     (10, 1 << 20, _('%.1f MB')),
     (1, 1 << 20, _('%.2f MB')),
     (100, 1 << 10, _('%.0f KB')),
     (10, 1 << 10, _('%.1f KB')),
     (1, 1 << 10, _('%.2f KB')),
     (1, 1, _('%.0f bytes')),
     )
 
 def bytecount(nbytes):
     '''return byte count formatted as readable string, with units'''
 
     for multiplier, divisor, format in _byteunits:
         if nbytes >= divisor * multiplier:
             return format % (nbytes / float(divisor))
     return _byteunits[-1][2] % nbytes
 
 def uirepr(s):
     # Avoid double backslash in Windows path repr()
     return repr(s).replace('\\\\', '\\')
 
 # delay import of textwrap
 def MBTextWrapper(**kwargs):
     class tw(textwrap.TextWrapper):
         """
         Extend TextWrapper for width-awareness.
 
         Neither number of 'bytes' in any encoding nor 'characters' is
         appropriate to calculate terminal columns for specified string.
 
         Original TextWrapper implementation uses built-in 'len()' directly,
         so overriding is needed to use width information of each characters.
 
         In addition, characters classified into 'ambiguous' width are
         treated as wide in East Asian area, but as narrow in other.
 
         This requires use decision to determine width of such characters.
         """
         def __init__(self, **kwargs):
             textwrap.TextWrapper.__init__(self, **kwargs)
 
             # for compatibility between 2.4 and 2.6
             if getattr(self, 'drop_whitespace', None) is None:
                 self.drop_whitespace = kwargs.get('drop_whitespace', True)
 
         def _cutdown(self, ucstr, space_left):
             l = 0
             colwidth = encoding.ucolwidth
             for i in xrange(len(ucstr)):
                 l += colwidth(ucstr[i])
                 if space_left < l:
                     return (ucstr[:i], ucstr[i:])
             return ucstr, ''
 
         # overriding of base class
         def _handle_long_word(self, reversed_chunks, cur_line, cur_len, width):
             space_left = max(width - cur_len, 1)
 
             if self.break_long_words:
                 cut, res = self._cutdown(reversed_chunks[-1], space_left)
                 cur_line.append(cut)
                 reversed_chunks[-1] = res
             elif not cur_line:
                 cur_line.append(reversed_chunks.pop())
 
         # this overriding code is imported from TextWrapper of python 2.6
         # to calculate columns of string by 'encoding.ucolwidth()'
         def _wrap_chunks(self, chunks):
             colwidth = encoding.ucolwidth
 
             lines = []
             if self.width <= 0:
                 raise ValueError("invalid width %r (must be > 0)" % self.width)
 
             # Arrange in reverse order so items can be efficiently popped
             # from a stack of chucks.
             chunks.reverse()
 
             while chunks:
 
                 # Start the list of chunks that will make up the current line.
                 # cur_len is just the length of all the chunks in cur_line.
                 cur_line = []
                 cur_len = 0
 
                 # Figure out which static string will prefix this line.
                 if lines:
                     indent = self.subsequent_indent
                 else:
                     indent = self.initial_indent
 
                 # Maximum width for this line.
                 width = self.width - len(indent)
 
                 # First chunk on line is whitespace -- drop it, unless this
                 # is the very beginning of the text (i.e. no lines started yet).
                 if self.drop_whitespace and chunks[-1].strip() == '' and lines:
                     del chunks[-1]
 
                 while chunks:
                     l = colwidth(chunks[-1])
 
                     # Can at least squeeze this chunk onto the current line.
                     if cur_len + l <= width:
                         cur_line.append(chunks.pop())
                         cur_len += l
 
                     # Nope, this line is full.
                     else:
                         break
 
                 # The current line is full, and the next chunk is too big to
                 # fit on *any* line (not just this one).
                 if chunks and colwidth(chunks[-1]) > width:
                     self._handle_long_word(chunks, cur_line, cur_len, width)
 
                 # If the last chunk on this line is all whitespace, drop it.
                 if (self.drop_whitespace and
                     cur_line and cur_line[-1].strip() == ''):
                     del cur_line[-1]
 
                 # Convert current line back to a string and store it in list
                 # of all lines (return value).
                 if cur_line:
                     lines.append(indent + ''.join(cur_line))
 
             return lines
 
     global MBTextWrapper
     MBTextWrapper = tw
     return tw(**kwargs)
 
 def wrap(line, width, initindent='', hangindent=''):
     maxindent = max(len(hangindent), len(initindent))
     if width <= maxindent:
         # adjust for weird terminal size
         width = max(78, maxindent + 1)
     line = line.decode(encoding.encoding, encoding.encodingmode)
     initindent = initindent.decode(encoding.encoding, encoding.encodingmode)
     hangindent = hangindent.decode(encoding.encoding, encoding.encodingmode)
     wrapper = MBTextWrapper(width=width,
                             initial_indent=initindent,
                             subsequent_indent=hangindent)
     return wrapper.fill(line).encode(encoding.encoding)
 
 def iterlines(iterator):
     for chunk in iterator:
         for line in chunk.splitlines():
             yield line
 
 def expandpath(path):
     return os.path.expanduser(os.path.expandvars(path))
 
 def hgcmd():
     """Return the command used to execute current hg
 
     This is different from hgexecutable() because on Windows we want
     to avoid things opening new shell windows like batch files, so we
     get either the python call or current executable.
     """
     if mainfrozen():
         return [sys.executable]
     return gethgcmd()
 
 def rundetached(args, condfn):
     """Execute the argument list in a detached process.
 
     condfn is a callable which is called repeatedly and should return
     True once the child process is known to have started successfully.
     At this point, the child process PID is returned. If the child
     process fails to start or finishes before condfn() evaluates to
     True, return -1.
     """
     # Windows case is easier because the child process is either
     # successfully starting and validating the condition or exiting
     # on failure. We just poll on its PID. On Unix, if the child
     # process fails to start, it will be left in a zombie state until
     # the parent wait on it, which we cannot do since we expect a long
     # running process on success. Instead we listen for SIGCHLD telling
     # us our child process terminated.
     terminated = set()
     def handler(signum, frame):
         terminated.add(os.wait())
     prevhandler = None
     SIGCHLD = getattr(signal, 'SIGCHLD', None)
     if SIGCHLD is not None:
         prevhandler = signal.signal(SIGCHLD, handler)
     try:
         pid = spawndetached(args)
         while not condfn():
             if ((pid in terminated or not testpid(pid))
                 and not condfn()):
                 return -1
             time.sleep(0.1)
         return pid
     finally:
         if prevhandler is not None:
             signal.signal(signal.SIGCHLD, prevhandler)
 
 try:
     any, all = any, all
 except NameError:
     def any(iterable):
         for i in iterable:
             if i:
                 return True
         return False
 
     def all(iterable):
         for i in iterable:
             if not i:
                 return False
         return True
 
 def interpolate(prefix, mapping, s, fn=None, escape_prefix=False):
     """Return the result of interpolating items in the mapping into string s.
 
     prefix is a single character string, or a two character string with
     a backslash as the first character if the prefix needs to be escaped in
     a regular expression.
 
     fn is an optional function that will be applied to the replacement text
     just before replacement.
 
     escape_prefix is an optional flag that allows using doubled prefix for
     its escaping.
     """
     fn = fn or (lambda s: s)
     patterns = '|'.join(mapping.keys())
     if escape_prefix:
         patterns += '|' + prefix
         if len(prefix) > 1:
             prefix_char = prefix[1:]
         else:
             prefix_char = prefix
         mapping[prefix_char] = prefix_char
     r = re.compile(r'%s(%s)' % (prefix, patterns))
     return r.sub(lambda x: fn(mapping[x.group()[1:]]), s)
 
 def getport(port):
     """Return the port for a given network service.
 
     If port is an integer, it's returned as is. If it's a string, it's
     looked up using socket.getservbyname(). If there's no matching
     service, util.Abort is raised.
     """
     try:
         return int(port)
     except ValueError:
         pass
 
     try:
         return socket.getservbyname(port)
     except socket.error:
         raise Abort(_("no port number associated with service '%s'") % port)
 
 _booleans = {'1': True, 'yes': True, 'true': True, 'on': True, 'always': True,
              '0': False, 'no': False, 'false': False, 'off': False,
              'never': False}
 
 def parsebool(s):
     """Parse s into a boolean.
 
     If s is not a valid boolean, returns None.
     """
     return _booleans.get(s.lower(), None)
 
 _hexdig = '0123456789ABCDEFabcdef'
 _hextochr = dict((a + b, chr(int(a + b, 16)))
                  for a in _hexdig for b in _hexdig)
 
 def _urlunquote(s):
     """Decode HTTP/HTML % encoding.
 
     >>> _urlunquote('abc%20def')
     'abc def'
     """
     res = s.split('%')
     # fastpath
     if len(res) == 1:
         return s
     s = res[0]
     for item in res[1:]:
         try:
             s += _hextochr[item[:2]] + item[2:]
         except KeyError:
             s += '%' + item
         except UnicodeDecodeError:
             s += unichr(int(item[:2], 16)) + item[2:]
     return s
 
 class url(object):
     r"""Reliable URL parser.
 
     This parses URLs and provides attributes for the following
     components:
 
     <scheme>://<user>:<passwd>@<host>:<port>/<path>?<query>#<fragment>
 
     Missing components are set to None. The only exception is
     fragment, which is set to '' if present but empty.
 
     If parsefragment is False, fragment is included in query. If
     parsequery is False, query is included in path. If both are
     False, both fragment and query are included in path.
 
     See http://www.ietf.org/rfc/rfc2396.txt for more information.
 
     Note that for backward compatibility reasons, bundle URLs do not
     take host names. That means 'bundle://../' has a path of '../'.
 
     Examples:
 
     >>> url('http://www.ietf.org/rfc/rfc2396.txt')
     <url scheme: 'http', host: 'www.ietf.org', path: 'rfc/rfc2396.txt'>
     >>> url('ssh://[::1]:2200//home/joe/repo')
     <url scheme: 'ssh', host: '[::1]', port: '2200', path: '/home/joe/repo'>
     >>> url('file:///home/joe/repo')
     <url scheme: 'file', path: '/home/joe/repo'>
     >>> url('file:///c:/temp/foo/')
     <url scheme: 'file', path: 'c:/temp/foo/'>
     >>> url('bundle:foo')
     <url scheme: 'bundle', path: 'foo'>
     >>> url('bundle://../foo')
     <url scheme: 'bundle', path: '../foo'>
     >>> url(r'c:\foo\bar')
     <url path: 'c:\\foo\\bar'>
     >>> url(r'\\blah\blah\blah')
     <url path: '\\\\blah\\blah\\blah'>
     >>> url(r'\\blah\blah\blah#baz')
     <url path: '\\\\blah\\blah\\blah', fragment: 'baz'>
 
     Authentication credentials:
 
     >>> url('ssh://joe:xyz@x/repo')
     <url scheme: 'ssh', user: 'joe', passwd: 'xyz', host: 'x', path: 'repo'>
     >>> url('ssh://joe@x/repo')
     <url scheme: 'ssh', user: 'joe', host: 'x', path: 'repo'>
 
     Query strings and fragments:
 
     >>> url('http://host/a?b#c')
     <url scheme: 'http', host: 'host', path: 'a', query: 'b', fragment: 'c'>
     >>> url('http://host/a?b#c', parsequery=False, parsefragment=False)
     <url scheme: 'http', host: 'host', path: 'a?b#c'>
     """
 
     _safechars = "!~*'()+"
     _safepchars = "/!~*'()+:"
     _matchscheme = re.compile(r'^[a-zA-Z0-9+.\-]+:').match
 
     def __init__(self, path, parsequery=True, parsefragment=True):
         # We slowly chomp away at path until we have only the path left
         self.scheme = self.user = self.passwd = self.host = None
         self.port = self.path = self.query = self.fragment = None
         self._localpath = True
         self._hostport = ''
         self._origpath = path
 
         if parsefragment and '#' in path:
             path, self.fragment = path.split('#', 1)
             if not path:
                 path = None
 
         # special case for Windows drive letters and UNC paths
         if hasdriveletter(path) or path.startswith(r'\\'):
             self.path = path
             return
 
         # For compatibility reasons, we can't handle bundle paths as
         # normal URLS
         if path.startswith('bundle:'):
             self.scheme = 'bundle'
             path = path[7:]
             if path.startswith('//'):
                 path = path[2:]
             self.path = path
             return
 
         if self._matchscheme(path):
             parts = path.split(':', 1)
             if parts[0]:
                 self.scheme, path = parts
                 self._localpath = False
 
         if not path:
             path = None
             if self._localpath:
                 self.path = ''
                 return
         else:
             if self._localpath:
                 self.path = path
                 return
 
             if parsequery and '?' in path:
                 path, self.query = path.split('?', 1)
                 if not path:
                     path = None
                 if not self.query:
                     self.query = None
 
             # // is required to specify a host/authority
             if path and path.startswith('//'):
                 parts = path[2:].split('/', 1)
                 if len(parts) > 1:
                     self.host, path = parts
                     path = path
                 else:
                     self.host = parts[0]
                     path = None
                 if not self.host:
                     self.host = None
                     # path of file:///d is /d
                     # path of file:///d:/ is d:/, not /d:/
                     if path and not hasdriveletter(path):
                         path = '/' + path
 
             if self.host and '@' in self.host:
                 self.user, self.host = self.host.rsplit('@', 1)
                 if ':' in self.user:
                     self.user, self.passwd = self.user.split(':', 1)
                 if not self.host:
                     self.host = None
 
             # Don't split on colons in IPv6 addresses without ports
             if (self.host and ':' in self.host and
                 not (self.host.startswith('[') and self.host.endswith(']'))):
                 self._hostport = self.host
                 self.host, self.port = self.host.rsplit(':', 1)
                 if not self.host:
                     self.host = None
 
             if (self.host and self.scheme == 'file' and
                 self.host not in ('localhost', '127.0.0.1', '[::1]')):
                 raise Abort(_('file:// URLs can only refer to localhost'))
 
         self.path = path
 
         # leave the query string escaped
         for a in ('user', 'passwd', 'host', 'port',
                   'path', 'fragment'):
             v = getattr(self, a)
             if v is not None:
                 setattr(self, a, _urlunquote(v))
 
     def __repr__(self):
         attrs = []
         for a in ('scheme', 'user', 'passwd', 'host', 'port', 'path',
                   'query', 'fragment'):
             v = getattr(self, a)
             if v is not None:
                 attrs.append('%s: %r' % (a, v))
         return '<url %s>' % ', '.join(attrs)
 
     def __str__(self):
         r"""Join the URL's components back into a URL string.
 
         Examples:
 
         >>> str(url('http://user:pw@host:80/c:/bob?fo:oo#ba:ar'))
         'http://user:pw@host:80/c:/bob?fo:oo#ba:ar'
         >>> str(url('http://user:pw@host:80/?foo=bar&baz=42'))
         'http://user:pw@host:80/?foo=bar&baz=42'
         >>> str(url('http://user:pw@host:80/?foo=bar%3dbaz'))
         'http://user:pw@host:80/?foo=bar%3dbaz'
         >>> str(url('ssh://user:pw@[::1]:2200//home/joe#'))
         'ssh://user:pw@[::1]:2200//home/joe#'
         >>> str(url('http://localhost:80//'))
         'http://localhost:80//'
         >>> str(url('http://localhost:80/'))
         'http://localhost:80/'
         >>> str(url('http://localhost:80'))
         'http://localhost:80/'
         >>> str(url('bundle:foo'))
         'bundle:foo'
         >>> str(url('bundle://../foo'))
         'bundle:../foo'
         >>> str(url('path'))
         'path'
         >>> str(url('file:///tmp/foo/bar'))
         'file:///tmp/foo/bar'
         >>> str(url('file:///c:/tmp/foo/bar'))
         'file:///c:/tmp/foo/bar'
         >>> print url(r'bundle:foo\bar')
         bundle:foo\bar
         """
         if self._localpath:
             s = self.path
             if self.scheme == 'bundle':
                 s = 'bundle:' + s
             if self.fragment:
                 s += '#' + self.fragment
             return s
 
         s = self.scheme + ':'
         if self.user or self.passwd or self.host:
             s += '//'
         elif self.scheme and (not self.path or self.path.startswith('/')
                               or hasdriveletter(self.path)):
             s += '//'
             if hasdriveletter(self.path):
                 s += '/'
         if self.user:
             s += urllib.quote(self.user, safe=self._safechars)
         if self.passwd:
             s += ':' + urllib.quote(self.passwd, safe=self._safechars)
         if self.user or self.passwd:
             s += '@'
         if self.host:
             if not (self.host.startswith('[') and self.host.endswith(']')):
                 s += urllib.quote(self.host)
             else:
                 s += self.host
         if self.port:
             s += ':' + urllib.quote(self.port)
         if self.host:
             s += '/'
         if self.path:
             # TODO: similar to the query string, we should not unescape the
             # path when we store it, the path might contain '%2f' = '/',
             # which we should *not* escape.
             s += urllib.quote(self.path, safe=self._safepchars)
         if self.query:
             # we store the query in escaped form.
             s += '?' + self.query
         if self.fragment is not None:
             s += '#' + urllib.quote(self.fragment, safe=self._safepchars)
         return s
 
     def authinfo(self):
         user, passwd = self.user, self.passwd
         try:
             self.user, self.passwd = None, None
             s = str(self)
         finally:
             self.user, self.passwd = user, passwd
         if not self.user:
             return (s, None)
         # authinfo[1] is passed to urllib2 password manager, and its
         # URIs must not contain credentials. The host is passed in the
         # URIs list because Python < 2.4.3 uses only that to search for
         # a password.
         return (s, (None, (s, self.host),
                     self.user, self.passwd or ''))
 
     def isabs(self):
         if self.scheme and self.scheme != 'file':
             return True # remote URL
         if hasdriveletter(self.path):
             return True # absolute for our purposes - can't be joined()
         if self.path.startswith(r'\\'):
             return True # Windows UNC path
         if self.path.startswith('/'):
             return True # POSIX-style
         return False
 
     def localpath(self):
         if self.scheme == 'file' or self.scheme == 'bundle':
             path = self.path or '/'
             # For Windows, we need to promote hosts containing drive
             # letters to paths with drive letters.
             if hasdriveletter(self._hostport):
                 path = self._hostport + '/' + self.path
             elif (self.host is not None and self.path
                   and not hasdriveletter(path)):
                 path = '/' + path
             return path
         return self._origpath
 
 def hasscheme(path):
     return bool(url(path).scheme)
 
 def hasdriveletter(path):
     return path and path[1:2] == ':' and path[0:1].isalpha()
 
 def urllocalpath(path):
     return url(path, parsequery=False, parsefragment=False).localpath()
 
 def hidepassword(u):
     '''hide user credential in a url string'''
     u = url(u)
     if u.passwd:
         u.passwd = '***'
     return str(u)
 
 def removeauth(u):
     '''remove all authentication information from a url string'''
     u = url(u)
     u.user = u.passwd = None
     return str(u)
 
 def isatty(fd):
     try:
         return fd.isatty()
     except AttributeError:
         return False

tests/test-phases.t

@@ -1,479 +1,488 b''
   $ hglog() { hg log --template "{rev} {phaseidx} {desc}\n" $*; }
   $ mkcommit() {
   >    echo "$1" > "$1"
   >    hg add "$1"
   >    message="$1"
   >    shift
   >    hg ci -m "$message" $*
   > }
 
   $ hg init initialrepo
   $ cd initialrepo
 
 Cannot change null revision phase
 
   $ hg phase --force --secret null
   abort: cannot change null revision phase
   [255]
   $ hg phase null
   -1: public
 
   $ mkcommit A
 
 New commit are draft by default
 
   $ hglog
   0 1 A
 
 Following commit are draft too
 
   $ mkcommit B
 
   $ hglog
   1 1 B
   0 1 A
 
 Draft commit are properly created over public one:
 
   $ hg phase --public .
   $ hglog
   1 0 B
   0 0 A
 
   $ mkcommit C
   $ mkcommit D
 
   $ hglog
   3 1 D
   2 1 C
   1 0 B
   0 0 A
 
 Test creating changeset as secret
 
   $ mkcommit E --config phases.new-commit='secret'
   $ hglog
   4 2 E
   3 1 D
   2 1 C
   1 0 B
   0 0 A
 
 Test the secret property is inherited
 
   $ mkcommit H
   $ hglog
   5 2 H
   4 2 E
   3 1 D
   2 1 C
   1 0 B
   0 0 A
 
 Even on merge
 
   $ hg up -q 1
   $ mkcommit "B'"
   created new head
   $ hglog
   6 1 B'
   5 2 H
   4 2 E
   3 1 D
   2 1 C
   1 0 B
   0 0 A
   $ hg merge 4 # E
   3 files updated, 0 files merged, 0 files removed, 0 files unresolved
   (branch merge, don't forget to commit)
   $ hg ci -m "merge B' and E"
   $ hglog
   7 2 merge B' and E
   6 1 B'
   5 2 H
   4 2 E
   3 1 D
   2 1 C
   1 0 B
   0 0 A
 
 Test secret changeset are not pushed
 
   $ hg init ../push-dest
   $ cat > ../push-dest/.hg/hgrc << EOF
   > [phases]
   > publish=False
   > EOF
   $ hg outgoing ../push-dest --template='{rev} {phase} {desc|firstline}\n'
   comparing with ../push-dest
   searching for changes
   0 public A
   1 public B
   2 draft C
   3 draft D
   6 draft B'
   $ hg outgoing -r 'branch(default)' ../push-dest --template='{rev} {phase} {desc|firstline}\n'
   comparing with ../push-dest
   searching for changes
   0 public A
   1 public B
   2 draft C
   3 draft D
   6 draft B'
 
   $ hg push ../push-dest -f # force because we push multiple heads
   pushing to ../push-dest
   searching for changes
   adding changesets
   adding manifests
   adding file changes
   added 5 changesets with 5 changes to 5 files (+1 heads)
   $ hglog
   7 2 merge B' and E
   6 1 B'
   5 2 H
   4 2 E
   3 1 D
   2 1 C
   1 0 B
   0 0 A
   $ cd ../push-dest
   $ hglog
   4 1 B'
   3 1 D
   2 1 C
   1 0 B
   0 0 A
 
 (Issue3303)
 Check that remote secret changeset are ignore when checking creation of remote heads
 
 We add a secret head into the push destination.  This secreat head shadow a
 visible shared between the initial repo and the push destination.
 
   $ hg up -q 4 # B'
   $ mkcommit Z --config phases.new-commit=secret
   $ hg phase .
   5: secret
 
 # We now try to push a new public changeset that descend from the common public
 # head shadowed by the remote secret head.
 
   $ cd ../initialrepo
   $ hg up -q 6 #B'
   $ mkcommit I
   created new head
   $ hg push ../push-dest
   pushing to ../push-dest
   searching for changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 1 changes to 1 files (+1 heads)
 
 :note: The "(+1 heads)" is wrong as we do not had any visible head
 
 
 Restore condition prior extra insertion.
   $ hg -q --config extensions.mq= strip .
   $ hg up -q 7
   $ cd ..
 
 Test secret changeset are not pull
 
   $ hg init pull-dest
   $ cd pull-dest
   $ hg pull ../initialrepo
   pulling from ../initialrepo
   requesting all changes
   adding changesets
   adding manifests
   adding file changes
   added 5 changesets with 5 changes to 5 files (+1 heads)
   (run 'hg heads' to see heads, 'hg merge' to merge)
   $ hglog
   4 0 B'
   3 0 D
   2 0 C
   1 0 B
   0 0 A
   $ cd ..
 
 But secret can still be bundled explicitly
 
   $ cd initialrepo
   $ hg bundle --base '4^' -r 'children(4)' ../secret-bundle.hg
   4 changesets found
   $ cd ..
 
 Test secret changeset are not cloned
 (during local clone)
 
   $ hg clone -qU initialrepo clone-dest
   $ hglog -R clone-dest
   4 0 B'
   3 0 D
   2 0 C
   1 0 B
   0 0 A
 
 Test revset
 
   $ cd initialrepo
   $ hglog -r 'public()'
   0 0 A
   1 0 B
   $ hglog -r 'draft()'
   2 1 C
   3 1 D
   6 1 B'
   $ hglog -r 'secret()'
   4 2 E
   5 2 H
   7 2 merge B' and E
 
 test that phase are displayed in log at debug level
 
   $ hg log --debug
   changeset:   7:17a481b3bccb796c0521ae97903d81c52bfee4af
   tag:         tip
   phase:       secret
   parent:      6:cf9fe039dfd67e829edf6522a45de057b5c86519
   parent:      4:a603bfb5a83e312131cebcd05353c217d4d21dde
   manifest:    7:5e724ffacba267b2ab726c91fc8b650710deaaa8
   user:        test
   date:        Thu Jan 01 00:00:00 1970 +0000
   files+:      C D E
   extra:       branch=default
   description:
   merge B' and E
   
   
   changeset:   6:cf9fe039dfd67e829edf6522a45de057b5c86519
   phase:       draft
   parent:      1:27547f69f25460a52fff66ad004e58da7ad3fb56
   parent:      -1:0000000000000000000000000000000000000000
   manifest:    6:ab8bfef2392903058bf4ebb9e7746e8d7026b27a
   user:        test
   date:        Thu Jan 01 00:00:00 1970 +0000
   files+:      B'
   extra:       branch=default
   description:
   B'
   
   
   changeset:   5:a030c6be5127abc010fcbff1851536552e6951a8
   phase:       secret
   parent:      4:a603bfb5a83e312131cebcd05353c217d4d21dde
   parent:      -1:0000000000000000000000000000000000000000
   manifest:    5:5c710aa854874fe3d5fa7192e77bdb314cc08b5a
   user:        test
   date:        Thu Jan 01 00:00:00 1970 +0000
   files+:      H
   extra:       branch=default
   description:
   H
   
   
   changeset:   4:a603bfb5a83e312131cebcd05353c217d4d21dde
   phase:       secret
   parent:      3:b3325c91a4d916bcc4cdc83ea3fe4ece46a42f6e
   parent:      -1:0000000000000000000000000000000000000000
   manifest:    4:7173fd1c27119750b959e3a0f47ed78abe75d6dc
   user:        test
   date:        Thu Jan 01 00:00:00 1970 +0000
   files+:      E
   extra:       branch=default
   description:
   E
   
   
   changeset:   3:b3325c91a4d916bcc4cdc83ea3fe4ece46a42f6e
   phase:       draft
   parent:      2:f838bfaca5c7226600ebcfd84f3c3c13a28d3757
   parent:      -1:0000000000000000000000000000000000000000
   manifest:    3:6e1f4c47ecb533ffd0c8e52cdc88afb6cd39e20c
   user:        test
   date:        Thu Jan 01 00:00:00 1970 +0000
   files+:      D
   extra:       branch=default
   description:
   D
   
   
   changeset:   2:f838bfaca5c7226600ebcfd84f3c3c13a28d3757
   phase:       draft
   parent:      1:27547f69f25460a52fff66ad004e58da7ad3fb56
   parent:      -1:0000000000000000000000000000000000000000
   manifest:    2:66a5a01817fdf5239c273802b5b7618d051c89e4
   user:        test
   date:        Thu Jan 01 00:00:00 1970 +0000
   files+:      C
   extra:       branch=default
   description:
   C
   
   
   changeset:   1:27547f69f25460a52fff66ad004e58da7ad3fb56
   parent:      0:4a2df7238c3b48766b5e22fafbb8a2f506ec8256
   parent:      -1:0000000000000000000000000000000000000000
   manifest:    1:cb5cbbc1bfbf24cc34b9e8c16914e9caa2d2a7fd
   user:        test
   date:        Thu Jan 01 00:00:00 1970 +0000
   files+:      B
   extra:       branch=default
   description:
   B
   
   
   changeset:   0:4a2df7238c3b48766b5e22fafbb8a2f506ec8256
   parent:      -1:0000000000000000000000000000000000000000
   parent:      -1:0000000000000000000000000000000000000000
   manifest:    0:007d8c9d88841325f5c6b06371b35b4e8a2b1a83
   user:        test
   date:        Thu Jan 01 00:00:00 1970 +0000
   files+:      A
   extra:       branch=default
   description:
   A
   
   
 
+(Issue3707)
+test invalid phase name
+
+  $ mkcommit I --config phases.new-commit='babar'
+  transaction abort!
+  rollback completed
+  abort: phases.new-commit: not a valid phase name ('babar')
+  [255]
+  
 Test phase command
 ===================
 
 initial picture
 
   $ cat >> $HGRCPATH << EOF
   > [extensions]
   > hgext.graphlog=
   > EOF
   $ hg log -G --template "{rev} {phase} {desc}\n"
   @    7 secret merge B' and E
   |\
   | o  6 draft B'
   | |
   +---o  5 secret H
   | |
   o |  4 secret E
   | |
   o |  3 draft D
   | |
   o |  2 draft C
   |/
   o  1 public B
   |
   o  0 public A
   
 
 display changesets phase
 
 (mixing -r and plain rev specification)
 
   $ hg phase 1::4 -r 7
   1: public
   2: draft
   3: draft
   4: secret
   7: secret
 
 
 move changeset forward
 
 (with -r option)
 
   $ hg phase --public -r 2
   $ hg log -G --template "{rev} {phase} {desc}\n"
   @    7 secret merge B' and E
   |\
   | o  6 draft B'
   | |
   +---o  5 secret H
   | |
   o |  4 secret E
   | |
   o |  3 draft D
   | |
   o |  2 public C
   |/
   o  1 public B
   |
   o  0 public A
   
 
 move changeset backward
 
 (without -r option)
 
   $ hg phase --draft --force 2
   $ hg log -G --template "{rev} {phase} {desc}\n"
   @    7 secret merge B' and E
   |\
   | o  6 draft B'
   | |
   +---o  5 secret H
   | |
   o |  4 secret E
   | |
   o |  3 draft D
   | |
   o |  2 draft C
   |/
   o  1 public B
   |
   o  0 public A
   
 
 move changeset forward and backward
 
   $ hg phase --draft --force 1::4
   $ hg log -G --template "{rev} {phase} {desc}\n"
   @    7 secret merge B' and E
   |\
   | o  6 draft B'
   | |
   +---o  5 secret H
   | |
   o |  4 draft E
   | |
   o |  3 draft D
   | |
   o |  2 draft C
   |/
   o  1 draft B
   |
   o  0 public A
   
 test partial failure
 
   $ hg phase --public 7
   $ hg phase --draft '5 or 7'
   cannot move 1 changesets to a more permissive phase, use --force
   phase changed for 1 changesets
   [1]
   $ hg log -G --template "{rev} {phase} {desc}\n"
   @    7 public merge B' and E
   |\
   | o  6 public B'
   | |
   +---o  5 draft H
   | |
   o |  4 public E
   | |
   o |  3 public D
   | |
   o |  2 public C
   |/
   o  1 public B
   |
   o  0 public A
   
 
 test complete failure
 
   $ hg phase --draft 7
   cannot move 1 changesets to a more permissive phase, use --force
   no phases changed
   [1]
 
   $ cd ..