smartset: extract method to slice abstractsmartset

Sub classes can provide optimized implementations.

mercurial/revset.py

@@ -1,2357 +1,2339
 # 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.
 
 from __future__ import absolute_import
 
 import heapq
 import re
 
 from .i18n import _
 from . import (
     destutil,
     encoding,
     error,
     hbisect,
     match as matchmod,
     node,
     obsolete as obsmod,
     pathutil,
     phases,
     registrar,
     repoview,
     revsetlang,
     scmutil,
     smartset,
     util,
 )
 
 # helpers for processing parsed tree
 getsymbol = revsetlang.getsymbol
 getstring = revsetlang.getstring
 getinteger = revsetlang.getinteger
 getboolean = revsetlang.getboolean
 getlist = revsetlang.getlist
 getrange = revsetlang.getrange
 getargs = revsetlang.getargs
 getargsdict = revsetlang.getargsdict
 
 # constants used as an argument of match() and matchany()
 anyorder = revsetlang.anyorder
 defineorder = revsetlang.defineorder
 followorder = revsetlang.followorder
 
 baseset = smartset.baseset
 generatorset = smartset.generatorset
 spanset = smartset.spanset
 fullreposet = smartset.fullreposet
 
 def _revancestors(repo, revs, followfirst):
     """Like revlog.ancestors(), but supports followfirst."""
     if followfirst:
         cut = 1
     else:
         cut = None
     cl = repo.changelog
 
     def iterate():
         revs.sort(reverse=True)
         irevs = iter(revs)
         h = []
 
         inputrev = next(irevs, None)
         if inputrev is not None:
             heapq.heappush(h, -inputrev)
 
         seen = set()
         while h:
             current = -heapq.heappop(h)
             if current == inputrev:
                 inputrev = next(irevs, None)
                 if inputrev is not None:
                     heapq.heappush(h, -inputrev)
             if current not in seen:
                 seen.add(current)
                 yield current
                 try:
                     for parent in cl.parentrevs(current)[:cut]:
                         if parent != node.nullrev:
                             heapq.heappush(h, -parent)
                 except error.WdirUnsupported:
                     for parent in repo[current].parents()[:cut]:
                         if parent.rev() != node.nullrev:
                             heapq.heappush(h, -parent.rev())
 
     return generatorset(iterate(), iterasc=False)
 
 def _revdescendants(repo, revs, followfirst):
     """Like revlog.descendants() but supports followfirst."""
     if followfirst:
         cut = 1
     else:
         cut = None
 
     def iterate():
         cl = repo.changelog
         # XXX this should be 'parentset.min()' assuming 'parentset' is a
         # smartset (and if it is not, it should.)
         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
         else:
             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
 
     return generatorset(iterate(), iterasc=True)
 
 def _reachablerootspure(repo, minroot, roots, heads, includepath):
     """return (heads(::<roots> and ::<heads>))
 
     If includepath is True, return (<roots>::<heads>)."""
     if not roots:
         return []
     parentrevs = repo.changelog.parentrevs
     roots = set(roots)
     visit = list(heads)
     reachable = set()
     seen = {}
     # prefetch all the things! (because python is slow)
     reached = reachable.add
     dovisit = visit.append
     nextvisit = visit.pop
     # open-code the post-order traversal due to the tiny size of
     # sys.getrecursionlimit()
     while visit:
         rev = nextvisit()
         if rev in roots:
             reached(rev)
             if not includepath:
                 continue
         parents = parentrevs(rev)
         seen[rev] = parents
         for parent in parents:
             if parent >= minroot and parent not in seen:
                 dovisit(parent)
     if not reachable:
         return baseset()
     if not includepath:
         return reachable
     for rev in sorted(seen):
         for parent in seen[rev]:
             if parent in reachable:
                 reached(rev)
     return reachable
 
 def reachableroots(repo, roots, heads, includepath=False):
     """return (heads(::<roots> and ::<heads>))
 
     If includepath is True, return (<roots>::<heads>)."""
     if not roots:
         return baseset()
     minroot = roots.min()
     roots = list(roots)
     heads = list(heads)
     try:
         revs = repo.changelog.reachableroots(minroot, heads, roots, includepath)
     except AttributeError:
         revs = _reachablerootspure(repo, minroot, roots, heads, includepath)
     revs = baseset(revs)
     revs.sort()
     return revs
 
 # helpers
 
 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 = scmutil.intrev(repo[x])
     if (x in subset
         or x == node.nullrev and isinstance(subset, fullreposet)):
         return baseset([x])
     return baseset()
 
 def rangeset(repo, subset, x, y, order):
     m = getset(repo, fullreposet(repo), x)
     n = getset(repo, fullreposet(repo), y)
 
     if not m or not n:
         return baseset()
     return _makerangeset(repo, subset, m.first(), n.last(), order)
 
 def rangeall(repo, subset, x, order):
     assert x is None
     return _makerangeset(repo, subset, 0, len(repo) - 1, order)
 
 def rangepre(repo, subset, y, order):
     # ':y' can't be rewritten to '0:y' since '0' may be hidden
     n = getset(repo, fullreposet(repo), y)
     if not n:
         return baseset()
     return _makerangeset(repo, subset, 0, n.last(), order)
 
 def rangepost(repo, subset, x, order):
     m = getset(repo, fullreposet(repo), x)
     if not m:
         return baseset()
     return _makerangeset(repo, subset, m.first(), len(repo) - 1, order)
 
 def _makerangeset(repo, subset, m, n, order):
     if m == n:
         r = baseset([m])
     elif n == node.wdirrev:
         r = spanset(repo, m, len(repo)) + baseset([n])
     elif m == node.wdirrev:
         r = baseset([m]) + spanset(repo, len(repo) - 1, n - 1)
     elif m < n:
         r = spanset(repo, m, n + 1)
     else:
         r = spanset(repo, m, n - 1)
 
     if order == defineorder:
         return r & subset
     else:
         # carrying the sorting over when possible would be more efficient
         return subset & r
 
 def dagrange(repo, subset, x, y, order):
     r = fullreposet(repo)
     xs = reachableroots(repo, getset(repo, r, x), getset(repo, r, y),
                          includepath=True)
     return subset & xs
 
 def andset(repo, subset, x, y, order):
     return getset(repo, getset(repo, subset, x), y)
 
 def differenceset(repo, subset, x, y, order):
     return getset(repo, subset, x) - getset(repo, subset, y)
 
 def _orsetlist(repo, subset, xs):
     assert xs
     if len(xs) == 1:
         return getset(repo, subset, xs[0])
     p = len(xs) // 2
     a = _orsetlist(repo, subset, xs[:p])
     b = _orsetlist(repo, subset, xs[p:])
     return a + b
 
 def orset(repo, subset, x, order):
     xs = getlist(x)
     if order == followorder:
         # slow path to take the subset order
         return subset & _orsetlist(repo, fullreposet(repo), xs)
     else:
         return _orsetlist(repo, subset, xs)
 
 def notset(repo, subset, x, order):
     return subset - getset(repo, subset, x)
 
 def listset(repo, subset, *xs):
     raise error.ParseError(_("can't use a list in this context"),
                            hint=_('see hg help "revsets.x or y"'))
 
 def keyvaluepair(repo, subset, k, v):
     raise error.ParseError(_("can't use a key-value pair in this context"))
 
 def func(repo, subset, a, b, order):
     f = getsymbol(a)
     if f in symbols:
         func = symbols[f]
         if getattr(func, '_takeorder', False):
             return func(repo, subset, b, order)
         return func(repo, subset, b)
 
     keep = lambda fn: getattr(fn, '__doc__', None) is not None
 
     syms = [s for (s, fn) in symbols.items() if keep(fn)]
     raise error.UnknownIdentifier(f, syms)
 
 # functions
 
 # symbols are callables like:
 #   fn(repo, subset, x)
 # with:
 #   repo - current repository instance
 #   subset - of revisions to be examined
 #   x - argument in tree form
 symbols = {}
 
 # symbols which can't be used for a DoS attack for any given input
 # (e.g. those which accept regexes as plain strings shouldn't be included)
 # functions that just return a lot of changesets (like all) don't count here
 safesymbols = set()
 
 predicate = registrar.revsetpredicate()
 
 @predicate('_destupdate')
 def _destupdate(repo, subset, x):
     # experimental revset for update destination
     args = getargsdict(x, 'limit', 'clean')
     return subset & baseset([destutil.destupdate(repo, **args)[0]])
 
 @predicate('_destmerge')
 def _destmerge(repo, subset, x):
     # experimental revset for merge destination
     sourceset = None
     if x is not None:
         sourceset = getset(repo, fullreposet(repo), x)
     return subset & baseset([destutil.destmerge(repo, sourceset=sourceset)])
 
 @predicate('adds(pattern)', safe=True)
 def adds(repo, subset, x):
     """Changesets that add a file matching pattern.
 
     The pattern without explicit kind like ``glob:`` is expected to be
     relative to the current directory and match against a file or a
     directory.
     """
     # i18n: "adds" is a keyword
     pat = getstring(x, _("adds requires a pattern"))
     return checkstatus(repo, subset, pat, 1)
 
 @predicate('ancestor(*changeset)', safe=True)
 def ancestor(repo, subset, x):
     """A greatest common ancestor of the changesets.
 
     Accepts 0 or more changesets.
     Will return empty list when passed no args.
     Greatest common ancestor of a single changeset is that changeset.
     """
     # i18n: "ancestor" is a keyword
     l = getlist(x)
     rl = fullreposet(repo)
     anc = None
 
     # (getset(repo, rl, i) for i in l) generates a list of lists
     for revs in (getset(repo, rl, i) for i in l):
         for r in revs:
             if anc is None:
                 anc = repo[r]
             else:
                 anc = anc.ancestor(repo[r])
 
     if anc is not None and anc.rev() in subset:
         return baseset([anc.rev()])
     return baseset()
 
 def _ancestors(repo, subset, x, followfirst=False):
     heads = getset(repo, fullreposet(repo), x)
     if not heads:
         return baseset()
     s = _revancestors(repo, heads, followfirst)
     return subset & s
 
 @predicate('ancestors(set)', safe=True)
 def ancestors(repo, subset, x):
     """Changesets that are ancestors of a changeset in set.
     """
     return _ancestors(repo, subset, x)
 
 @predicate('_firstancestors', safe=True)
 def _firstancestors(repo, subset, x):
     # ``_firstancestors(set)``
     # Like ``ancestors(set)`` but follows only the first parents.
     return _ancestors(repo, subset, x, followfirst=True)
 
 def _childrenspec(repo, subset, x, n, order):
     """Changesets that are the Nth child of a changeset
     in set.
     """
     cs = set()
     for r in getset(repo, fullreposet(repo), x):
         for i in range(n):
             c = repo[r].children()
             if len(c) == 0:
                 break
             if len(c) > 1:
                 raise error.RepoLookupError(
                     _("revision in set has more than one child"))
             r = c[0]
         else:
             cs.add(r)
     return subset & cs
 
 def ancestorspec(repo, subset, x, n, order):
     """``set~n``
     Changesets that are the Nth ancestor (first parents only) of a changeset
     in set.
     """
     n = getinteger(n, _("~ expects a number"))
     if n < 0:
         # children lookup
         return _childrenspec(repo, subset, x, -n, order)
     ps = set()
     cl = repo.changelog
     for r in getset(repo, fullreposet(repo), x):
         for i in range(n):
             try:
                 r = cl.parentrevs(r)[0]
             except error.WdirUnsupported:
                 r = repo[r].parents()[0].rev()
         ps.add(r)
     return subset & ps
 
 @predicate('author(string)', safe=True)
 def author(repo, subset, x):
     """Alias for ``user(string)``.
     """
     # i18n: "author" is a keyword
     n = getstring(x, _("author requires a string"))
     kind, pattern, matcher = _substringmatcher(n, casesensitive=False)
     return subset.filter(lambda x: matcher(repo[x].user()),
                          condrepr=('<user %r>', n))
 
 @predicate('bisect(string)', safe=True)
 def bisect(repo, subset, x):
     """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 subset & state
 
 # Backward-compatibility
 # - no help entry so that we do not advertise it any more
 @predicate('bisected', safe=True)
 def bisected(repo, subset, x):
     return bisect(repo, subset, x)
 
 @predicate('bookmark([name])', safe=True)
 def bookmark(repo, subset, x):
     """The named bookmark or all bookmarks.
 
     Pattern matching is supported for `name`. See :hg:`help revisions.patterns`.
     """
     # 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 = util.stringmatcher(bm)
         bms = set()
         if kind == 'literal':
             bmrev = repo._bookmarks.get(pattern, None)
             if not bmrev:
                 raise error.RepoLookupError(_("bookmark '%s' does not exist")
                                             % pattern)
             bms.add(repo[bmrev].rev())
         else:
             matchrevs = set()
             for name, bmrev in repo._bookmarks.iteritems():
                 if matcher(name):
                     matchrevs.add(bmrev)
             if not matchrevs:
                 raise error.RepoLookupError(_("no bookmarks exist"
                                               " that match '%s'") % pattern)
             for bmrev in matchrevs:
                 bms.add(repo[bmrev].rev())
     else:
         bms = {repo[r].rev() for r in repo._bookmarks.values()}
     bms -= {node.nullrev}
     return subset & bms
 
 @predicate('branch(string or set)', safe=True)
 def branch(repo, subset, x):
     """
     All changesets belonging to the given branch or the branches of the given
     changesets.
 
     Pattern matching is supported for `string`. See
     :hg:`help revisions.patterns`.
     """
     getbi = repo.revbranchcache().branchinfo
     def getbranch(r):
         try:
             return getbi(r)[0]
         except error.WdirUnsupported:
             return repo[r].branch()
 
     try:
         b = getstring(x, '')
     except error.ParseError:
         # not a string, but another revspec, e.g. tip()
         pass
     else:
         kind, pattern, matcher = util.stringmatcher(b)
         if kind == 'literal':
             # note: falls through to the revspec case if no branch with
             # this name exists and pattern kind is not specified explicitly
             if pattern in repo.branchmap():
                 return subset.filter(lambda r: matcher(getbranch(r)),
                                      condrepr=('<branch %r>', b))
             if b.startswith('literal:'):
                 raise error.RepoLookupError(_("branch '%s' does not exist")
                                             % pattern)
         else:
             return subset.filter(lambda r: matcher(getbranch(r)),
                                  condrepr=('<branch %r>', b))
 
     s = getset(repo, fullreposet(repo), x)
     b = set()
     for r in s:
         b.add(getbranch(r))
     c = s.__contains__
     return subset.filter(lambda r: c(r) or getbranch(r) in b,
                          condrepr=lambda: '<branch %r>' % sorted(b))
 
 @predicate('bumped()', safe=True)
 def bumped(repo, subset, x):
     """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 subset & bumped
 
 @predicate('bundle()', safe=True)
 def bundle(repo, subset, x):
     """Changesets in the bundle.
 
     Bundle must be specified by the -R option."""
 
     try:
         bundlerevs = repo.changelog.bundlerevs
     except AttributeError:
         raise error.Abort(_("no bundle provided - specify with -R"))
     return subset & bundlerevs
 
 def checkstatus(repo, subset, pat, field):
     hasset = matchmod.patkind(pat) == 'set'
 
     mcache = [None]
     def matches(x):
         c = repo[x]
         if not mcache[0] or hasset:
             mcache[0] = matchmod.match(repo.root, repo.getcwd(), [pat], ctx=c)
         m = mcache[0]
         fname = None
         if not m.anypats() and len(m.files()) == 1:
             fname = m.files()[0]
         if fname is not None:
             if fname not in c.files():
                 return False
         else:
             for f in c.files():
                 if m(f):
                     break
             else:
                 return False
         files = repo.status(c.p1().node(), c.node())[field]
         if fname is not None:
             if fname in files:
                 return True
         else:
             for f in files:
                 if m(f):
                     return True
 
     return subset.filter(matches, condrepr=('<status[%r] %r>', field, pat))
 
 def _children(repo, subset, parentset):
     if not parentset:
         return baseset()
     cs = set()
     pr = repo.changelog.parentrevs
     minrev = parentset.min()
     nullrev = node.nullrev
     for r in subset:
         if r <= minrev:
             continue
         p1, p2 = pr(r)
         if p1 in parentset:
             cs.add(r)
         if p2 != nullrev and p2 in parentset:
             cs.add(r)
     return baseset(cs)
 
 @predicate('children(set)', safe=True)
 def children(repo, subset, x):
     """Child changesets of changesets in set.
     """
     s = getset(repo, fullreposet(repo), x)
     cs = _children(repo, subset, s)
     return subset & cs
 
 @predicate('closed()', safe=True)
 def closed(repo, subset, x):
     """Changeset is closed.
     """
     # i18n: "closed" is a keyword
     getargs(x, 0, 0, _("closed takes no arguments"))
     return subset.filter(lambda r: repo[r].closesbranch(),
                          condrepr='<branch closed>')
 
 @predicate('contains(pattern)')
 def contains(repo, subset, x):
     """The revision's manifest contains a file matching pattern (but might not
     modify it). See :hg:`help patterns` for information about file patterns.
 
     The pattern without explicit kind like ``glob:`` is expected to be
     relative to the current directory and match against a file exactly
     for efficiency.
     """
     # i18n: "contains" is a keyword
     pat = getstring(x, _("contains requires a pattern"))
 
     def matches(x):
         if not matchmod.patkind(pat):
             pats = pathutil.canonpath(repo.root, repo.getcwd(), pat)
             if pats in repo[x]:
                 return True
         else:
             c = repo[x]
             m = matchmod.match(repo.root, repo.getcwd(), [pat], ctx=c)
             for f in c.manifest():
                 if m(f):
                     return True
         return False
 
     return subset.filter(matches, condrepr=('<contains %r>', pat))
 
 @predicate('converted([id])', safe=True)
 def converted(repo, subset, x):
     """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 subset.filter(lambda r: _matchvalue(r),
                          condrepr=('<converted %r>', rev))
 
 @predicate('date(interval)', safe=True)
 def date(repo, subset, x):
     """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 subset.filter(lambda x: dm(repo[x].date()[0]),
                          condrepr=('<date %r>', ds))
 
 @predicate('desc(string)', safe=True)
 def desc(repo, subset, x):
     """Search commit message for string. The match is case-insensitive.
 
     Pattern matching is supported for `string`. See
     :hg:`help revisions.patterns`.
     """
     # i18n: "desc" is a keyword
     ds = getstring(x, _("desc requires a string"))
 
     kind, pattern, matcher = _substringmatcher(ds, casesensitive=False)
 
     return subset.filter(lambda r: matcher(repo[r].description()),
                          condrepr=('<desc %r>', ds))
 
 def _descendants(repo, subset, x, followfirst=False):
     roots = getset(repo, fullreposet(repo), x)
     if not roots:
         return baseset()
     s = _revdescendants(repo, roots, followfirst)
 
     # Both sets need to be ascending in order to lazily return the union
     # in the correct order.
     base = subset & roots
     desc = subset & s
     result = base + desc
     if subset.isascending():
         result.sort()
     elif subset.isdescending():
         result.sort(reverse=True)
     else:
         result = subset & result
     return result
 
 @predicate('descendants(set)', safe=True)
 def descendants(repo, subset, x):
     """Changesets which are descendants of changesets in set.
     """
     return _descendants(repo, subset, x)
 
 @predicate('_firstdescendants', safe=True)
 def _firstdescendants(repo, subset, x):
     # ``_firstdescendants(set)``
     # Like ``descendants(set)`` but follows only the first parents.
     return _descendants(repo, subset, x, followfirst=True)
 
 @predicate('destination([set])', safe=True)
 def destination(repo, subset, x):
     """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:
         sources = getset(repo, fullreposet(repo), x)
     else:
         sources = fullreposet(repo)
 
     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 arg set.
     # Even if the immediate src of r is not in the arg set, 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 sources or src in dests:
                 dests.update(lineage)
                 break
 
             r = src
             src = _getrevsource(repo, r)
 
     return subset.filter(dests.__contains__,
                          condrepr=lambda: '<destination %r>' % sorted(dests))
 
 @predicate('divergent()', safe=True)
 def divergent(repo, subset, x):
     """
     Final successors of changesets with an alternative set of final successors.
     """
     # i18n: "divergent" is a keyword
     getargs(x, 0, 0, _("divergent takes no arguments"))
     divergent = obsmod.getrevs(repo, 'divergent')
     return subset & divergent
 
 @predicate('extinct()', safe=True)
 def extinct(repo, subset, x):
     """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 subset & extincts
 
 @predicate('extra(label, [value])', safe=True)
 def extra(repo, subset, x):
     """Changesets with the given label in the extra metadata, with the given
     optional value.
 
     Pattern matching is supported for `value`. See
     :hg:`help revisions.patterns`.
     """
     args = getargsdict(x, 'extra', 'label value')
     if 'label' not in args:
         # i18n: "extra" is a keyword
         raise error.ParseError(_('extra takes at least 1 argument'))
     # i18n: "extra" is a keyword
     label = getstring(args['label'], _('first argument to extra must be '
                                        'a string'))
     value = None
 
     if 'value' in args:
         # i18n: "extra" is a keyword
         value = getstring(args['value'], _('second argument to extra must be '
                                            'a string'))
         kind, value, matcher = util.stringmatcher(value)
 
     def _matchvalue(r):
         extra = repo[r].extra()
         return label in extra and (value is None or matcher(extra[label]))
 
     return subset.filter(lambda r: _matchvalue(r),
                          condrepr=('<extra[%r] %r>', label, value))
 
 @predicate('filelog(pattern)', safe=True)
 def filelog(repo, subset, x):
     """Changesets connected to the specified filelog.
 
     For performance reasons, visits only revisions mentioned in the file-level
     filelog, rather than filtering through all changesets (much faster, but
     doesn't include deletes or duplicate changes). For a slower, more accurate
     result, use ``file()``.
 
     The pattern without explicit kind like ``glob:`` is expected to be
     relative to the current directory and match against a file exactly
     for efficiency.
 
     If some linkrev points to revisions filtered by the current repoview, we'll
     work around it to return a non-filtered value.
     """
 
     # i18n: "filelog" is a keyword
     pat = getstring(x, _("filelog requires a pattern"))
     s = set()
     cl = repo.changelog
 
     if not matchmod.patkind(pat):
         f = pathutil.canonpath(repo.root, repo.getcwd(), pat)
         files = [f]
     else:
         m = matchmod.match(repo.root, repo.getcwd(), [pat], ctx=repo[None])
         files = (f for f in repo[None] if m(f))
 
     for f in files:
         fl = repo.file(f)
         known = {}
         scanpos = 0
         for fr in list(fl):
             fn = fl.node(fr)
             if fn in known:
                 s.add(known[fn])
                 continue
 
             lr = fl.linkrev(fr)
             if lr in cl:
                 s.add(lr)
             elif scanpos is not None:
                 # lowest matching changeset is filtered, scan further
                 # ahead in changelog
                 start = max(lr, scanpos) + 1
                 scanpos = None
                 for r in cl.revs(start):
                     # minimize parsing of non-matching entries
                     if f in cl.revision(r) and f in cl.readfiles(r):
                         try:
                             # try to use manifest delta fastpath
                             n = repo[r].filenode(f)
                             if n not in known:
                                 if n == fn:
                                     s.add(r)
                                     scanpos = r
                                     break
                                 else:
                                     known[n] = r
                         except error.ManifestLookupError:
                             # deletion in changelog
                             continue
 
     return subset & s
 
 @predicate('first(set, [n])', safe=True, takeorder=True)
 def first(repo, subset, x, order):
     """An alias for limit().
     """
     return limit(repo, subset, x, order)
 
 def _follow(repo, subset, x, name, followfirst=False):
     l = getargs(x, 0, 2, _("%s takes no arguments or a pattern "
                            "and an optional revset") % name)
     c = repo['.']
     if l:
         x = getstring(l[0], _("%s expected a pattern") % name)
         rev = None
         if len(l) >= 2:
             revs = getset(repo, fullreposet(repo), l[1])
             if len(revs) != 1:
                 raise error.RepoLookupError(
                         _("%s expected one starting revision") % name)
             rev = revs.last()
             c = repo[rev]
         matcher = matchmod.match(repo.root, repo.getcwd(), [x],
                                  ctx=repo[rev], default='path')
 
         files = c.manifest().walk(matcher)
 
         s = set()
         for fname in files:
             fctx = c[fname]
             s = s.union(set(c.rev() for c in fctx.ancestors(followfirst)))
             # include the revision responsible for the most recent version
             s.add(fctx.introrev())
     else:
         s = _revancestors(repo, baseset([c.rev()]), followfirst)
 
     return subset & s
 
 @predicate('follow([pattern[, startrev]])', safe=True)
 def follow(repo, subset, x):
     """
     An alias for ``::.`` (ancestors of the working directory's first parent).
     If pattern is specified, the histories of files matching given
     pattern in the revision given by startrev are followed, including copies.
     """
     return _follow(repo, subset, x, 'follow')
 
 @predicate('_followfirst', safe=True)
 def _followfirst(repo, subset, x):
     # ``followfirst([pattern[, startrev]])``
     # Like ``follow([pattern[, startrev]])`` but follows only the first parent
     # of every revisions or files revisions.
     return _follow(repo, subset, x, '_followfirst', followfirst=True)
 
 @predicate('followlines(file, fromline:toline[, startrev=., descend=False])',
            safe=True)
 def followlines(repo, subset, x):
     """Changesets modifying `file` in line range ('fromline', 'toline').
 
     Line range corresponds to 'file' content at 'startrev' and should hence be
     consistent with file size. If startrev is not specified, working directory's
     parent is used.
 
     By default, ancestors of 'startrev' are returned. If 'descend' is True,
     descendants of 'startrev' are returned though renames are (currently) not
     followed in this direction.
     """
     from . import context  # avoid circular import issues
 
     args = getargsdict(x, 'followlines', 'file *lines startrev descend')
     if len(args['lines']) != 1:
         raise error.ParseError(_("followlines requires a line range"))
 
     rev = '.'
     if 'startrev' in args:
         revs = getset(repo, fullreposet(repo), args['startrev'])
         if len(revs) != 1:
             raise error.ParseError(
                 # i18n: "followlines" is a keyword
                 _("followlines expects exactly one revision"))
         rev = revs.last()
 
     pat = getstring(args['file'], _("followlines requires a pattern"))
     if not matchmod.patkind(pat):
         fname = pathutil.canonpath(repo.root, repo.getcwd(), pat)
     else:
         m = matchmod.match(repo.root, repo.getcwd(), [pat], ctx=repo[rev])
         files = [f for f in repo[rev] if m(f)]
         if len(files) != 1:
             # i18n: "followlines" is a keyword
             raise error.ParseError(_("followlines expects exactly one file"))
         fname = files[0]
 
     # i18n: "followlines" is a keyword
     lr = getrange(args['lines'][0], _("followlines expects a line range"))
     fromline, toline = [getinteger(a, _("line range bounds must be integers"))
                         for a in lr]
     fromline, toline = util.processlinerange(fromline, toline)
 
     fctx = repo[rev].filectx(fname)
     descend = False
     if 'descend' in args:
         descend = getboolean(args['descend'],
                              # i18n: "descend" is a keyword
                              _("descend argument must be a boolean"))
     if descend:
         rs = generatorset(
             (c.rev() for c, _linerange
              in context.blockdescendants(fctx, fromline, toline)),
             iterasc=True)
     else:
         rs = generatorset(
             (c.rev() for c, _linerange
              in context.blockancestors(fctx, fromline, toline)),
             iterasc=False)
     return subset & rs
 
 @predicate('all()', safe=True)
 def getall(repo, subset, x):
     """All changesets, the same as ``0:tip``.
     """
     # i18n: "all" is a keyword
     getargs(x, 0, 0, _("all takes no arguments"))
     return subset & spanset(repo)  # drop "null" if any
 
 @predicate('grep(regex)')
 def grep(repo, subset, x):
     """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 as e:
         raise error.ParseError(_('invalid match pattern: %s') % e)
 
     def matches(x):
         c = repo[x]
         for e in c.files() + [c.user(), c.description()]:
             if gr.search(e):
                 return True
         return False
 
     return subset.filter(matches, condrepr=('<grep %r>', gr.pattern))
 
 @predicate('_matchfiles', safe=True)
 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.
 
     l = getargs(x, 1, -1, "_matchfiles requires at least one argument")
     pats, inc, exc = [], [], []
     rev, default = None, None
     for arg in l:
         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:
                 raise error.ParseError('_matchfiles expected at most one '
                                        'revision')
             if value != '': # empty means working directory; leave rev as None
                 rev = value
         elif prefix == 'd:':
             if default is not None:
                 raise error.ParseError('_matchfiles expected at most one '
                                        'default mode')
             default = value
         else:
             raise error.ParseError('invalid _matchfiles prefix: %s' % prefix)
     if not default:
         default = 'glob'
 
     m = matchmod.match(repo.root, repo.getcwd(), pats, include=inc,
                        exclude=exc, ctx=repo[rev], default=default)
 
     # This directly read the changelog data as creating changectx for all
     # revisions is quite expensive.
     getfiles = repo.changelog.readfiles
     wdirrev = node.wdirrev
     def matches(x):
         if x == wdirrev:
             files = repo[x].files()
         else:
             files = getfiles(x)
         for f in files:
             if m(f):
                 return True
         return False
 
     return subset.filter(matches,
                          condrepr=('<matchfiles patterns=%r, include=%r '
                                    'exclude=%r, default=%r, rev=%r>',
                                    pats, inc, exc, default, rev))
 
 @predicate('file(pattern)', safe=True)
 def hasfile(repo, subset, x):
     """Changesets affecting files matched by pattern.
 
     For a faster but less accurate result, consider using ``filelog()``
     instead.
 
     This predicate uses ``glob:`` as the default kind of pattern.
     """
     # i18n: "file" is a keyword
     pat = getstring(x, _("file requires a pattern"))
     return _matchfiles(repo, subset, ('string', 'p:' + pat))
 
 @predicate('head()', safe=True)
 def head(repo, subset, x):
     """Changeset is a named branch head.
     """
     # i18n: "head" is a keyword
     getargs(x, 0, 0, _("head takes no arguments"))
     hs = set()
     cl = repo.changelog
     for ls in repo.branchmap().itervalues():
         hs.update(cl.rev(h) for h in ls)
     return subset & baseset(hs)
 
 @predicate('heads(set)', safe=True)
 def heads(repo, subset, x):
     """Members of set with no children in set.
     """
     s = getset(repo, subset, x)
     ps = parents(repo, subset, x)
     return s - ps
 
 @predicate('hidden()', safe=True)
 def hidden(repo, subset, x):
     """Hidden changesets.
     """
     # i18n: "hidden" is a keyword
     getargs(x, 0, 0, _("hidden takes no arguments"))
     hiddenrevs = repoview.filterrevs(repo, 'visible')
     return subset & hiddenrevs
 
 @predicate('keyword(string)', safe=True)
 def keyword(repo, subset, x):
     """Search commit message, user name, and names of changed files for
     string. The match is case-insensitive.
 
     For a regular expression or case sensitive search of these fields, use
     ``grep(regex)``.
     """
     # i18n: "keyword" is a keyword
     kw = encoding.lower(getstring(x, _("keyword requires a string")))
 
     def matches(r):
         c = repo[r]
         return any(kw in encoding.lower(t)
                    for t in c.files() + [c.user(), c.description()])
 
     return subset.filter(matches, condrepr=('<keyword %r>', kw))
 
 @predicate('limit(set[, n[, offset]])', safe=True, takeorder=True)
 def limit(repo, subset, x, order):
     """First n members of set, defaulting to 1, starting from offset.
     """
     args = getargsdict(x, 'limit', 'set n offset')
     if 'set' not in args:
         # i18n: "limit" is a keyword
         raise error.ParseError(_("limit requires one to three arguments"))
     # i18n: "limit" is a keyword
     lim = getinteger(args.get('n'), _("limit expects a number"), default=1)
     if lim < 0:
         raise error.ParseError(_("negative number to select"))
     # i18n: "limit" is a keyword
     ofs = getinteger(args.get('offset'), _("limit expects a number"), default=0)
     if ofs < 0:
         raise error.ParseError(_("negative offset"))
     os = getset(repo, fullreposet(repo), args['set'])
-    result = []
-    it = iter(os)
-    for x in xrange(ofs):
-        y = next(it, None)
-        if y is None:
-            break
-    for x in xrange(lim):
-        y = next(it, None)
-        if y is None:
-            break
-        result.append(y)
-    ls = baseset(result, datarepr=('<limit n=%d, offset=%d, %r>', lim, ofs, os))
+    ls = os.slice(ofs, ofs + lim)
     if order == followorder and lim > 1:
         return subset & ls
     return ls & subset
 
 @predicate('last(set, [n])', safe=True, takeorder=True)
 def last(repo, subset, x, order):
     """Last n members of set, defaulting to 1.
     """
     # i18n: "last" is a keyword
     l = getargs(x, 1, 2, _("last requires one or two arguments"))
     lim = 1
     if len(l) == 2:
         # i18n: "last" is a keyword
         lim = getinteger(l[1], _("last expects a number"))
     if lim < 0:
         raise error.ParseError(_("negative number to select"))
     os = getset(repo, fullreposet(repo), l[0])
     os.reverse()
-    result = []
-    it = iter(os)
-    for x in xrange(lim):
-        y = next(it, None)
-        if y is None:
-            break
-        result.append(y)
-    ls = baseset(result, datarepr=('<last n=%d, %r>', lim, os))
+    ls = os.slice(0, lim)
     if order == followorder and lim > 1:
         return subset & ls
     ls.reverse()
     return ls & subset
 
 @predicate('max(set)', safe=True)
 def maxrev(repo, subset, x):
     """Changeset with highest revision number in set.
     """
     os = getset(repo, fullreposet(repo), x)
     try:
         m = os.max()
         if m in subset:
             return baseset([m], datarepr=('<max %r, %r>', subset, os))
     except ValueError:
         # os.max() throws a ValueError when the collection is empty.
         # Same as python's max().
         pass
     return baseset(datarepr=('<max %r, %r>', subset, os))
 
 @predicate('merge()', safe=True)
 def merge(repo, subset, x):
     """Changeset is a merge changeset.
     """
     # i18n: "merge" is a keyword
     getargs(x, 0, 0, _("merge takes no arguments"))
     cl = repo.changelog
     return subset.filter(lambda r: cl.parentrevs(r)[1] != -1,
                          condrepr='<merge>')
 
 @predicate('branchpoint()', safe=True)
 def branchpoint(repo, subset, x):
     """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 baseset()
     # XXX this should be 'parentset.min()' assuming 'parentset' is a smartset
     # (and if it is not, it should.)
     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 subset.filter(lambda r: parentscount[r - baserev] > 1,
                          condrepr='<branchpoint>')
 
 @predicate('min(set)', safe=True)
 def minrev(repo, subset, x):
     """Changeset with lowest revision number in set.
     """
     os = getset(repo, fullreposet(repo), x)
     try:
         m = os.min()
         if m in subset:
             return baseset([m], datarepr=('<min %r, %r>', subset, os))
     except ValueError:
         # os.min() throws a ValueError when the collection is empty.
         # Same as python's min().
         pass
     return baseset(datarepr=('<min %r, %r>', subset, os))
 
 @predicate('modifies(pattern)', safe=True)
 def modifies(repo, subset, x):
     """Changesets modifying files matched by pattern.
 
     The pattern without explicit kind like ``glob:`` is expected to be
     relative to the current directory and match against a file or a
     directory.
     """
     # i18n: "modifies" is a keyword
     pat = getstring(x, _("modifies requires a pattern"))
     return checkstatus(repo, subset, pat, 0)
 
 @predicate('named(namespace)')
 def named(repo, subset, x):
     """The changesets in a given namespace.
 
     Pattern matching is supported for `namespace`. See
     :hg:`help revisions.patterns`.
     """
     # i18n: "named" is a keyword
     args = getargs(x, 1, 1, _('named requires a namespace argument'))
 
     ns = getstring(args[0],
                    # i18n: "named" is a keyword
                    _('the argument to named must be a string'))
     kind, pattern, matcher = util.stringmatcher(ns)
     namespaces = set()
     if kind == 'literal':
         if pattern not in repo.names:
             raise error.RepoLookupError(_("namespace '%s' does not exist")
                                         % ns)
         namespaces.add(repo.names[pattern])
     else:
         for name, ns in repo.names.iteritems():
             if matcher(name):
                 namespaces.add(ns)
         if not namespaces:
             raise error.RepoLookupError(_("no namespace exists"
                                           " that match '%s'") % pattern)
 
     names = set()
     for ns in namespaces:
         for name in ns.listnames(repo):
             if name not in ns.deprecated:
                 names.update(repo[n].rev() for n in ns.nodes(repo, name))
 
     names -= {node.nullrev}
     return subset & names
 
 @predicate('id(string)', safe=True)
 def node_(repo, subset, x):
     """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:
         try:
             rn = repo.changelog.rev(node.bin(n))
         except error.WdirUnsupported:
             rn = node.wdirrev
         except (LookupError, TypeError):
             rn = None
     else:
         rn = None
         try:
             pm = repo.changelog._partialmatch(n)
             if pm is not None:
                 rn = repo.changelog.rev(pm)
         except error.WdirUnsupported:
             rn = node.wdirrev
 
     if rn is None:
         return baseset()
     result = baseset([rn])
     return result & subset
 
 @predicate('obsolete()', safe=True)
 def obsolete(repo, subset, x):
     """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 subset & obsoletes
 
 @predicate('only(set, [set])', safe=True)
 def only(repo, subset, x):
     """Changesets that are ancestors of the first set that are not ancestors
     of any other head in the repo. If a second set is specified, the result
     is ancestors of the first set that are not ancestors of the second set
     (i.e. ::<set1> - ::<set2>).
     """
     cl = repo.changelog
     # i18n: "only" is a keyword
     args = getargs(x, 1, 2, _('only takes one or two arguments'))
     include = getset(repo, fullreposet(repo), args[0])
     if len(args) == 1:
         if not include:
             return baseset()
 
         descendants = set(_revdescendants(repo, include, False))
         exclude = [rev for rev in cl.headrevs()
             if not rev in descendants and not rev in include]
     else:
         exclude = getset(repo, fullreposet(repo), args[1])
 
     results = set(cl.findmissingrevs(common=exclude, heads=include))
     # XXX we should turn this into a baseset instead of a set, smartset may do
     # some optimizations from the fact this is a baseset.
     return subset & results
 
 @predicate('origin([set])', safe=True)
 def origin(repo, subset, x):
     """
     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:
         dests = getset(repo, fullreposet(repo), x)
     else:
         dests = fullreposet(repo)
 
     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 = {_firstsrc(r) for r in dests}
     o -= {None}
     # XXX we should turn this into a baseset instead of a set, smartset may do
     # some optimizations from the fact this is a baseset.
     return subset & o
 
 @predicate('outgoing([path])', safe=False)
 def outgoing(repo, subset, x):
     """Changesets not found in the specified destination repository, or the
     default push location.
     """
     # Avoid cycles.
     from . import (
         discovery,
         hg,
     )
     # 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 = {cl.rev(r) for r in outgoing.missing}
     return subset & o
 
 @predicate('p1([set])', safe=True)
 def p1(repo, subset, x):
     """First parent of changesets in set, or the working directory.
     """
     if x is None:
         p = repo[x].p1().rev()
         if p >= 0:
             return subset & baseset([p])
         return baseset()
 
     ps = set()
     cl = repo.changelog
     for r in getset(repo, fullreposet(repo), x):
         try:
             ps.add(cl.parentrevs(r)[0])
         except error.WdirUnsupported:
             ps.add(repo[r].parents()[0].rev())
     ps -= {node.nullrev}
     # XXX we should turn this into a baseset instead of a set, smartset may do
     # some optimizations from the fact this is a baseset.
     return subset & ps
 
 @predicate('p2([set])', safe=True)
 def p2(repo, subset, x):
     """Second parent of changesets in set, or the working directory.
     """
     if x is None:
         ps = repo[x].parents()
         try:
             p = ps[1].rev()
             if p >= 0:
                 return subset & baseset([p])
             return baseset()
         except IndexError:
             return baseset()
 
     ps = set()
     cl = repo.changelog
     for r in getset(repo, fullreposet(repo), x):
         try:
             ps.add(cl.parentrevs(r)[1])
         except error.WdirUnsupported:
             parents = repo[r].parents()
             if len(parents) == 2:
                 ps.add(parents[1])
     ps -= {node.nullrev}
     # XXX we should turn this into a baseset instead of a set, smartset may do
     # some optimizations from the fact this is a baseset.
     return subset & ps
 
 def parentpost(repo, subset, x, order):
     return p1(repo, subset, x)
 
 @predicate('parents([set])', safe=True)
 def parents(repo, subset, x):
     """
     The set of all parents for all changesets in set, or the working directory.
     """
     if x is None:
         ps = set(p.rev() for p in repo[x].parents())
     else:
         ps = set()
         cl = repo.changelog
         up = ps.update
         parentrevs = cl.parentrevs
         for r in getset(repo, fullreposet(repo), x):
             try:
                 up(parentrevs(r))
             except error.WdirUnsupported:
                 up(p.rev() for p in repo[r].parents())
     ps -= {node.nullrev}
     return subset & ps
 
 def _phase(repo, subset, *targets):
     """helper to select all rev in <targets> phases"""
     s = repo._phasecache.getrevset(repo, targets)
     return subset & s
 
 @predicate('draft()', safe=True)
 def draft(repo, subset, x):
     """Changeset in draft phase."""
     # i18n: "draft" is a keyword
     getargs(x, 0, 0, _("draft takes no arguments"))
     target = phases.draft
     return _phase(repo, subset, target)
 
 @predicate('secret()', safe=True)
 def secret(repo, subset, x):
     """Changeset in secret phase."""
     # i18n: "secret" is a keyword
     getargs(x, 0, 0, _("secret takes no arguments"))
     target = phases.secret
     return _phase(repo, subset, target)
 
 def parentspec(repo, subset, x, n, order):
     """``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, fullreposet(repo), x):
         if n == 0:
             ps.add(r)
         elif n == 1:
             try:
                 ps.add(cl.parentrevs(r)[0])
             except error.WdirUnsupported:
                 ps.add(repo[r].parents()[0].rev())
         else:
             try:
                 parents = cl.parentrevs(r)
                 if parents[1] != node.nullrev:
                     ps.add(parents[1])
             except error.WdirUnsupported:
                 parents = repo[r].parents()
                 if len(parents) == 2:
                     ps.add(parents[1].rev())
     return subset & ps
 
 @predicate('present(set)', safe=True)
 def present(repo, subset, x):
     """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 baseset()
 
 # for internal use
 @predicate('_notpublic', safe=True)
 def _notpublic(repo, subset, x):
     getargs(x, 0, 0, "_notpublic takes no arguments")
     return _phase(repo, subset, phases.draft, phases.secret)
 
 @predicate('public()', safe=True)
 def public(repo, subset, x):
     """Changeset in public phase."""
     # i18n: "public" is a keyword
     getargs(x, 0, 0, _("public takes no arguments"))
     phase = repo._phasecache.phase
     target = phases.public
     condition = lambda r: phase(repo, r) == target
     return subset.filter(condition, condrepr=('<phase %r>', target),
                          cache=False)
 
 @predicate('remote([id [,path]])', safe=False)
 def remote(repo, subset, x):
     """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.
     """
 
     from . import hg # avoid start-up nasties
     # i18n: "remote" is a keyword
     l = getargs(x, 0, 2, _("remote takes zero, one, or two 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 baseset([r])
     return baseset()
 
 @predicate('removes(pattern)', safe=True)
 def removes(repo, subset, x):
     """Changesets which remove files matching pattern.
 
     The pattern without explicit kind like ``glob:`` is expected to be
     relative to the current directory and match against a file or a
     directory.
     """
     # i18n: "removes" is a keyword
     pat = getstring(x, _("removes requires a pattern"))
     return checkstatus(repo, subset, pat, 2)
 
 @predicate('rev(number)', safe=True)
 def rev(repo, subset, x):
     """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"))
     if l not in repo.changelog and l not in (node.nullrev, node.wdirrev):
         return baseset()
     return subset & baseset([l])
 
 @predicate('matching(revision [, field])', safe=True)
 def matching(repo, subset, x):
     """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, fullreposet(repo), l[0])
 
     fieldlist = ['metadata']
     if len(l) > 1:
             fieldlist = getstring(l[1],
                 # i18n: "matching" is a keyword
                 _("matching requires a string "
                 "as its second argument")).split()
 
     # Make sure that there are no repeated fields,
     # expand the 'special' 'metadata' field type
     # and check the 'files' whenever we check the 'diff'
     fields = []
     for field in fieldlist:
         if field == 'metadata':
             fields += ['user', 'description', 'date']
         elif field == 'diff':
             # a revision matching the diff must also match the files
             # since matching the diff is very costly, make sure to
             # also match the files first
             fields += ['files', 'diff']
         else:
             if field == 'author':
                 field = 'user'
             fields.append(field)
     fields = set(fields)
     if 'summary' in fields and 'description' in fields:
         # If a revision matches its description it also matches its summary
         fields.discard('summary')
 
     # We may want to match more than one field
     # Not all fields take the same amount of time to be matched
     # Sort the selected fields in order of increasing matching cost
     fieldorder = ['phase', 'parents', 'user', 'date', 'branch', 'summary',
         'files', 'description', 'substate', 'diff']
     def fieldkeyfunc(f):
         try:
             return fieldorder.index(f)
         except ValueError:
             # assume an unknown field is very costly
             return len(fieldorder)
     fields = list(fields)
     fields.sort(key=fieldkeyfunc)
 
     # Each field will be matched with its own "getfield" function
     # which will be added to the getfieldfuncs array of functions
     getfieldfuncs = []
     _funcs = {
         'user': lambda r: repo[r].user(),
         'branch': lambda r: repo[r].branch(),
         'date': lambda r: repo[r].date(),
         'description': lambda r: repo[r].description(),
         'files': lambda r: repo[r].files(),
         'parents': lambda r: repo[r].parents(),
         'phase': lambda r: repo[r].phase(),
         'substate': lambda r: repo[r].substate,
         'summary': lambda r: repo[r].description().splitlines()[0],
         'diff': lambda r: list(repo[r].diff(git=True),)
     }
     for info in fields:
         getfield = _funcs.get(info, None)
         if getfield is None:
             raise error.ParseError(
                 # i18n: "matching" is a keyword
                 _("unexpected field name passed to matching: %s") % info)
         getfieldfuncs.append(getfield)
     # convert the getfield array of functions into a "getinfo" function
     # which returns an array of field values (or a single value if there
     # is only one field to match)
     getinfo = lambda r: [f(r) for f in getfieldfuncs]
 
     def matches(x):
         for rev in revs:
             target = getinfo(rev)
             match = True
             for n, f in enumerate(getfieldfuncs):
                 if target[n] != f(x):
                     match = False
             if match:
                 return True
         return False
 
     return subset.filter(matches, condrepr=('<matching%r %r>', fields, revs))
 
 @predicate('reverse(set)', safe=True, takeorder=True)
 def reverse(repo, subset, x, order):
     """Reverse order of set.
     """
     l = getset(repo, subset, x)
     if order == defineorder:
         l.reverse()
     return l
 
 @predicate('roots(set)', safe=True)
 def roots(repo, subset, x):
     """Changesets in set with no parent changeset in set.
     """
     s = getset(repo, fullreposet(repo), x)
     parents = repo.changelog.parentrevs
     def filter(r):
         for p in parents(r):
             if 0 <= p and p in s:
                 return False
         return True
     return subset & s.filter(filter, condrepr='<roots>')
 
 _sortkeyfuncs = {
     'rev': lambda c: c.rev(),
     'branch': lambda c: c.branch(),
     'desc': lambda c: c.description(),
     'user': lambda c: c.user(),
     'author': lambda c: c.user(),
     'date': lambda c: c.date()[0],
 }
 
 def _getsortargs(x):
     """Parse sort options into (set, [(key, reverse)], opts)"""
     args = getargsdict(x, 'sort', 'set keys topo.firstbranch')
     if 'set' not in args:
         # i18n: "sort" is a keyword
         raise error.ParseError(_('sort requires one or two arguments'))
     keys = "rev"
     if 'keys' in args:
         # i18n: "sort" is a keyword
         keys = getstring(args['keys'], _("sort spec must be a string"))
 
     keyflags = []
     for k in keys.split():
         fk = k
         reverse = (k[0] == '-')
         if reverse:
             k = k[1:]
         if k not in _sortkeyfuncs and k != 'topo':
             raise error.ParseError(_("unknown sort key %r") % fk)
         keyflags.append((k, reverse))
 
     if len(keyflags) > 1 and any(k == 'topo' for k, reverse in keyflags):
         # i18n: "topo" is a keyword
         raise error.ParseError(_('topo sort order cannot be combined '
                                  'with other sort keys'))
 
     opts = {}
     if 'topo.firstbranch' in args:
         if any(k == 'topo' for k, reverse in keyflags):
             opts['topo.firstbranch'] = args['topo.firstbranch']
         else:
             # i18n: "topo" and "topo.firstbranch" are keywords
             raise error.ParseError(_('topo.firstbranch can only be used '
                                      'when using the topo sort key'))
 
     return args['set'], keyflags, opts
 
 @predicate('sort(set[, [-]key... [, ...]])', safe=True, takeorder=True)
 def sort(repo, subset, x, order):
     """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
     - ``topo`` for a reverse topographical sort
 
     The ``topo`` sort order cannot be combined with other sort keys. This sort
     takes one optional argument, ``topo.firstbranch``, which takes a revset that
     specifies what topographical branches to prioritize in the sort.
 
     """
     s, keyflags, opts = _getsortargs(x)
     revs = getset(repo, subset, s)
 
     if not keyflags or order != defineorder:
         return revs
     if len(keyflags) == 1 and keyflags[0][0] == "rev":
         revs.sort(reverse=keyflags[0][1])
         return revs
     elif keyflags[0][0] == "topo":
         firstbranch = ()
         if 'topo.firstbranch' in opts:
             firstbranch = getset(repo, subset, opts['topo.firstbranch'])
         revs = baseset(_toposort(revs, repo.changelog.parentrevs, firstbranch),
                        istopo=True)
         if keyflags[0][1]:
             revs.reverse()
         return revs
 
     # sort() is guaranteed to be stable
     ctxs = [repo[r] for r in revs]
     for k, reverse in reversed(keyflags):
         ctxs.sort(key=_sortkeyfuncs[k], reverse=reverse)
     return baseset([c.rev() for c in ctxs])
 
 def _toposort(revs, parentsfunc, firstbranch=()):
     """Yield revisions from heads to roots one (topo) branch at a time.
 
     This function aims to be used by a graph generator that wishes to minimize
     the number of parallel branches and their interleaving.
 
     Example iteration order (numbers show the "true" order in a changelog):
 
       o  4
       |
       o  1
       |
       | o  3
       | |
       | o  2
       |/
       o  0
 
     Note that the ancestors of merges are understood by the current
     algorithm to be on the same branch. This means no reordering will
     occur behind a merge.
     """
 
     ### Quick summary of the algorithm
     #
     # This function is based around a "retention" principle. We keep revisions
     # in memory until we are ready to emit a whole branch that immediately
     # "merges" into an existing one. This reduces the number of parallel
     # branches with interleaved revisions.
     #
     # During iteration revs are split into two groups:
     # A) revision already emitted
     # B) revision in "retention". They are stored as different subgroups.
     #
     # for each REV, we do the following logic:
     #
     #   1) if REV is a parent of (A), we will emit it. If there is a
     #   retention group ((B) above) that is blocked on REV being
     #   available, we emit all the revisions out of that retention
     #   group first.
     #
     #   2) else, we'll search for a subgroup in (B) awaiting for REV to be
     #   available, if such subgroup exist, we add REV to it and the subgroup is
     #   now awaiting for REV.parents() to be available.
     #
     #   3) finally if no such group existed in (B), we create a new subgroup.
     #
     #
     # To bootstrap the algorithm, we emit the tipmost revision (which
     # puts it in group (A) from above).
 
     revs.sort(reverse=True)
 
     # Set of parents of revision that have been emitted. They can be considered
     # unblocked as the graph generator is already aware of them so there is no
     # need to delay the revisions that reference them.
     #
     # If someone wants to prioritize a branch over the others, pre-filling this
     # set will force all other branches to wait until this branch is ready to be
     # emitted.
     unblocked = set(firstbranch)
 
     # list of groups waiting to be displayed, each group is defined by:
     #
     #   (revs:    lists of revs waiting to be displayed,
     #    blocked: set of that cannot be displayed before those in 'revs')
     #
     # The second value ('blocked') correspond to parents of any revision in the
     # group ('revs') that is not itself contained in the group. The main idea
     # of this algorithm is to delay as much as possible the emission of any
     # revision.  This means waiting for the moment we are about to display
     # these parents to display the revs in a group.
     #
     # This first implementation is smart until it encounters a merge: it will
     # emit revs as soon as any parent is about to be emitted and can grow an
     # arbitrary number of revs in 'blocked'. In practice this mean we properly
     # retains new branches but gives up on any special ordering for ancestors
     # of merges. The implementation can be improved to handle this better.
     #
     # The first subgroup is special. It corresponds to all the revision that
     # were already emitted. The 'revs' lists is expected to be empty and the
     # 'blocked' set contains the parents revisions of already emitted revision.
     #
     # You could pre-seed the <parents> set of groups[0] to a specific
     # changesets to select what the first emitted branch should be.
     groups = [([], unblocked)]
     pendingheap = []
     pendingset = set()
 
     heapq.heapify(pendingheap)
     heappop = heapq.heappop
     heappush = heapq.heappush
     for currentrev in revs:
         # Heap works with smallest element, we want highest so we invert
         if currentrev not in pendingset:
             heappush(pendingheap, -currentrev)
             pendingset.add(currentrev)
         # iterates on pending rev until after the current rev have been
         # processed.
         rev = None
         while rev != currentrev:
             rev = -heappop(pendingheap)
             pendingset.remove(rev)
 
             # Seek for a subgroup blocked, waiting for the current revision.
             matching = [i for i, g in enumerate(groups) if rev in g[1]]
 
             if matching:
                 # The main idea is to gather together all sets that are blocked
                 # on the same revision.
                 #
                 # Groups are merged when a common blocking ancestor is
                 # observed. For example, given two groups:
                 #
                 # revs [5, 4] waiting for 1
                 # revs [3, 2] waiting for 1
                 #
                 # These two groups will be merged when we process
                 # 1. In theory, we could have merged the groups when
                 # we added 2 to the group it is now in (we could have
                 # noticed the groups were both blocked on 1 then), but
                 # the way it works now makes the algorithm simpler.
                 #
                 # We also always keep the oldest subgroup first. We can
                 # probably improve the behavior by having the longest set
                 # first. That way, graph algorithms could minimise the length
                 # of parallel lines their drawing. This is currently not done.
                 targetidx = matching.pop(0)
                 trevs, tparents = groups[targetidx]
                 for i in matching:
                     gr = groups[i]
                     trevs.extend(gr[0])
                     tparents |= gr[1]
                 # delete all merged subgroups (except the one we kept)
                 # (starting from the last subgroup for performance and
                 # sanity reasons)
                 for i in reversed(matching):
                     del groups[i]
             else:
                 # This is a new head. We create a new subgroup for it.
                 targetidx = len(groups)
                 groups.append(([], {rev}))
 
             gr = groups[targetidx]
 
             # We now add the current nodes to this subgroups. This is done
             # after the subgroup merging because all elements from a subgroup
             # that relied on this rev must precede it.
             #
             # we also update the <parents> set to include the parents of the
             # new nodes.
             if rev == currentrev: # only display stuff in rev
                 gr[0].append(rev)
             gr[1].remove(rev)
             parents = [p for p in parentsfunc(rev) if p > node.nullrev]
             gr[1].update(parents)
             for p in parents:
                 if p not in pendingset:
                     pendingset.add(p)
                     heappush(pendingheap, -p)
 
             # Look for a subgroup to display
             #
             # When unblocked is empty (if clause), we were not waiting for any
             # revisions during the first iteration (if no priority was given) or
             # if we emitted a whole disconnected set of the graph (reached a
             # root).  In that case we arbitrarily take the oldest known
             # subgroup. The heuristic could probably be better.
             #
             # Otherwise (elif clause) if the subgroup is blocked on
             # a revision we just emitted, we can safely emit it as
             # well.
             if not unblocked:
                 if len(groups) > 1:  # display other subset
                     targetidx = 1
                     gr = groups[1]
             elif not gr[1] & unblocked:
                 gr = None
 
             if gr is not None:
                 # update the set of awaited revisions with the one from the
                 # subgroup
                 unblocked |= gr[1]
                 # output all revisions in the subgroup
                 for r in gr[0]:
                     yield r
                 # delete the subgroup that you just output
                 # unless it is groups[0] in which case you just empty it.
                 if targetidx:
                     del groups[targetidx]
                 else:
                     gr[0][:] = []
     # Check if we have some subgroup waiting for revisions we are not going to
     # iterate over
     for g in groups:
         for r in g[0]:
             yield r
 
 @predicate('subrepo([pattern])')
 def subrepo(repo, subset, x):
     """Changesets that add, modify or remove the given subrepo.  If no subrepo
     pattern is named, any subrepo changes are returned.
     """
     # i18n: "subrepo" is a keyword
     args = getargs(x, 0, 1, _('subrepo takes at most one argument'))
     pat = None
     if len(args) != 0:
         pat = getstring(args[0], _("subrepo requires a pattern"))
 
     m = matchmod.exact(repo.root, repo.root, ['.hgsubstate'])
 
     def submatches(names):
         k, p, m = util.stringmatcher(pat)
         for name in names:
             if m(name):
                 yield name
 
     def matches(x):
         c = repo[x]
         s = repo.status(c.p1().node(), c.node(), match=m)
 
         if pat is None:
             return s.added or s.modified or s.removed
 
         if s.added:
             return any(submatches(c.substate.keys()))
 
         if s.modified:
             subs = set(c.p1().substate.keys())
             subs.update(c.substate.keys())
 
             for path in submatches(subs):
                 if c.p1().substate.get(path) != c.substate.get(path):
                     return True
 
         if s.removed:
             return any(submatches(c.p1().substate.keys()))
 
         return False
 
     return subset.filter(matches, condrepr=('<subrepo %r>', pat))
 
 def _substringmatcher(pattern, casesensitive=True):
     kind, pattern, matcher = util.stringmatcher(pattern,
                                                 casesensitive=casesensitive)
     if kind == 'literal':
         if not casesensitive:
             pattern = encoding.lower(pattern)
             matcher = lambda s: pattern in encoding.lower(s)
         else:
             matcher = lambda s: pattern in s
     return kind, pattern, matcher
 
 @predicate('tag([name])', safe=True)
 def tag(repo, subset, x):
     """The specified tag by name, or all tagged revisions if no name is given.
 
     Pattern matching is supported for `name`. See
     :hg:`help revisions.patterns`.
     """
     # 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 = util.stringmatcher(pattern)
         if kind == 'literal':
             # avoid resolving all tags
             tn = repo._tagscache.tags.get(pattern, None)
             if tn is None:
                 raise error.RepoLookupError(_("tag '%s' does not exist")
                                             % pattern)
             s = {repo[tn].rev()}
         else:
             s = {cl.rev(n) for t, n in repo.tagslist() if matcher(t)}
     else:
         s = {cl.rev(n) for t, n in repo.tagslist() if t != 'tip'}
     return subset & s
 
 @predicate('tagged', safe=True)
 def tagged(repo, subset, x):
     return tag(repo, subset, x)
 
 @predicate('unstable()', safe=True)
 def unstable(repo, subset, x):
     """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 subset & unstables
 
 
 @predicate('user(string)', safe=True)
 def user(repo, subset, x):
     """User name contains string. The match is case-insensitive.
 
     Pattern matching is supported for `string`. See
     :hg:`help revisions.patterns`.
     """
     return author(repo, subset, x)
 
 @predicate('wdir()', safe=True)
 def wdir(repo, subset, x):
     """Working directory. (EXPERIMENTAL)"""
     # i18n: "wdir" is a keyword
     getargs(x, 0, 0, _("wdir takes no arguments"))
     if node.wdirrev in subset or isinstance(subset, fullreposet):
         return baseset([node.wdirrev])
     return baseset()
 
 def _orderedlist(repo, subset, x):
     s = getstring(x, "internal error")
     if not s:
         return baseset()
     # remove duplicates here. it's difficult for caller to deduplicate sets
     # because different symbols can point to the same rev.
     cl = repo.changelog
     ls = []
     seen = set()
     for t in s.split('\0'):
         try:
             # fast path for integer revision
             r = int(t)
             if str(r) != t or r not in cl:
                 raise ValueError
             revs = [r]
         except ValueError:
             revs = stringset(repo, subset, t)
 
         for r in revs:
             if r in seen:
                 continue
             if (r in subset
                 or r == node.nullrev and isinstance(subset, fullreposet)):
                 ls.append(r)
             seen.add(r)
     return baseset(ls)
 
 # for internal use
 @predicate('_list', safe=True, takeorder=True)
 def _list(repo, subset, x, order):
     if order == followorder:
         # slow path to take the subset order
         return subset & _orderedlist(repo, fullreposet(repo), x)
     else:
         return _orderedlist(repo, subset, x)
 
 def _orderedintlist(repo, subset, x):
     s = getstring(x, "internal error")
     if not s:
         return baseset()
     ls = [int(r) for r in s.split('\0')]
     s = subset
     return baseset([r for r in ls if r in s])
 
 # for internal use
 @predicate('_intlist', safe=True, takeorder=True)
 def _intlist(repo, subset, x, order):
     if order == followorder:
         # slow path to take the subset order
         return subset & _orderedintlist(repo, fullreposet(repo), x)
     else:
         return _orderedintlist(repo, subset, x)
 
 def _orderedhexlist(repo, subset, x):
     s = getstring(x, "internal error")
     if not s:
         return baseset()
     cl = repo.changelog
     ls = [cl.rev(node.bin(r)) for r in s.split('\0')]
     s = subset
     return baseset([r for r in ls if r in s])
 
 # for internal use
 @predicate('_hexlist', safe=True, takeorder=True)
 def _hexlist(repo, subset, x, order):
     if order == followorder:
         # slow path to take the subset order
         return subset & _orderedhexlist(repo, fullreposet(repo), x)
     else:
         return _orderedhexlist(repo, subset, x)
 
 methods = {
     "range": rangeset,
     "rangeall": rangeall,
     "rangepre": rangepre,
     "rangepost": rangepost,
     "dagrange": dagrange,
     "string": stringset,
     "symbol": stringset,
     "and": andset,
     "or": orset,
     "not": notset,
     "difference": differenceset,
     "list": listset,
     "keyvalue": keyvaluepair,
     "func": func,
     "ancestor": ancestorspec,
     "parent": parentspec,
     "parentpost": parentpost,
 }
 
 def posttreebuilthook(tree, repo):
     # hook for extensions to execute code on the optimized tree
     pass
 
 def match(ui, spec, repo=None, order=defineorder):
     """Create a matcher for a single revision spec
 
     If order=followorder, a matcher takes the ordering specified by the input
     set.
     """
     return matchany(ui, [spec], repo=repo, order=order)
 
 def matchany(ui, specs, repo=None, order=defineorder):
     """Create a matcher that will include any revisions matching one of the
     given specs
 
     If order=followorder, a matcher takes the ordering specified by the input
     set.
     """
     if not specs:
         def mfunc(repo, subset=None):
             return baseset()
         return mfunc
     if not all(specs):
         raise error.ParseError(_("empty query"))
     lookup = None
     if repo:
         lookup = repo.__contains__
     if len(specs) == 1:
         tree = revsetlang.parse(specs[0], lookup)
     else:
         tree = ('or',
                 ('list',) + tuple(revsetlang.parse(s, lookup) for s in specs))
 
     if ui:
         tree = revsetlang.expandaliases(ui, tree)
     tree = revsetlang.foldconcat(tree)
     tree = revsetlang.analyze(tree, order)
     tree = revsetlang.optimize(tree)
     posttreebuilthook(tree, repo)
     return makematcher(tree)
 
 def makematcher(tree):
     """Create a matcher from an evaluatable tree"""
     def mfunc(repo, subset=None):
         if subset is None:
             subset = fullreposet(repo)
         return getset(repo, subset, tree)
     return mfunc
 
 def loadpredicate(ui, extname, registrarobj):
     """Load revset predicates from specified registrarobj
     """
     for name, func in registrarobj._table.iteritems():
         symbols[name] = func
         if func._safe:
             safesymbols.add(name)
 
 # load built-in predicates explicitly to setup safesymbols
 loadpredicate(None, None, predicate)
 
 # tell hggettext to extract docstrings from these functions:
 i18nfunctions = symbols.values()

mercurial/smartset.py

@@ -1,1070 +1,1093
 # smartset.py - data structure for revision set
 #
 # 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.
 
 from __future__ import absolute_import
 
 from . import (
+    error,
     util,
 )
 
 def _formatsetrepr(r):
     """Format an optional printable representation of a set
 
     ========  =================================
     type(r)   example
     ========  =================================
     tuple     ('<not %r>', other)
     str       '<branch closed>'
     callable  lambda: '<branch %r>' % sorted(b)
     object    other
     ========  =================================
     """
     if r is None:
         return ''
     elif isinstance(r, tuple):
         return r[0] % r[1:]
     elif isinstance(r, str):
         return r
     elif callable(r):
         return r()
     else:
         return repr(r)
 
 class abstractsmartset(object):
 
     def __nonzero__(self):
         """True if the smartset is not empty"""
         raise NotImplementedError()
 
     __bool__ = __nonzero__
 
     def __contains__(self, rev):
         """provide fast membership testing"""
         raise NotImplementedError()
 
     def __iter__(self):
         """iterate the set in the order it is supposed to be iterated"""
         raise NotImplementedError()
 
     # Attributes containing a function to perform a fast iteration in a given
     # direction. A smartset can have none, one, or both defined.
     #
     # Default value is None instead of a function returning None to avoid
     # initializing an iterator just for testing if a fast method exists.
     fastasc = None
     fastdesc = None
 
     def isascending(self):
         """True if the set will iterate in ascending order"""
         raise NotImplementedError()
 
     def isdescending(self):
         """True if the set will iterate in descending order"""
         raise NotImplementedError()
 
     def istopo(self):
         """True if the set will iterate in topographical order"""
         raise NotImplementedError()
 
     def min(self):
         """return the minimum element in the set"""
         if self.fastasc is None:
             v = min(self)
         else:
             for v in self.fastasc():
                 break
             else:
                 raise ValueError('arg is an empty sequence')
         self.min = lambda: v
         return v
 
     def max(self):
         """return the maximum element in the set"""
         if self.fastdesc is None:
             return max(self)
         else:
             for v in self.fastdesc():
                 break
             else:
                 raise ValueError('arg is an empty sequence')
         self.max = lambda: v
         return v
 
     def first(self):
         """return the first element in the set (user iteration perspective)
 
         Return None if the set is empty"""
         raise NotImplementedError()
 
     def last(self):
         """return the last element in the set (user iteration perspective)
 
         Return None if the set is empty"""
         raise NotImplementedError()
 
     def __len__(self):
         """return the length of the smartsets
 
         This can be expensive on smartset that could be lazy otherwise."""
         raise NotImplementedError()
 
     def reverse(self):
         """reverse the expected iteration order"""
         raise NotImplementedError()
 
     def sort(self, reverse=True):
         """get the set to iterate in an ascending or descending order"""
         raise NotImplementedError()
 
     def __and__(self, other):
         """Returns a new object with the intersection of the two collections.
 
         This is part of the mandatory API for smartset."""
         if isinstance(other, fullreposet):
             return self
         return self.filter(other.__contains__, condrepr=other, cache=False)
 
     def __add__(self, other):
         """Returns a new object with the union of the two collections.
 
         This is part of the mandatory API for smartset."""
         return addset(self, other)
 
     def __sub__(self, other):
         """Returns a new object with the substraction of the two collections.
 
         This is part of the mandatory API for smartset."""
         c = other.__contains__
         return self.filter(lambda r: not c(r), condrepr=('<not %r>', other),
                            cache=False)
 
     def filter(self, condition, condrepr=None, cache=True):
         """Returns this smartset filtered by condition as a new smartset.
 
         `condition` is a callable which takes a revision number and returns a
         boolean. Optional `condrepr` provides a printable representation of
         the given `condition`.
 
         This is part of the mandatory API for smartset."""
         # builtin cannot be cached. but do not needs to
         if cache and util.safehasattr(condition, 'func_code'):
             condition = util.cachefunc(condition)
         return filteredset(self, condition, condrepr)
 
+    def slice(self, start, stop):
+        """Return new smartset that contains selected elements from this set"""
+        if start < 0 or stop < 0:
+            raise error.ProgrammingError('negative index not allowed')
+        return self._slice(start, stop)
+
+    def _slice(self, start, stop):
+        # sub classes may override this. start and stop must not be negative,
+        # but start > stop is allowed, which should be an empty set.
+        ys = []
+        it = iter(self)
+        for x in xrange(start):
+            y = next(it, None)
+            if y is None:
+                break
+        for x in xrange(stop - start):
+            y = next(it, None)
+            if y is None:
+                break
+            ys.append(y)
+        return baseset(ys, datarepr=('slice=%d:%d %r', start, stop, self))
+
 class baseset(abstractsmartset):
     """Basic data structure that represents a revset and contains the basic
     operation that it should be able to perform.
 
     Every method in this class should be implemented by any smartset class.
 
     This class could be constructed by an (unordered) set, or an (ordered)
     list-like object. If a set is provided, it'll be sorted lazily.
 
     >>> x = [4, 0, 7, 6]
     >>> y = [5, 6, 7, 3]
 
     Construct by a set:
     >>> xs = baseset(set(x))
     >>> ys = baseset(set(y))
     >>> [list(i) for i in [xs + ys, xs & ys, xs - ys]]
     [[0, 4, 6, 7, 3, 5], [6, 7], [0, 4]]
     >>> [type(i).__name__ for i in [xs + ys, xs & ys, xs - ys]]
     ['addset', 'baseset', 'baseset']
 
     Construct by a list-like:
     >>> xs = baseset(x)
     >>> ys = baseset(i for i in y)
     >>> [list(i) for i in [xs + ys, xs & ys, xs - ys]]
     [[4, 0, 7, 6, 5, 3], [7, 6], [4, 0]]
     >>> [type(i).__name__ for i in [xs + ys, xs & ys, xs - ys]]
     ['addset', 'filteredset', 'filteredset']
 
     Populate "_set" fields in the lists so set optimization may be used:
     >>> [1 in xs, 3 in ys]
     [False, True]
 
     Without sort(), results won't be changed:
     >>> [list(i) for i in [xs + ys, xs & ys, xs - ys]]
     [[4, 0, 7, 6, 5, 3], [7, 6], [4, 0]]
     >>> [type(i).__name__ for i in [xs + ys, xs & ys, xs - ys]]
     ['addset', 'filteredset', 'filteredset']
 
     With sort(), set optimization could be used:
     >>> xs.sort(reverse=True)
     >>> [list(i) for i in [xs + ys, xs & ys, xs - ys]]
     [[7, 6, 4, 0, 5, 3], [7, 6], [4, 0]]
     >>> [type(i).__name__ for i in [xs + ys, xs & ys, xs - ys]]
     ['addset', 'baseset', 'baseset']
 
     >>> ys.sort()
     >>> [list(i) for i in [xs + ys, xs & ys, xs - ys]]
     [[7, 6, 4, 0, 3, 5], [7, 6], [4, 0]]
     >>> [type(i).__name__ for i in [xs + ys, xs & ys, xs - ys]]
     ['addset', 'baseset', 'baseset']
 
     istopo is preserved across set operations
     >>> xs = baseset(set(x), istopo=True)
     >>> rs = xs & ys
     >>> type(rs).__name__
     'baseset'
     >>> rs._istopo
     True
     """
     def __init__(self, data=(), datarepr=None, istopo=False):
         """
         datarepr: a tuple of (format, obj, ...), a function or an object that
                   provides a printable representation of the given data.
         """
         self._ascending = None
         self._istopo = istopo
         if isinstance(data, set):
             # converting set to list has a cost, do it lazily
             self._set = data
             # set has no order we pick one for stability purpose
             self._ascending = True
         else:
             if not isinstance(data, list):
                 data = list(data)
             self._list = data
         self._datarepr = datarepr
 
     @util.propertycache
     def _set(self):
         return set(self._list)
 
     @util.propertycache
     def _asclist(self):
         asclist = self._list[:]
         asclist.sort()
         return asclist
 
     @util.propertycache
     def _list(self):
         # _list is only lazily constructed if we have _set
         assert r'_set' in self.__dict__
         return list(self._set)
 
     def __iter__(self):
         if self._ascending is None:
             return iter(self._list)
         elif self._ascending:
             return iter(self._asclist)
         else:
             return reversed(self._asclist)
 
     def fastasc(self):
         return iter(self._asclist)
 
     def fastdesc(self):
         return reversed(self._asclist)
 
     @util.propertycache
     def __contains__(self):
         return self._set.__contains__
 
     def __nonzero__(self):
         return bool(len(self))
 
     __bool__ = __nonzero__
 
     def sort(self, reverse=False):
         self._ascending = not bool(reverse)
         self._istopo = False
 
     def reverse(self):
         if self._ascending is None:
             self._list.reverse()
         else:
             self._ascending = not self._ascending
         self._istopo = False
 
     def __len__(self):
         if '_list' in self.__dict__:
             return len(self._list)
         else:
             return len(self._set)
 
     def isascending(self):
         """Returns True if the collection is ascending order, False if not.
 
         This is part of the mandatory API for smartset."""
         if len(self) <= 1:
             return True
         return self._ascending is not None and self._ascending
 
     def isdescending(self):
         """Returns True if the collection is descending order, False if not.
 
         This is part of the mandatory API for smartset."""
         if len(self) <= 1:
             return True
         return self._ascending is not None and not self._ascending
 
     def istopo(self):
         """Is the collection is in topographical order or not.
 
         This is part of the mandatory API for smartset."""
         if len(self) <= 1:
             return True
         return self._istopo
 
     def first(self):
         if self:
             if self._ascending is None:
                 return self._list[0]
             elif self._ascending:
                 return self._asclist[0]
             else:
                 return self._asclist[-1]
         return None
 
     def last(self):
         if self:
             if self._ascending is None:
                 return self._list[-1]
             elif self._ascending:
                 return self._asclist[-1]
             else:
                 return self._asclist[0]
         return None
 
     def _fastsetop(self, other, op):
         # try to use native set operations as fast paths
         if (type(other) is baseset and '_set' in other.__dict__ and '_set' in
             self.__dict__ and self._ascending is not None):
             s = baseset(data=getattr(self._set, op)(other._set),
                         istopo=self._istopo)
             s._ascending = self._ascending
         else:
             s = getattr(super(baseset, self), op)(other)
         return s
 
     def __and__(self, other):
         return self._fastsetop(other, '__and__')
 
     def __sub__(self, other):
         return self._fastsetop(other, '__sub__')
 
     def __repr__(self):
         d = {None: '', False: '-', True: '+'}[self._ascending]
         s = _formatsetrepr(self._datarepr)
         if not s:
             l = self._list
             # if _list has been built from a set, it might have a different
             # order from one python implementation to another.
             # We fallback to the sorted version for a stable output.
             if self._ascending is not None:
                 l = self._asclist
             s = repr(l)
         return '<%s%s %s>' % (type(self).__name__, d, s)
 
 class filteredset(abstractsmartset):
     """Duck type for baseset class which iterates lazily over the revisions in
     the subset and contains a function which tests for membership in the
     revset
     """
     def __init__(self, subset, condition=lambda x: True, condrepr=None):
         """
         condition: a function that decide whether a revision in the subset
                    belongs to the revset or not.
         condrepr: a tuple of (format, obj, ...), a function or an object that
                   provides a printable representation of the given condition.
         """
         self._subset = subset
         self._condition = condition
         self._condrepr = condrepr
 
     def __contains__(self, x):
         return x in self._subset and self._condition(x)
 
     def __iter__(self):
         return self._iterfilter(self._subset)
 
     def _iterfilter(self, it):
         cond = self._condition
         for x in it:
             if cond(x):
                 yield x
 
     @property
     def fastasc(self):
         it = self._subset.fastasc
         if it is None:
             return None
         return lambda: self._iterfilter(it())
 
     @property
     def fastdesc(self):
         it = self._subset.fastdesc
         if it is None:
             return None
         return lambda: self._iterfilter(it())
 
     def __nonzero__(self):
         fast = None
         candidates = [self.fastasc if self.isascending() else None,
                       self.fastdesc if self.isdescending() else None,
                       self.fastasc,
                       self.fastdesc]
         for candidate in candidates:
             if candidate is not None:
                 fast = candidate
                 break
 
         if fast is not None:
             it = fast()
         else:
             it = self
 
         for r in it:
             return True
         return False
 
     __bool__ = __nonzero__
 
     def __len__(self):
         # Basic implementation to be changed in future patches.
         # until this gets improved, we use generator expression
         # here, since list comprehensions are free to call __len__ again
         # causing infinite recursion
         l = baseset(r for r in self)
         return len(l)
 
     def sort(self, reverse=False):
         self._subset.sort(reverse=reverse)
 
     def reverse(self):
         self._subset.reverse()
 
     def isascending(self):
         return self._subset.isascending()
 
     def isdescending(self):
         return self._subset.isdescending()
 
     def istopo(self):
         return self._subset.istopo()
 
     def first(self):
         for x in self:
             return x
         return None
 
     def last(self):
         it = None
         if self.isascending():
             it = self.fastdesc
         elif self.isdescending():
             it = self.fastasc
         if it is not None:
             for x in it():
                 return x
             return None #empty case
         else:
             x = None
             for x in self:
                 pass
             return x
 
     def __repr__(self):
         xs = [repr(self._subset)]
         s = _formatsetrepr(self._condrepr)
         if s:
             xs.append(s)
         return '<%s %s>' % (type(self).__name__, ', '.join(xs))
 
 def _iterordered(ascending, iter1, iter2):
     """produce an ordered iteration from two iterators with the same order
 
     The ascending is used to indicated the iteration direction.
     """
     choice = max
     if ascending:
         choice = min
 
     val1 = None
     val2 = None
     try:
         # Consume both iterators in an ordered way until one is empty
         while True:
             if val1 is None:
                 val1 = next(iter1)
             if val2 is None:
                 val2 = next(iter2)
             n = choice(val1, val2)
             yield n
             if val1 == n:
                 val1 = None
             if val2 == n:
                 val2 = None
     except StopIteration:
         # Flush any remaining values and consume the other one
         it = iter2
         if val1 is not None:
             yield val1
             it = iter1
         elif val2 is not None:
             # might have been equality and both are empty
             yield val2
         for val in it:
             yield val
 
 class addset(abstractsmartset):
     """Represent the addition of two sets
 
     Wrapper structure for lazily adding two structures without losing much
     performance on the __contains__ method
 
     If the ascending attribute is set, that means the two structures are
     ordered in either an ascending or descending way. Therefore, we can add
     them maintaining the order by iterating over both at the same time
 
     >>> xs = baseset([0, 3, 2])
     >>> ys = baseset([5, 2, 4])
 
     >>> rs = addset(xs, ys)
     >>> bool(rs), 0 in rs, 1 in rs, 5 in rs, rs.first(), rs.last()
     (True, True, False, True, 0, 4)
     >>> rs = addset(xs, baseset([]))
     >>> bool(rs), 0 in rs, 1 in rs, rs.first(), rs.last()
     (True, True, False, 0, 2)
     >>> rs = addset(baseset([]), baseset([]))
     >>> bool(rs), 0 in rs, rs.first(), rs.last()
     (False, False, None, None)
 
     iterate unsorted:
     >>> rs = addset(xs, ys)
     >>> # (use generator because pypy could call len())
     >>> list(x for x in rs)  # without _genlist
     [0, 3, 2, 5, 4]
     >>> assert not rs._genlist
     >>> len(rs)
     5
     >>> [x for x in rs]  # with _genlist
     [0, 3, 2, 5, 4]
     >>> assert rs._genlist
 
     iterate ascending:
     >>> rs = addset(xs, ys, ascending=True)
     >>> # (use generator because pypy could call len())
     >>> list(x for x in rs), list(x for x in rs.fastasc())  # without _asclist
     ([0, 2, 3, 4, 5], [0, 2, 3, 4, 5])
     >>> assert not rs._asclist
     >>> len(rs)
     5
     >>> [x for x in rs], [x for x in rs.fastasc()]
     ([0, 2, 3, 4, 5], [0, 2, 3, 4, 5])
     >>> assert rs._asclist
 
     iterate descending:
     >>> rs = addset(xs, ys, ascending=False)
     >>> # (use generator because pypy could call len())
     >>> list(x for x in rs), list(x for x in rs.fastdesc())  # without _asclist
     ([5, 4, 3, 2, 0], [5, 4, 3, 2, 0])
     >>> assert not rs._asclist
     >>> len(rs)
     5
     >>> [x for x in rs], [x for x in rs.fastdesc()]
     ([5, 4, 3, 2, 0], [5, 4, 3, 2, 0])
     >>> assert rs._asclist
 
     iterate ascending without fastasc:
     >>> rs = addset(xs, generatorset(ys), ascending=True)
     >>> assert rs.fastasc is None
     >>> [x for x in rs]
     [0, 2, 3, 4, 5]
 
     iterate descending without fastdesc:
     >>> rs = addset(generatorset(xs), ys, ascending=False)
     >>> assert rs.fastdesc is None
     >>> [x for x in rs]
     [5, 4, 3, 2, 0]
     """
     def __init__(self, revs1, revs2, ascending=None):
         self._r1 = revs1
         self._r2 = revs2
         self._iter = None
         self._ascending = ascending
         self._genlist = None
         self._asclist = None
 
     def __len__(self):
         return len(self._list)
 
     def __nonzero__(self):
         return bool(self._r1) or bool(self._r2)
 
     __bool__ = __nonzero__
 
     @util.propertycache
     def _list(self):
         if not self._genlist:
             self._genlist = baseset(iter(self))
         return self._genlist
 
     def __iter__(self):
         """Iterate over both collections without repeating elements
 
         If the ascending attribute is not set, iterate over the first one and
         then over the second one checking for membership on the first one so we
         dont yield any duplicates.
 
         If the ascending attribute is set, iterate over both collections at the
         same time, yielding only one value at a time in the given order.
         """
         if self._ascending is None:
             if self._genlist:
                 return iter(self._genlist)
             def arbitraryordergen():
                 for r in self._r1:
                     yield r
                 inr1 = self._r1.__contains__
                 for r in self._r2:
                     if not inr1(r):
                         yield r
             return arbitraryordergen()
         # try to use our own fast iterator if it exists
         self._trysetasclist()
         if self._ascending:
             attr = 'fastasc'
         else:
             attr = 'fastdesc'
         it = getattr(self, attr)
         if it is not None:
             return it()
         # maybe half of the component supports fast
         # get iterator for _r1
         iter1 = getattr(self._r1, attr)
         if iter1 is None:
             # let's avoid side effect (not sure it matters)
             iter1 = iter(sorted(self._r1, reverse=not self._ascending))
         else:
             iter1 = iter1()
         # get iterator for _r2
         iter2 = getattr(self._r2, attr)
         if iter2 is None:
             # let's avoid side effect (not sure it matters)
             iter2 = iter(sorted(self._r2, reverse=not self._ascending))
         else:
             iter2 = iter2()
         return _iterordered(self._ascending, iter1, iter2)
 
     def _trysetasclist(self):
         """populate the _asclist attribute if possible and necessary"""
         if self._genlist is not None and self._asclist is None:
             self._asclist = sorted(self._genlist)
 
     @property
     def fastasc(self):
         self._trysetasclist()
         if self._asclist is not None:
             return self._asclist.__iter__
         iter1 = self._r1.fastasc
         iter2 = self._r2.fastasc
         if None in (iter1, iter2):
             return None
         return lambda: _iterordered(True, iter1(), iter2())
 
     @property
     def fastdesc(self):
         self._trysetasclist()
         if self._asclist is not None:
             return self._asclist.__reversed__
         iter1 = self._r1.fastdesc
         iter2 = self._r2.fastdesc
         if None in (iter1, iter2):
             return None
         return lambda: _iterordered(False, iter1(), iter2())
 
     def __contains__(self, x):
         return x in self._r1 or x in self._r2
 
     def sort(self, reverse=False):
         """Sort the added set
 
         For this we use the cached list with all the generated values and if we
         know they are ascending or descending we can sort them in a smart way.
         """
         self._ascending = not reverse
 
     def isascending(self):
         return self._ascending is not None and self._ascending
 
     def isdescending(self):
         return self._ascending is not None and not self._ascending
 
     def istopo(self):
         # not worth the trouble asserting if the two sets combined are still
         # in topographical order. Use the sort() predicate to explicitly sort
         # again instead.
         return False
 
     def reverse(self):
         if self._ascending is None:
             self._list.reverse()
         else:
             self._ascending = not self._ascending
 
     def first(self):
         for x in self:
             return x
         return None
 
     def last(self):
         self.reverse()
         val = self.first()
         self.reverse()
         return val
 
     def __repr__(self):
         d = {None: '', False: '-', True: '+'}[self._ascending]
         return '<%s%s %r, %r>' % (type(self).__name__, d, self._r1, self._r2)
 
 class generatorset(abstractsmartset):
     """Wrap a generator for lazy iteration
 
     Wrapper structure for generators that provides lazy membership and can
     be iterated more than once.
     When asked for membership it generates values until either it finds the
     requested one or has gone through all the elements in the generator
     """
     def __init__(self, gen, iterasc=None):
         """
         gen: a generator producing the values for the generatorset.
         """
         self._gen = gen
         self._asclist = None
         self._cache = {}
         self._genlist = []
         self._finished = False
         self._ascending = True
         if iterasc is not None:
             if iterasc:
                 self.fastasc = self._iterator
                 self.__contains__ = self._asccontains
             else:
                 self.fastdesc = self._iterator
                 self.__contains__ = self._desccontains
 
     def __nonzero__(self):
         # Do not use 'for r in self' because it will enforce the iteration
         # order (default ascending), possibly unrolling a whole descending
         # iterator.
         if self._genlist:
             return True
         for r in self._consumegen():
             return True
         return False
 
     __bool__ = __nonzero__
 
     def __contains__(self, x):
         if x in self._cache:
             return self._cache[x]
 
         # Use new values only, as existing values would be cached.
         for l in self._consumegen():
             if l == x:
                 return True
 
         self._cache[x] = False
         return False
 
     def _asccontains(self, x):
         """version of contains optimised for ascending generator"""
         if x in self._cache:
             return self._cache[x]
 
         # Use new values only, as existing values would be cached.
         for l in self._consumegen():
             if l == x:
                 return True
             if l > x:
                 break
 
         self._cache[x] = False
         return False
 
     def _desccontains(self, x):
         """version of contains optimised for descending generator"""
         if x in self._cache:
             return self._cache[x]
 
         # Use new values only, as existing values would be cached.
         for l in self._consumegen():
             if l == x:
                 return True
             if l < x:
                 break
 
         self._cache[x] = False
         return False
 
     def __iter__(self):
         if self._ascending:
             it = self.fastasc
         else:
             it = self.fastdesc
         if it is not None:
             return it()
         # we need to consume the iterator
         for x in self._consumegen():
             pass
         # recall the same code
         return iter(self)
 
     def _iterator(self):
         if self._finished:
             return iter(self._genlist)
 
         # We have to use this complex iteration strategy to allow multiple
         # iterations at the same time. We need to be able to catch revision
         # removed from _consumegen and added to genlist in another instance.
         #
         # Getting rid of it would provide an about 15% speed up on this
         # iteration.
         genlist = self._genlist
         nextgen = self._consumegen()
         _len, _next = len, next # cache global lookup
         def gen():
             i = 0
             while True:
                 if i < _len(genlist):
                     yield genlist[i]
                 else:
                     yield _next(nextgen)
                 i += 1
         return gen()
 
     def _consumegen(self):
         cache = self._cache
         genlist = self._genlist.append
         for item in self._gen:
             cache[item] = True
             genlist(item)
             yield item
         if not self._finished:
             self._finished = True
             asc = self._genlist[:]
             asc.sort()
             self._asclist = asc
             self.fastasc = asc.__iter__
             self.fastdesc = asc.__reversed__
 
     def __len__(self):
         for x in self._consumegen():
             pass
         return len(self._genlist)
 
     def sort(self, reverse=False):
         self._ascending = not reverse
 
     def reverse(self):
         self._ascending = not self._ascending
 
     def isascending(self):
         return self._ascending
 
     def isdescending(self):
         return not self._ascending
 
     def istopo(self):
         # not worth the trouble asserting if the two sets combined are still
         # in topographical order. Use the sort() predicate to explicitly sort
         # again instead.
         return False
 
     def first(self):
         if self._ascending:
             it = self.fastasc
         else:
             it = self.fastdesc
         if it is None:
             # we need to consume all and try again
             for x in self._consumegen():
                 pass
             return self.first()
         return next(it(), None)
 
     def last(self):
         if self._ascending:
             it = self.fastdesc
         else:
             it = self.fastasc
         if it is None:
             # we need to consume all and try again
             for x in self._consumegen():
                 pass
             return self.first()
         return next(it(), None)
 
     def __repr__(self):
         d = {False: '-', True: '+'}[self._ascending]
         return '<%s%s>' % (type(self).__name__, d)
 
 def spanset(repo, start=0, end=None):
     """Create a spanset that represents a range of repository revisions
 
     start: first revision included the set (default to 0)
     end:   first revision excluded (last+1) (default to len(repo))
 
     Spanset will be descending if `end` < `start`.
     """
     if end is None:
         end = len(repo)
     ascending = start <= end
     if not ascending:
         start, end = end + 1, start + 1
     return _spanset(start, end, ascending, repo.changelog.filteredrevs)
 
 class _spanset(abstractsmartset):
     """Duck type for baseset class which represents a range of revisions and
     can work lazily and without having all the range in memory
 
     Note that spanset(x, y) behave almost like xrange(x, y) except for two
     notable points:
     - when x < y it will be automatically descending,
     - revision filtered with this repoview will be skipped.
 
     """
     def __init__(self, start, end, ascending, hiddenrevs):
         self._start = start
         self._end = end
         self._ascending = ascending
         self._hiddenrevs = hiddenrevs
 
     def sort(self, reverse=False):
         self._ascending = not reverse
 
     def reverse(self):
         self._ascending = not self._ascending
 
     def istopo(self):
         # not worth the trouble asserting if the two sets combined are still
         # in topographical order. Use the sort() predicate to explicitly sort
         # again instead.
         return False
 
     def _iterfilter(self, iterrange):
         s = self._hiddenrevs
         for r in iterrange:
             if r not in s:
                 yield r
 
     def __iter__(self):
         if self._ascending:
             return self.fastasc()
         else:
             return self.fastdesc()
 
     def fastasc(self):
         iterrange = xrange(self._start, self._end)
         if self._hiddenrevs:
             return self._iterfilter(iterrange)
         return iter(iterrange)
 
     def fastdesc(self):
         iterrange = xrange(self._end - 1, self._start - 1, -1)
         if self._hiddenrevs:
             return self._iterfilter(iterrange)
         return iter(iterrange)
 
     def __contains__(self, rev):
         hidden = self._hiddenrevs
         return ((self._start <= rev < self._end)
                 and not (hidden and rev in hidden))
 
     def __nonzero__(self):
         for r in self:
             return True
         return False
 
     __bool__ = __nonzero__
 
     def __len__(self):
         if not self._hiddenrevs:
             return abs(self._end - self._start)
         else:
             count = 0
             start = self._start
             end = self._end
             for rev in self._hiddenrevs:
                 if (end < rev <= start) or (start <= rev < end):
                     count += 1
             return abs(self._end - self._start) - count
 
     def isascending(self):
         return self._ascending
 
     def isdescending(self):
         return not self._ascending
 
     def first(self):
         if self._ascending:
             it = self.fastasc
         else:
             it = self.fastdesc
         for x in it():
             return x
         return None
 
     def last(self):
         if self._ascending:
             it = self.fastdesc
         else:
             it = self.fastasc
         for x in it():
             return x
         return None
 
     def __repr__(self):
         d = {False: '-', True: '+'}[self._ascending]
         return '<%s%s %d:%d>' % (type(self).__name__.lstrip('_'), d,
                                  self._start, self._end)
 
 class fullreposet(_spanset):
     """a set containing all revisions in the repo
 
     This class exists to host special optimization and magic to handle virtual
     revisions such as "null".
     """
 
     def __init__(self, repo):
         super(fullreposet, self).__init__(0, len(repo), True,
                                           repo.changelog.filteredrevs)
 
     def __and__(self, other):
         """As self contains the whole repo, all of the other set should also be
         in self. Therefore `self & other = other`.
 
         This boldly assumes the other contains valid revs only.
         """
         # other not a smartset, make is so
         if not util.safehasattr(other, 'isascending'):
             # filter out hidden revision
             # (this boldly assumes all smartset are pure)
             #
             # `other` was used with "&", let's assume this is a set like
             # object.
             other = baseset(other - self._hiddenrevs)
 
         other.sort(reverse=self.isdescending())
         return other
 
 def prettyformat(revs):
     lines = []
     rs = repr(revs)
     p = 0
     while p < len(rs):
         q = rs.find('<', p + 1)
         if q < 0:
             q = len(rs)
         l = rs.count('<', 0, p) - rs.count('>', 0, p)
         assert l >= 0
         lines.append((l, rs[p:q].rstrip()))
         p = q
     return '\n'.join('  ' * l + s for l, s in lines)

tests/test-revset.t

@@ -1,3922 +1,3935
   $ HGENCODING=utf-8
   $ export HGENCODING
   $ cat > testrevset.py << EOF
   > import mercurial.revset
   > 
   > baseset = mercurial.revset.baseset
   > 
   > def r3232(repo, subset, x):
   >     """"simple revset that return [3,2,3,2]
   > 
   >     revisions duplicated on purpose.
   >     """
   >     if 3 not in subset:
   >        if 2 in subset:
   >            return baseset([2,2])
   >        return baseset()
   >     return baseset([3,3,2,2])
   > 
   > mercurial.revset.symbols['r3232'] = r3232
   > EOF
   $ cat >> $HGRCPATH << EOF
   > [extensions]
   > testrevset=$TESTTMP/testrevset.py
   > EOF
 
   $ try() {
   >   hg debugrevspec --debug "$@"
   > }
 
   $ log() {
   >   hg log --template '{rev}\n' -r "$1"
   > }
 
 extension to build '_intlist()' and '_hexlist()', which is necessary because
 these predicates use '\0' as a separator:
 
   $ cat <<EOF > debugrevlistspec.py
   > from __future__ import absolute_import
   > from mercurial import (
   >     node as nodemod,
   >     registrar,
   >     revset,
   >     revsetlang,
   >     smartset,
   > )
   > cmdtable = {}
   > command = registrar.command(cmdtable)
   > @command('debugrevlistspec',
   >     [('', 'optimize', None, 'print parsed tree after optimizing'),
   >      ('', 'bin', None, 'unhexlify arguments')])
   > def debugrevlistspec(ui, repo, fmt, *args, **opts):
   >     if opts['bin']:
   >         args = map(nodemod.bin, args)
   >     expr = revsetlang.formatspec(fmt, list(args))
   >     if ui.verbose:
   >         tree = revsetlang.parse(expr, lookup=repo.__contains__)
   >         ui.note(revsetlang.prettyformat(tree), "\n")
   >         if opts["optimize"]:
   >             opttree = revsetlang.optimize(revsetlang.analyze(tree))
   >             ui.note("* optimized:\n", revsetlang.prettyformat(opttree),
   >                     "\n")
   >     func = revset.match(ui, expr, repo)
   >     revs = func(repo)
   >     if ui.verbose:
   >         ui.note("* set:\n", smartset.prettyformat(revs), "\n")
   >     for c in revs:
   >         ui.write("%s\n" % c)
   > EOF
   $ cat <<EOF >> $HGRCPATH
   > [extensions]
   > debugrevlistspec = $TESTTMP/debugrevlistspec.py
   > EOF
   $ trylist() {
   >   hg debugrevlistspec --debug "$@"
   > }
 
   $ hg init repo
   $ cd repo
 
   $ echo a > a
   $ hg branch a
   marked working directory as branch a
   (branches are permanent and global, did you want a bookmark?)
   $ hg ci -Aqm0
 
   $ echo b > b
   $ hg branch b
   marked working directory as branch b
   $ hg ci -Aqm1
 
   $ rm a
   $ hg branch a-b-c-
   marked working directory as branch a-b-c-
   $ hg ci -Aqm2 -u Bob
 
   $ hg log -r "extra('branch', 'a-b-c-')" --template '{rev}\n'
   2
   $ hg log -r "extra('branch')" --template '{rev}\n'
   0
   1
   2
   $ hg log -r "extra('branch', 're:a')" --template '{rev} {branch}\n'
   0 a
   2 a-b-c-
 
   $ hg co 1
   1 files updated, 0 files merged, 0 files removed, 0 files unresolved
   $ hg branch +a+b+c+
   marked working directory as branch +a+b+c+
   $ hg ci -Aqm3
 
   $ hg co 2  # interleave
   0 files updated, 0 files merged, 1 files removed, 0 files unresolved
   $ echo bb > b
   $ hg branch -- -a-b-c-
   marked working directory as branch -a-b-c-
   $ hg ci -Aqm4 -d "May 12 2005"
 
   $ hg co 3
   2 files updated, 0 files merged, 0 files removed, 0 files unresolved
   $ hg branch !a/b/c/
   marked working directory as branch !a/b/c/
   $ hg ci -Aqm"5 bug"
 
   $ hg merge 4
   1 files updated, 0 files merged, 1 files removed, 0 files unresolved
   (branch merge, don't forget to commit)
   $ hg branch _a_b_c_
   marked working directory as branch _a_b_c_
   $ hg ci -Aqm"6 issue619"
 
   $ hg branch .a.b.c.
   marked working directory as branch .a.b.c.
   $ hg ci -Aqm7
 
   $ hg branch all
   marked working directory as branch all
 
   $ hg co 4
   0 files updated, 0 files merged, 0 files removed, 0 files unresolved
   $ hg branch é
   marked working directory as branch \xc3\xa9 (esc)
   $ hg ci -Aqm9
 
   $ hg tag -r6 1.0
   $ hg bookmark -r6 xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
 
   $ hg clone --quiet -U -r 7 . ../remote1
   $ hg clone --quiet -U -r 8 . ../remote2
   $ echo "[paths]" >> .hg/hgrc
   $ echo "default = ../remote1" >> .hg/hgrc
 
 trivial
 
   $ try 0:1
   (range
     ('symbol', '0')
     ('symbol', '1'))
   * set:
   <spanset+ 0:2>
   0
   1
   $ try --optimize :
   (rangeall
     None)
   * optimized:
   (rangeall
     None
     define)
   * set:
   <spanset+ 0:10>
   0
   1
   2
   3
   4
   5
   6
   7
   8
   9
   $ try 3::6
   (dagrange
     ('symbol', '3')
     ('symbol', '6'))
   * set:
   <baseset+ [3, 5, 6]>
   3
   5
   6
   $ try '0|1|2'
   (or
     (list
       ('symbol', '0')
       ('symbol', '1')
       ('symbol', '2')))
   * set:
   <baseset [0, 1, 2]>
   0
   1
   2
 
 names that should work without quoting
 
   $ try a
   ('symbol', 'a')
   * set:
   <baseset [0]>
   0
   $ try b-a
   (minus
     ('symbol', 'b')
     ('symbol', 'a'))
   * set:
   <filteredset
     <baseset [1]>,
     <not
       <baseset [0]>>>
   1
   $ try _a_b_c_
   ('symbol', '_a_b_c_')
   * set:
   <baseset [6]>
   6
   $ try _a_b_c_-a
   (minus
     ('symbol', '_a_b_c_')
     ('symbol', 'a'))
   * set:
   <filteredset
     <baseset [6]>,
     <not
       <baseset [0]>>>
   6
   $ try .a.b.c.
   ('symbol', '.a.b.c.')
   * set:
   <baseset [7]>
   7
   $ try .a.b.c.-a
   (minus
     ('symbol', '.a.b.c.')
     ('symbol', 'a'))
   * set:
   <filteredset
     <baseset [7]>,
     <not
       <baseset [0]>>>
   7
 
 names that should be caught by fallback mechanism
 
   $ try -- '-a-b-c-'
   ('symbol', '-a-b-c-')
   * set:
   <baseset [4]>
   4
   $ log -a-b-c-
   4
   $ try '+a+b+c+'
   ('symbol', '+a+b+c+')
   * set:
   <baseset [3]>
   3
   $ try '+a+b+c+:'
   (rangepost
     ('symbol', '+a+b+c+'))
   * set:
   <spanset+ 3:10>
   3
   4
   5
   6
   7
   8
   9
   $ try ':+a+b+c+'
   (rangepre
     ('symbol', '+a+b+c+'))
   * set:
   <spanset+ 0:4>
   0
   1
   2
   3
   $ try -- '-a-b-c-:+a+b+c+'
   (range
     ('symbol', '-a-b-c-')
     ('symbol', '+a+b+c+'))
   * set:
   <spanset- 3:5>
   4
   3
   $ log '-a-b-c-:+a+b+c+'
   4
   3
 
   $ try -- -a-b-c--a # complains
   (minus
     (minus
       (minus
         (negate
           ('symbol', 'a'))
         ('symbol', 'b'))
       ('symbol', 'c'))
     (negate
       ('symbol', 'a')))
   abort: unknown revision '-a'!
   [255]
   $ try é
   ('symbol', '\xc3\xa9')
   * set:
   <baseset [9]>
   9
 
 no quoting needed
 
   $ log ::a-b-c-
   0
   1
   2
 
 quoting needed
 
   $ try '"-a-b-c-"-a'
   (minus
     ('string', '-a-b-c-')
     ('symbol', 'a'))
   * set:
   <filteredset
     <baseset [4]>,
     <not
       <baseset [0]>>>
   4
 
   $ log '1 or 2'
   1
   2
   $ log '1|2'
   1
   2
   $ log '1 and 2'
   $ log '1&2'
   $ try '1&2|3' # precedence - and is higher
   (or
     (list
       (and
         ('symbol', '1')
         ('symbol', '2'))
       ('symbol', '3')))
   * set:
   <addset
     <baseset []>,
     <baseset [3]>>
   3
   $ try '1|2&3'
   (or
     (list
       ('symbol', '1')
       (and
         ('symbol', '2')
         ('symbol', '3'))))
   * set:
   <addset
     <baseset [1]>,
     <baseset []>>
   1
   $ try '1&2&3' # associativity
   (and
     (and
       ('symbol', '1')
       ('symbol', '2'))
     ('symbol', '3'))
   * set:
   <baseset []>
   $ try '1|(2|3)'
   (or
     (list
       ('symbol', '1')
       (group
         (or
           (list
             ('symbol', '2')
             ('symbol', '3'))))))
   * set:
   <addset
     <baseset [1]>,
     <baseset [2, 3]>>
   1
   2
   3
   $ log '1.0' # tag
   6
   $ log 'a' # branch
   0
   $ log '2785f51ee'
   0
   $ log 'date(2005)'
   4
   $ log 'date(this is a test)'
   hg: parse error at 10: unexpected token: symbol
   [255]
   $ log 'date()'
   hg: parse error: date requires a string
   [255]
   $ log 'date'
   abort: unknown revision 'date'!
   [255]
   $ log 'date('
   hg: parse error at 5: not a prefix: end
   [255]
   $ log 'date("\xy")'
   hg: parse error: invalid \x escape
   [255]
   $ log 'date(tip)'
   hg: parse error: invalid date: 'tip'
   [255]
   $ log '0:date'
   abort: unknown revision 'date'!
   [255]
   $ log '::"date"'
   abort: unknown revision 'date'!
   [255]
   $ hg book date -r 4
   $ log '0:date'
   0
   1
   2
   3
   4
   $ log '::date'
   0
   1
   2
   4
   $ log '::"date"'
   0
   1
   2
   4
   $ log 'date(2005) and 1::'
   4
   $ hg book -d date
 
 function name should be a symbol
 
   $ log '"date"(2005)'
   hg: parse error: not a symbol
   [255]
 
 keyword arguments
 
   $ log 'extra(branch, value=a)'
   0
 
   $ log 'extra(branch, a, b)'
   hg: parse error: extra takes at most 2 positional arguments
   [255]
   $ log 'extra(a, label=b)'
   hg: parse error: extra got multiple values for keyword argument 'label'
   [255]
   $ log 'extra(label=branch, default)'
   hg: parse error: extra got an invalid argument
   [255]
   $ log 'extra(branch, foo+bar=baz)'
   hg: parse error: extra got an invalid argument
   [255]
   $ log 'extra(unknown=branch)'
   hg: parse error: extra got an unexpected keyword argument 'unknown'
   [255]
 
   $ try 'foo=bar|baz'
   (keyvalue
     ('symbol', 'foo')
     (or
       (list
         ('symbol', 'bar')
         ('symbol', 'baz'))))
   hg: parse error: can't use a key-value pair in this context
   [255]
 
  right-hand side should be optimized recursively
 
   $ try --optimize 'foo=(not public())'
   (keyvalue
     ('symbol', 'foo')
     (group
       (not
         (func
           ('symbol', 'public')
           None))))
   * optimized:
   (keyvalue
     ('symbol', 'foo')
     (func
       ('symbol', '_notpublic')
       None
       any))
   hg: parse error: can't use a key-value pair in this context
   [255]
 
 parsed tree at stages:
 
   $ hg debugrevspec -p all '()'
   * parsed:
   (group
     None)
   * expanded:
   (group
     None)
   * concatenated:
   (group
     None)
   * analyzed:
   None
   * optimized:
   None
   hg: parse error: missing argument
   [255]
 
   $ hg debugrevspec --no-optimized -p all '()'
   * parsed:
   (group
     None)
   * expanded:
   (group
     None)
   * concatenated:
   (group
     None)
   * analyzed:
   None
   hg: parse error: missing argument
   [255]
 
   $ hg debugrevspec -p parsed -p analyzed -p optimized '(0|1)-1'
   * parsed:
   (minus
     (group
       (or
         (list
           ('symbol', '0')
           ('symbol', '1'))))
     ('symbol', '1'))
   * analyzed:
   (and
     (or
       (list
         ('symbol', '0')
         ('symbol', '1'))
       define)
     (not
       ('symbol', '1')
       follow)
     define)
   * optimized:
   (difference
     (func
       ('symbol', '_list')
       ('string', '0\x001')
       define)
     ('symbol', '1')
     define)
   0
 
   $ hg debugrevspec -p unknown '0'
   abort: invalid stage name: unknown
   [255]
 
   $ hg debugrevspec -p all --optimize '0'
   abort: cannot use --optimize with --show-stage
   [255]
 
 verify optimized tree:
 
   $ hg debugrevspec --verify '0|1'
 
   $ hg debugrevspec --verify -v -p analyzed -p optimized 'r3232() & 2'
   * analyzed:
   (and
     (func
       ('symbol', 'r3232')
       None
       define)
     ('symbol', '2')
     define)
   * optimized:
   (and
     ('symbol', '2')
     (func
       ('symbol', 'r3232')
       None
       define)
     define)
   * analyzed set:
   <baseset [2]>
   * optimized set:
   <baseset [2, 2]>
   --- analyzed
   +++ optimized
    2
   +2
   [1]
 
   $ hg debugrevspec --no-optimized --verify-optimized '0'
   abort: cannot use --verify-optimized with --no-optimized
   [255]
 
 Test that symbols only get parsed as functions if there's an opening
 parenthesis.
 
   $ hg book only -r 9
   $ log 'only(only)'   # Outer "only" is a function, inner "only" is the bookmark
   8
   9
 
 ':y' behaves like '0:y', but can't be rewritten as such since the revision '0'
 may be hidden (issue5385)
 
   $ try -p parsed -p analyzed ':'
   * parsed:
   (rangeall
     None)
   * analyzed:
   (rangeall
     None
     define)
   * set:
   <spanset+ 0:10>
   0
   1
   2
   3
   4
   5
   6
   7
   8
   9
   $ try -p analyzed ':1'
   * analyzed:
   (rangepre
     ('symbol', '1')
     define)
   * set:
   <spanset+ 0:2>
   0
   1
   $ try -p analyzed ':(1|2)'
   * analyzed:
   (rangepre
     (or
       (list
         ('symbol', '1')
         ('symbol', '2'))
       define)
     define)
   * set:
   <spanset+ 0:3>
   0
   1
   2
   $ try -p analyzed ':(1&2)'
   * analyzed:
   (rangepre
     (and
       ('symbol', '1')
       ('symbol', '2')
       define)
     define)
   * set:
   <baseset []>
 
 infix/suffix resolution of ^ operator (issue2884):
 
  x^:y means (x^):y
 
   $ try '1^:2'
   (range
     (parentpost
       ('symbol', '1'))
     ('symbol', '2'))
   * set:
   <spanset+ 0:3>
   0
   1
   2
 
   $ try '1^::2'
   (dagrange
     (parentpost
       ('symbol', '1'))
     ('symbol', '2'))
   * set:
   <baseset+ [0, 1, 2]>
   0
   1
   2
 
   $ try '9^:'
   (rangepost
     (parentpost
       ('symbol', '9')))
   * set:
   <spanset+ 8:10>
   8
   9
 
  x^:y should be resolved before omitting group operators
 
   $ try '1^(:2)'
   (parent
     ('symbol', '1')
     (group
       (rangepre
         ('symbol', '2'))))
   hg: parse error: ^ expects a number 0, 1, or 2
   [255]
 
  x^:y should be resolved recursively
 
   $ try 'sort(1^:2)'
   (func
     ('symbol', 'sort')
     (range
       (parentpost
         ('symbol', '1'))
       ('symbol', '2')))
   * set:
   <spanset+ 0:3>
   0
   1
   2
 
   $ try '(3^:4)^:2'
   (range
     (parentpost
       (group
         (range
           (parentpost
             ('symbol', '3'))
           ('symbol', '4'))))
     ('symbol', '2'))
   * set:
   <spanset+ 0:3>
   0
   1
   2
 
   $ try '(3^::4)^::2'
   (dagrange
     (parentpost
       (group
         (dagrange
           (parentpost
             ('symbol', '3'))
           ('symbol', '4'))))
     ('symbol', '2'))
   * set:
   <baseset+ [0, 1, 2]>
   0
   1
   2
 
   $ try '(9^:)^:'
   (rangepost
     (parentpost
       (group
         (rangepost
           (parentpost
             ('symbol', '9'))))))
   * set:
   <spanset+ 4:10>
   4
   5
   6
   7
   8
   9
 
  x^ in alias should also be resolved
 
   $ try 'A' --config 'revsetalias.A=1^:2'
   ('symbol', 'A')
   * expanded:
   (range
     (parentpost
       ('symbol', '1'))
     ('symbol', '2'))
   * set:
   <spanset+ 0:3>
   0
   1
   2
 
   $ try 'A:2' --config 'revsetalias.A=1^'
   (range
     ('symbol', 'A')
     ('symbol', '2'))
   * expanded:
   (range
     (parentpost
       ('symbol', '1'))
     ('symbol', '2'))
   * set:
   <spanset+ 0:3>
   0
   1
   2
 
  but not beyond the boundary of alias expansion, because the resolution should
  be made at the parsing stage
 
   $ try '1^A' --config 'revsetalias.A=:2'
   (parent
     ('symbol', '1')
     ('symbol', 'A'))
   * expanded:
   (parent
     ('symbol', '1')
     (rangepre
       ('symbol', '2')))
   hg: parse error: ^ expects a number 0, 1, or 2
   [255]
 
 ancestor can accept 0 or more arguments
 
   $ log 'ancestor()'
   $ log 'ancestor(1)'
   1
   $ log 'ancestor(4,5)'
   1
   $ log 'ancestor(4,5) and 4'
   $ log 'ancestor(0,0,1,3)'
   0
   $ log 'ancestor(3,1,5,3,5,1)'
   1
   $ log 'ancestor(0,1,3,5)'
   0
   $ log 'ancestor(1,2,3,4,5)'
   1
 
 test ancestors
 
   $ log 'ancestors(5)'
   0
   1
   3
   5
   $ log 'ancestor(ancestors(5))'
   0
   $ log '::r3232()'
   0
   1
   2
   3
 
   $ log 'author(bob)'
   2
   $ log 'author("re:bob|test")'
   0
   1
   2
   3
   4
   5
   6
   7
   8
   9
   $ log 'author(r"re:\S")'
   0
   1
   2
   3
   4
   5
   6
   7
   8
   9
   $ log 'branch(é)'
   8
   9
   $ log 'branch(a)'
   0
   $ hg log -r 'branch("re:a")' --template '{rev} {branch}\n'
   0 a
   2 a-b-c-
   3 +a+b+c+
   4 -a-b-c-
   5 !a/b/c/
   6 _a_b_c_
   7 .a.b.c.
   $ log 'children(ancestor(4,5))'
   2
   3
 
   $ log 'children(4)'
   6
   8
   $ log 'children(null)'
   0
 
   $ log 'closed()'
   $ log 'contains(a)'
   0
   1
   3
   5
   $ log 'contains("../repo/a")'
   0
   1
   3
   5
   $ log 'desc(B)'
   5
   $ hg log -r 'desc(r"re:S?u")' --template "{rev} {desc|firstline}\n"
   5 5 bug
   6 6 issue619
   $ log 'descendants(2 or 3)'
   2
   3
   4
   5
   6
   7
   8
   9
   $ log 'file("b*")'
   1
   4
   $ log 'filelog("b")'
   1
   4
   $ log 'filelog("../repo/b")'
   1
   4
   $ log 'follow()'
   0
   1
   2
   4
   8
   9
   $ log 'grep("issue\d+")'
   6
   $ try 'grep("(")' # invalid regular expression
   (func
     ('symbol', 'grep')
     ('string', '('))
   hg: parse error: invalid match pattern: unbalanced parenthesis
   [255]
   $ try 'grep("\bissue\d+")'
   (func
     ('symbol', 'grep')
     ('string', '\x08issue\\d+'))
   * set:
   <filteredset
     <fullreposet+ 0:10>,
     <grep '\x08issue\\d+'>>
   $ try 'grep(r"\bissue\d+")'
   (func
     ('symbol', 'grep')
     ('string', '\\bissue\\d+'))
   * set:
   <filteredset
     <fullreposet+ 0:10>,
     <grep '\\bissue\\d+'>>
   6
   $ try 'grep(r"\")'
   hg: parse error at 7: unterminated string
   [255]
   $ log 'head()'
   0
   1
   2
   3
   4
   5
   6
   7
   9
   $ log 'heads(6::)'
   7
   $ log 'keyword(issue)'
   6
   $ log 'keyword("test a")'
 
 Test first (=limit) and last
 
   $ log 'limit(head(), 1)'
   0
   $ log 'limit(author("re:bob|test"), 3, 5)'
   5
   6
   7
   $ log 'limit(author("re:bob|test"), offset=6)'
   6
   $ log 'limit(author("re:bob|test"), offset=10)'
   $ log 'limit(all(), 1, -1)'
   hg: parse error: negative offset
   [255]
   $ log 'limit(all(), -1)'
   hg: parse error: negative number to select
   [255]
   $ log 'limit(all(), 0)'
 
   $ log 'last(all(), -1)'
   hg: parse error: negative number to select
   [255]
   $ log 'last(all(), 0)'
   $ log 'last(all(), 1)'
   9
   $ log 'last(all(), 2)'
   8
   9
 
+Test smartset.slice() by first/last()
+
+ (using unoptimized set, filteredset as example)
+
+  $ hg debugrevspec --no-show-revs -s '0:7 & branch("re:")'
+  * set:
+  <filteredset
+    <spanset+ 0:8>,
+    <branch 're:'>>
+  $ log 'limit(0:7 & branch("re:"), 3, 4)'
+  4
+  5
+  6
+  $ log 'limit(7:0 & branch("re:"), 3, 4)'
+  3
+  2
+  1
+  $ log 'last(0:7 & branch("re:"), 2)'
+  6
+  7
+
 Test order of first/last revisions
 
   $ hg debugrevspec -s 'first(4:0, 3) & 3:'
   * set:
   <filteredset
-    <baseset
-      <limit n=3, offset=0,
-        <spanset- 0:5>>>,
+    <baseset slice=0:3
+      <spanset- 0:5>>,
     <spanset+ 3:10>>
   4
   3
 
   $ hg debugrevspec -s '3: & first(4:0, 3)'
   * set:
   <filteredset
     <spanset+ 3:10>,
-    <baseset
-      <limit n=3, offset=0,
-        <spanset- 0:5>>>>
+    <baseset slice=0:3
+      <spanset- 0:5>>>
   3
   4
 
   $ hg debugrevspec -s 'last(4:0, 3) & :1'
   * set:
   <filteredset
-    <baseset
-      <last n=3,
-        <spanset+ 0:5>>>,
+    <baseset slice=0:3
+      <spanset+ 0:5>>,
     <spanset+ 0:2>>
   1
   0
 
   $ hg debugrevspec -s ':1 & last(4:0, 3)'
   * set:
   <filteredset
     <spanset+ 0:2>,
-    <baseset
-      <last n=3,
-        <spanset+ 0:5>>>>
+    <baseset slice=0:3
+      <spanset+ 0:5>>>
   0
   1
 
 Test matching
 
   $ log 'matching(6)'
   6
   $ log 'matching(6:7, "phase parents user date branch summary files description substate")'
   6
   7
 
 Testing min and max
 
 max: simple
 
   $ log 'max(contains(a))'
   5
 
 max: simple on unordered set)
 
   $ log 'max((4+0+2+5+7) and contains(a))'
   5
 
 max: no result
 
   $ log 'max(contains(stringthatdoesnotappearanywhere))'
 
 max: no result on unordered set
 
   $ log 'max((4+0+2+5+7) and contains(stringthatdoesnotappearanywhere))'
 
 min: simple
 
   $ log 'min(contains(a))'
   0
 
 min: simple on unordered set
 
   $ log 'min((4+0+2+5+7) and contains(a))'
   0
 
 min: empty
 
   $ log 'min(contains(stringthatdoesnotappearanywhere))'
 
 min: empty on unordered set
 
   $ log 'min((4+0+2+5+7) and contains(stringthatdoesnotappearanywhere))'
 
 
   $ log 'merge()'
   6
   $ log 'branchpoint()'
   1
   4
   $ log 'modifies(b)'
   4
   $ log 'modifies("path:b")'
   4
   $ log 'modifies("*")'
   4
   6
   $ log 'modifies("set:modified()")'
   4
   $ log 'id(5)'
   2
   $ log 'only(9)'
   8
   9
   $ log 'only(8)'
   8
   $ log 'only(9, 5)'
   2
   4
   8
   9
   $ log 'only(7 + 9, 5 + 2)'
   4
   6
   7
   8
   9
 
 Test empty set input
   $ log 'only(p2())'
   $ log 'only(p1(), p2())'
   0
   1
   2
   4
   8
   9
 
 Test '%' operator
 
   $ log '9%'
   8
   9
   $ log '9%5'
   2
   4
   8
   9
   $ log '(7 + 9)%(5 + 2)'
   4
   6
   7
   8
   9
 
 Test operand of '%' is optimized recursively (issue4670)
 
   $ try --optimize '8:9-8%'
   (onlypost
     (minus
       (range
         ('symbol', '8')
         ('symbol', '9'))
       ('symbol', '8')))
   * optimized:
   (func
     ('symbol', 'only')
     (difference
       (range
         ('symbol', '8')
         ('symbol', '9')
         define)
       ('symbol', '8')
       define)
     define)
   * set:
   <baseset+ [8, 9]>
   8
   9
   $ try --optimize '(9)%(5)'
   (only
     (group
       ('symbol', '9'))
     (group
       ('symbol', '5')))
   * optimized:
   (func
     ('symbol', 'only')
     (list
       ('symbol', '9')
       ('symbol', '5'))
     define)
   * set:
   <baseset+ [2, 4, 8, 9]>
   2
   4
   8
   9
 
 Test the order of operations
 
   $ log '7 + 9%5 + 2'
   7
   2
   4
   8
   9
 
 Test explicit numeric revision
   $ log 'rev(-2)'
   $ log 'rev(-1)'
   -1
   $ log 'rev(0)'
   0
   $ log 'rev(9)'
   9
   $ log 'rev(10)'
   $ log 'rev(tip)'
   hg: parse error: rev expects a number
   [255]
 
 Test hexadecimal revision
   $ log 'id(2)'
   abort: 00changelog.i@2: ambiguous identifier!
   [255]
   $ log 'id(23268)'
   4
   $ log 'id(2785f51eece)'
   0
   $ log 'id(d5d0dcbdc4d9ff5dbb2d336f32f0bb561c1a532c)'
   8
   $ log 'id(d5d0dcbdc4a)'
   $ log 'id(d5d0dcbdc4w)'
   $ log 'id(d5d0dcbdc4d9ff5dbb2d336f32f0bb561c1a532d)'
   $ log 'id(d5d0dcbdc4d9ff5dbb2d336f32f0bb561c1a532q)'
   $ log 'id(1.0)'
   $ log 'id(xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx)'
 
 Test null revision
   $ log '(null)'
   -1
   $ log '(null:0)'
   -1
   0
   $ log '(0:null)'
   0
   -1
   $ log 'null::0'
   -1
   0
   $ log 'null:tip - 0:'
   -1
   $ log 'null: and null::' | head -1
   -1
   $ log 'null: or 0:' | head -2
   -1
   0
   $ log 'ancestors(null)'
   -1
   $ log 'reverse(null:)' | tail -2
   0
   -1
   $ log 'first(null:)'
   -1
   $ log 'min(null:)'
 BROKEN: should be '-1'
   $ log 'tip:null and all()' | tail -2
   1
   0
 
 Test working-directory revision
   $ hg debugrevspec 'wdir()'
   2147483647
   $ hg debugrevspec 'wdir()^'
   9
   $ hg up 7
   0 files updated, 0 files merged, 1 files removed, 0 files unresolved
   $ hg debugrevspec 'wdir()^'
   7
   $ hg debugrevspec 'wdir()^0'
   2147483647
   $ hg debugrevspec 'wdir()~3'
   5
   $ hg debugrevspec 'ancestors(wdir())'
   0
   1
   2
   3
   4
   5
   6
   7
   2147483647
   $ hg debugrevspec 'wdir()~0'
   2147483647
   $ hg debugrevspec 'p1(wdir())'
   7
   $ hg debugrevspec 'p2(wdir())'
   $ hg debugrevspec 'parents(wdir())'
   7
   $ hg debugrevspec 'wdir()^1'
   7
   $ hg debugrevspec 'wdir()^2'
   $ hg debugrevspec 'wdir()^3'
   hg: parse error: ^ expects a number 0, 1, or 2
   [255]
 For tests consistency
   $ hg up 9
   1 files updated, 0 files merged, 0 files removed, 0 files unresolved
   $ hg debugrevspec 'tip or wdir()'
   9
   2147483647
   $ hg debugrevspec '0:tip and wdir()'
   $ log '0:wdir()' | tail -3
   8
   9
   2147483647
   $ log 'wdir():0' | head -3
   2147483647
   9
   8
   $ log 'wdir():wdir()'
   2147483647
   $ log '(all() + wdir()) & min(. + wdir())'
   9
   $ log '(all() + wdir()) & max(. + wdir())'
   2147483647
   $ log 'first(wdir() + .)'
   2147483647
   $ log 'last(. + wdir())'
   2147483647
 
 Test working-directory integer revision and node id
 (BUG: '0:wdir()' is still needed to populate wdir revision)
 
   $ hg debugrevspec '0:wdir() & 2147483647'
   2147483647
   $ hg debugrevspec '0:wdir() & rev(2147483647)'
   2147483647
   $ hg debugrevspec '0:wdir() & ffffffffffffffffffffffffffffffffffffffff'
   2147483647
   $ hg debugrevspec '0:wdir() & ffffffffffff'
   2147483647
   $ hg debugrevspec '0:wdir() & id(ffffffffffffffffffffffffffffffffffffffff)'
   2147483647
   $ hg debugrevspec '0:wdir() & id(ffffffffffff)'
   2147483647
 
   $ cd ..
 
 Test short 'ff...' hash collision
 (BUG: '0:wdir()' is still needed to populate wdir revision)
 
   $ hg init wdir-hashcollision
   $ cd wdir-hashcollision
   $ cat <<EOF >> .hg/hgrc
   > [experimental]
   > evolution = createmarkers
   > EOF
   $ echo 0 > a
   $ hg ci -qAm 0
   $ for i in 2463 2961 6726 78127; do
   >   hg up -q 0
   >   echo $i > a
   >   hg ci -qm $i
   > done
   $ hg up -q null
   $ hg log -r '0:wdir()' -T '{rev}:{node} {shortest(node, 3)}\n'
   0:b4e73ffab476aa0ee32ed81ca51e07169844bc6a b4e
   1:fffbae3886c8fbb2114296380d276fd37715d571 fffba
   2:fffb6093b00943f91034b9bdad069402c834e572 fffb6
   3:fff48a9b9de34a4d64120c29548214c67980ade3 fff4
   4:ffff85cff0ff78504fcdc3c0bc10de0c65379249 ffff8
   2147483647:ffffffffffffffffffffffffffffffffffffffff fffff
   $ hg debugobsolete fffbae3886c8fbb2114296380d276fd37715d571
 
   $ hg debugrevspec '0:wdir() & fff'
   abort: 00changelog.i@fff: ambiguous identifier!
   [255]
   $ hg debugrevspec '0:wdir() & ffff'
   abort: 00changelog.i@ffff: ambiguous identifier!
   [255]
   $ hg debugrevspec '0:wdir() & fffb'
   abort: 00changelog.i@fffb: ambiguous identifier!
   [255]
 BROKEN should be '2' (node lookup uses unfiltered repo since dc25ed84bee8)
   $ hg debugrevspec '0:wdir() & id(fffb)'
   2
   $ hg debugrevspec '0:wdir() & ffff8'
   4
   $ hg debugrevspec '0:wdir() & fffff'
   2147483647
 
   $ cd ..
 
 Test branch() with wdir()
 
   $ cd repo
 
   $ log '0:wdir() & branch("literal:é")'
   8
   9
   2147483647
   $ log '0:wdir() & branch("re:é")'
   8
   9
   2147483647
   $ log '0:wdir() & branch("re:^a")'
   0
   2
   $ log '0:wdir() & branch(8)'
   8
   9
   2147483647
 
 branch(wdir()) returns all revisions belonging to the working branch. The wdir
 itself isn't returned unless it is explicitly populated.
 
   $ log 'branch(wdir())'
   8
   9
   $ log '0:wdir() & branch(wdir())'
   8
   9
   2147483647
 
   $ log 'outgoing()'
   8
   9
   $ log 'outgoing("../remote1")'
   8
   9
   $ log 'outgoing("../remote2")'
   3
   5
   6
   7
   9
   $ log 'p1(merge())'
   5
   $ log 'p2(merge())'
   4
   $ log 'parents(merge())'
   4
   5
   $ log 'p1(branchpoint())'
   0
   2
   $ log 'p2(branchpoint())'
   $ log 'parents(branchpoint())'
   0
   2
   $ log 'removes(a)'
   2
   6
   $ log 'roots(all())'
   0
   $ log 'reverse(2 or 3 or 4 or 5)'
   5
   4
   3
   2
   $ log 'reverse(all())'
   9
   8
   7
   6
   5
   4
   3
   2
   1
   0
   $ log 'reverse(all()) & filelog(b)'
   4
   1
   $ log 'rev(5)'
   5
   $ log 'sort(limit(reverse(all()), 3))'
   7
   8
   9
   $ log 'sort(2 or 3 or 4 or 5, date)'
   2
   3
   5
   4
   $ log 'tagged()'
   6
   $ log 'tag()'
   6
   $ log 'tag(1.0)'
   6
   $ log 'tag(tip)'
   9
 
 Test order of revisions in compound expression
 ----------------------------------------------
 
 The general rule is that only the outermost (= leftmost) predicate can
 enforce its ordering requirement. The other predicates should take the
 ordering defined by it.
 
  'A & B' should follow the order of 'A':
 
   $ log '2:0 & 0::2'
   2
   1
   0
 
  'head()' combines sets in right order:
 
   $ log '2:0 & head()'
   2
   1
   0
 
  'x:y' takes ordering parameter into account:
 
   $ try -p optimized '3:0 & 0:3 & not 2:1'
   * optimized:
   (difference
     (and
       (range
         ('symbol', '3')
         ('symbol', '0')
         define)
       (range
         ('symbol', '0')
         ('symbol', '3')
         follow)
       define)
     (range
       ('symbol', '2')
       ('symbol', '1')
       any)
     define)
   * set:
   <filteredset
     <filteredset
       <spanset- 0:4>,
       <spanset+ 0:4>>,
     <not
       <spanset+ 1:3>>>
   3
   0
 
  'a + b', which is optimized to '_list(a b)', should take the ordering of
  the left expression:
 
   $ try --optimize '2:0 & (0 + 1 + 2)'
   (and
     (range
       ('symbol', '2')
       ('symbol', '0'))
     (group
       (or
         (list
           ('symbol', '0')
           ('symbol', '1')
           ('symbol', '2')))))
   * optimized:
   (and
     (range
       ('symbol', '2')
       ('symbol', '0')
       define)
     (func
       ('symbol', '_list')
       ('string', '0\x001\x002')
       follow)
     define)
   * set:
   <filteredset
     <spanset- 0:3>,
     <baseset [0, 1, 2]>>
   2
   1
   0
 
  'A + B' should take the ordering of the left expression:
 
   $ try --optimize '2:0 & (0:1 + 2)'
   (and
     (range
       ('symbol', '2')
       ('symbol', '0'))
     (group
       (or
         (list
           (range
             ('symbol', '0')
             ('symbol', '1'))
           ('symbol', '2')))))
   * optimized:
   (and
     (range
       ('symbol', '2')
       ('symbol', '0')
       define)
     (or
       (list
         ('symbol', '2')
         (range
           ('symbol', '0')
           ('symbol', '1')
           follow))
       follow)
     define)
   * set:
   <filteredset
     <spanset- 0:3>,
     <addset
       <baseset [2]>,
       <spanset+ 0:2>>>
   2
   1
   0
 
  '_intlist(a b)' should behave like 'a + b':
 
   $ trylist --optimize '2:0 & %ld' 0 1 2
   (and
     (range
       ('symbol', '2')
       ('symbol', '0'))
     (func
       ('symbol', '_intlist')
       ('string', '0\x001\x002')))
   * optimized:
   (and
     (func
       ('symbol', '_intlist')
       ('string', '0\x001\x002')
       follow)
     (range
       ('symbol', '2')
       ('symbol', '0')
       define)
     define)
   * set:
   <filteredset
     <spanset- 0:3>,
     <baseset+ [0, 1, 2]>>
   2
   1
   0
 
   $ trylist --optimize '%ld & 2:0' 0 2 1
   (and
     (func
       ('symbol', '_intlist')
       ('string', '0\x002\x001'))
     (range
       ('symbol', '2')
       ('symbol', '0')))
   * optimized:
   (and
     (func
       ('symbol', '_intlist')
       ('string', '0\x002\x001')
       define)
     (range
       ('symbol', '2')
       ('symbol', '0')
       follow)
     define)
   * set:
   <filteredset
     <baseset [0, 2, 1]>,
     <spanset- 0:3>>
   0
   2
   1
 
  '_hexlist(a b)' should behave like 'a + b':
 
   $ trylist --optimize --bin '2:0 & %ln' `hg log -T '{node} ' -r0:2`
   (and
     (range
       ('symbol', '2')
       ('symbol', '0'))
     (func
       ('symbol', '_hexlist')
       ('string', '*'))) (glob)
   * optimized:
   (and
     (range
       ('symbol', '2')
       ('symbol', '0')
       define)
     (func
       ('symbol', '_hexlist')
       ('string', '*') (glob)
       follow)
     define)
   * set:
   <filteredset
     <spanset- 0:3>,
     <baseset [0, 1, 2]>>
   2
   1
   0
 
   $ trylist --optimize --bin '%ln & 2:0' `hg log -T '{node} ' -r0+2+1`
   (and
     (func
       ('symbol', '_hexlist')
       ('string', '*')) (glob)
     (range
       ('symbol', '2')
       ('symbol', '0')))
   * optimized:
   (and
     (range
       ('symbol', '2')
       ('symbol', '0')
       follow)
     (func
       ('symbol', '_hexlist')
       ('string', '*') (glob)
       define)
     define)
   * set:
   <baseset [0, 2, 1]>
   0
   2
   1
 
  '_list' should not go through the slow follow-order path if order doesn't
  matter:
 
   $ try -p optimized '2:0 & not (0 + 1)'
   * optimized:
   (difference
     (range
       ('symbol', '2')
       ('symbol', '0')
       define)
     (func
       ('symbol', '_list')
       ('string', '0\x001')
       any)
     define)
   * set:
   <filteredset
     <spanset- 0:3>,
     <not
       <baseset [0, 1]>>>
   2
 
   $ try -p optimized '2:0 & not (0:2 & (0 + 1))'
   * optimized:
   (difference
     (range
       ('symbol', '2')
       ('symbol', '0')
       define)
     (and
       (range
         ('symbol', '0')
         ('symbol', '2')
         any)
       (func
         ('symbol', '_list')
         ('string', '0\x001')
         any)
       any)
     define)
   * set:
   <filteredset
     <spanset- 0:3>,
     <not
       <baseset [0, 1]>>>
   2
 
  because 'present()' does nothing other than suppressing an error, the
  ordering requirement should be forwarded to the nested expression
 
   $ try -p optimized 'present(2 + 0 + 1)'
   * optimized:
   (func
     ('symbol', 'present')
     (func
       ('symbol', '_list')
       ('string', '2\x000\x001')
       define)
     define)
   * set:
   <baseset [2, 0, 1]>
   2
   0
   1
 
   $ try --optimize '2:0 & present(0 + 1 + 2)'
   (and
     (range
       ('symbol', '2')
       ('symbol', '0'))
     (func
       ('symbol', 'present')
       (or
         (list
           ('symbol', '0')
           ('symbol', '1')
           ('symbol', '2')))))
   * optimized:
   (and
     (range
       ('symbol', '2')
       ('symbol', '0')
       define)
     (func
       ('symbol', 'present')
       (func
         ('symbol', '_list')
         ('string', '0\x001\x002')
         follow)
       follow)
     define)
   * set:
   <filteredset
     <spanset- 0:3>,
     <baseset [0, 1, 2]>>
   2
   1
   0
 
  'reverse()' should take effect only if it is the outermost expression:
 
   $ try --optimize '0:2 & reverse(all())'
   (and
     (range
       ('symbol', '0')
       ('symbol', '2'))
     (func
       ('symbol', 'reverse')
       (func
         ('symbol', 'all')
         None)))
   * optimized:
   (and
     (range
       ('symbol', '0')
       ('symbol', '2')
       define)
     (func
       ('symbol', 'reverse')
       (func
         ('symbol', 'all')
         None
         define)
       follow)
     define)
   * set:
   <filteredset
     <spanset+ 0:3>,
     <spanset+ 0:10>>
   0
   1
   2
 
  'sort()' should take effect only if it is the outermost expression:
 
   $ try --optimize '0:2 & sort(all(), -rev)'
   (and
     (range
       ('symbol', '0')
       ('symbol', '2'))
     (func
       ('symbol', 'sort')
       (list
         (func
           ('symbol', 'all')
           None)
         (negate
           ('symbol', 'rev')))))
   * optimized:
   (and
     (range
       ('symbol', '0')
       ('symbol', '2')
       define)
     (func
       ('symbol', 'sort')
       (list
         (func
           ('symbol', 'all')
           None
           define)
         ('string', '-rev'))
       follow)
     define)
   * set:
   <filteredset
     <spanset+ 0:3>,
     <spanset+ 0:10>>
   0
   1
   2
 
  invalid argument passed to noop sort():
 
   $ log '0:2 & sort()'
   hg: parse error: sort requires one or two arguments
   [255]
   $ log '0:2 & sort(all(), -invalid)'
   hg: parse error: unknown sort key '-invalid'
   [255]
 
  for 'A & f(B)', 'B' should not be affected by the order of 'A':
 
   $ try --optimize '2:0 & first(1 + 0 + 2)'
   (and
     (range
       ('symbol', '2')
       ('symbol', '0'))
     (func
       ('symbol', 'first')
       (or
         (list
           ('symbol', '1')
           ('symbol', '0')
           ('symbol', '2')))))
   * optimized:
   (and
     (range
       ('symbol', '2')
       ('symbol', '0')
       define)
     (func
       ('symbol', 'first')
       (func
         ('symbol', '_list')
         ('string', '1\x000\x002')
         define)
       follow)
     define)
   * set:
   <filteredset
-    <baseset
-      <limit n=1, offset=0,
-        <baseset [1, 0, 2]>>>,
+    <baseset slice=0:1
+      <baseset [1, 0, 2]>>,
     <spanset- 0:3>>
   1
 
   $ try --optimize '2:0 & not last(0 + 2 + 1)'
   (and
     (range
       ('symbol', '2')
       ('symbol', '0'))
     (not
       (func
         ('symbol', 'last')
         (or
           (list
             ('symbol', '0')
             ('symbol', '2')
             ('symbol', '1'))))))
   * optimized:
   (difference
     (range
       ('symbol', '2')
       ('symbol', '0')
       define)
     (func
       ('symbol', 'last')
       (func
         ('symbol', '_list')
         ('string', '0\x002\x001')
         define)
       any)
     define)
   * set:
   <filteredset
     <spanset- 0:3>,
     <not
-      <baseset
-        <last n=1,
-          <baseset [1, 2, 0]>>>>>
+      <baseset slice=0:1
+        <baseset [1, 2, 0]>>>>
   2
   0
 
  for 'A & (op)(B)', 'B' should not be affected by the order of 'A':
 
   $ try --optimize '2:0 & (1 + 0 + 2):(0 + 2 + 1)'
   (and
     (range
       ('symbol', '2')
       ('symbol', '0'))
     (range
       (group
         (or
           (list
             ('symbol', '1')
             ('symbol', '0')
             ('symbol', '2'))))
       (group
         (or
           (list
             ('symbol', '0')
             ('symbol', '2')
             ('symbol', '1'))))))
   * optimized:
   (and
     (range
       ('symbol', '2')
       ('symbol', '0')
       define)
     (range
       (func
         ('symbol', '_list')
         ('string', '1\x000\x002')
         define)
       (func
         ('symbol', '_list')
         ('string', '0\x002\x001')
         define)
       follow)
     define)
   * set:
   <filteredset
     <spanset- 0:3>,
     <baseset [1]>>
   1
 
  'A & B' can be rewritten as 'B & A' by weight, but that's fine as long as
  the ordering rule is determined before the rewrite; in this example,
  'B' follows the order of the initial set, which is the same order as 'A'
  since 'A' also follows the order:
 
   $ try --optimize 'contains("glob:*") & (2 + 0 + 1)'
   (and
     (func
       ('symbol', 'contains')
       ('string', 'glob:*'))
     (group
       (or
         (list
           ('symbol', '2')
           ('symbol', '0')
           ('symbol', '1')))))
   * optimized:
   (and
     (func
       ('symbol', '_list')
       ('string', '2\x000\x001')
       follow)
     (func
       ('symbol', 'contains')
       ('string', 'glob:*')
       define)
     define)
   * set:
   <filteredset
     <baseset+ [0, 1, 2]>,
     <contains 'glob:*'>>
   0
   1
   2
 
  and in this example, 'A & B' is rewritten as 'B & A', but 'A' overrides
  the order appropriately:
 
   $ try --optimize 'reverse(contains("glob:*")) & (0 + 2 + 1)'
   (and
     (func
       ('symbol', 'reverse')
       (func
         ('symbol', 'contains')
         ('string', 'glob:*')))
     (group
       (or
         (list
           ('symbol', '0')
           ('symbol', '2')
           ('symbol', '1')))))
   * optimized:
   (and
     (func
       ('symbol', '_list')
       ('string', '0\x002\x001')
       follow)
     (func
       ('symbol', 'reverse')
       (func
         ('symbol', 'contains')
         ('string', 'glob:*')
         define)
       define)
     define)
   * set:
   <filteredset
     <baseset- [0, 1, 2]>,
     <contains 'glob:*'>>
   2
   1
   0
 
  'A + B' can be rewritten to 'B + A' by weight only when the order doesn't
  matter (e.g. 'X & (A + B)' can be 'X & (B + A)', but '(A + B) & X' can't):
 
   $ try -p optimized '0:2 & (reverse(contains("a")) + 2)'
   * optimized:
   (and
     (range
       ('symbol', '0')
       ('symbol', '2')
       define)
     (or
       (list
         ('symbol', '2')
         (func
           ('symbol', 'reverse')
           (func
             ('symbol', 'contains')
             ('string', 'a')
             define)
           follow))
       follow)
     define)
   * set:
   <filteredset
     <spanset+ 0:3>,
     <addset
       <baseset [2]>,
       <filteredset
         <fullreposet+ 0:10>,
         <contains 'a'>>>>
   0
   1
   2
 
   $ try -p optimized '(reverse(contains("a")) + 2) & 0:2'
   * optimized:
   (and
     (range
       ('symbol', '0')
       ('symbol', '2')
       follow)
     (or
       (list
         (func
           ('symbol', 'reverse')
           (func
             ('symbol', 'contains')
             ('string', 'a')
             define)
           define)
         ('symbol', '2'))
       define)
     define)
   * set:
   <addset
     <filteredset
       <spanset- 0:3>,
       <contains 'a'>>,
     <baseset [2]>>
   1
   0
   2
 
 test sort revset
 --------------------------------------------
 
 test when adding two unordered revsets
 
   $ log 'sort(keyword(issue) or modifies(b))'
   4
   6
 
 test when sorting a reversed collection in the same way it is
 
   $ log 'sort(reverse(all()), -rev)'
   9
   8
   7
   6
   5
   4
   3
   2
   1
   0
 
 test when sorting a reversed collection
 
   $ log 'sort(reverse(all()), rev)'
   0
   1
   2
   3
   4
   5
   6
   7
   8
   9
 
 
 test sorting two sorted collections in different orders
 
   $ log 'sort(outgoing() or reverse(removes(a)), rev)'
   2
   6
   8
   9
 
 test sorting two sorted collections in different orders backwards
 
   $ log 'sort(outgoing() or reverse(removes(a)), -rev)'
   9
   8
   6
   2
 
 test empty sort key which is noop
 
   $ log 'sort(0 + 2 + 1, "")'
   0
   2
   1
 
 test invalid sort keys
 
   $ log 'sort(all(), -invalid)'
   hg: parse error: unknown sort key '-invalid'
   [255]
 
   $ cd ..
 
 test sorting by multiple keys including variable-length strings
 
   $ hg init sorting
   $ cd sorting
   $ cat <<EOF >> .hg/hgrc
   > [ui]
   > logtemplate = '{rev} {branch|p5}{desc|p5}{author|p5}{date|hgdate}\n'
   > [templatealias]
   > p5(s) = pad(s, 5)
   > EOF
   $ hg branch -qf b12
   $ hg ci -m m111 -u u112 -d '111 10800'
   $ hg branch -qf b11
   $ hg ci -m m12 -u u111 -d '112 7200'
   $ hg branch -qf b111
   $ hg ci -m m11 -u u12 -d '111 3600'
   $ hg branch -qf b112
   $ hg ci -m m111 -u u11 -d '120 0'
   $ hg branch -qf b111
   $ hg ci -m m112 -u u111 -d '110 14400'
   created new head
 
  compare revisions (has fast path):
 
   $ hg log -r 'sort(all(), rev)'
   0 b12  m111 u112 111 10800
   1 b11  m12  u111 112 7200
   2 b111 m11  u12  111 3600
   3 b112 m111 u11  120 0
   4 b111 m112 u111 110 14400
 
   $ hg log -r 'sort(all(), -rev)'
   4 b111 m112 u111 110 14400
   3 b112 m111 u11  120 0
   2 b111 m11  u12  111 3600
   1 b11  m12  u111 112 7200
   0 b12  m111 u112 111 10800
 
  compare variable-length strings (issue5218):
 
   $ hg log -r 'sort(all(), branch)'
   1 b11  m12  u111 112 7200
   2 b111 m11  u12  111 3600
   4 b111 m112 u111 110 14400
   3 b112 m111 u11  120 0
   0 b12  m111 u112 111 10800
 
   $ hg log -r 'sort(all(), -branch)'
   0 b12  m111 u112 111 10800
   3 b112 m111 u11  120 0
   2 b111 m11  u12  111 3600
   4 b111 m112 u111 110 14400
   1 b11  m12  u111 112 7200
 
   $ hg log -r 'sort(all(), desc)'
   2 b111 m11  u12  111 3600
   0 b12  m111 u112 111 10800
   3 b112 m111 u11  120 0
   4 b111 m112 u111 110 14400
   1 b11  m12  u111 112 7200
 
   $ hg log -r 'sort(all(), -desc)'
   1 b11  m12  u111 112 7200
   4 b111 m112 u111 110 14400
   0 b12  m111 u112 111 10800
   3 b112 m111 u11  120 0
   2 b111 m11  u12  111 3600
 
   $ hg log -r 'sort(all(), user)'
   3 b112 m111 u11  120 0
   1 b11  m12  u111 112 7200
   4 b111 m112 u111 110 14400
   0 b12  m111 u112 111 10800
   2 b111 m11  u12  111 3600
 
   $ hg log -r 'sort(all(), -user)'
   2 b111 m11  u12  111 3600
   0 b12  m111 u112 111 10800
   1 b11  m12  u111 112 7200
   4 b111 m112 u111 110 14400
   3 b112 m111 u11  120 0
 
  compare dates (tz offset should have no effect):
 
   $ hg log -r 'sort(all(), date)'
   4 b111 m112 u111 110 14400
   0 b12  m111 u112 111 10800
   2 b111 m11  u12  111 3600
   1 b11  m12  u111 112 7200
   3 b112 m111 u11  120 0
 
   $ hg log -r 'sort(all(), -date)'
   3 b112 m111 u11  120 0
   1 b11  m12  u111 112 7200
   0 b12  m111 u112 111 10800
   2 b111 m11  u12  111 3600
   4 b111 m112 u111 110 14400
 
  be aware that 'sort(x, -k)' is not exactly the same as 'reverse(sort(x, k))'
  because '-k' reverses the comparison, not the list itself:
 
   $ hg log -r 'sort(0 + 2, date)'
   0 b12  m111 u112 111 10800
   2 b111 m11  u12  111 3600
 
   $ hg log -r 'sort(0 + 2, -date)'
   0 b12  m111 u112 111 10800
   2 b111 m11  u12  111 3600
 
   $ hg log -r 'reverse(sort(0 + 2, date))'
   2 b111 m11  u12  111 3600
   0 b12  m111 u112 111 10800
 
  sort by multiple keys:
 
   $ hg log -r 'sort(all(), "branch -rev")'
   1 b11  m12  u111 112 7200
   4 b111 m112 u111 110 14400
   2 b111 m11  u12  111 3600
   3 b112 m111 u11  120 0
   0 b12  m111 u112 111 10800
 
   $ hg log -r 'sort(all(), "-desc -date")'
   1 b11  m12  u111 112 7200
   4 b111 m112 u111 110 14400
   3 b112 m111 u11  120 0
   0 b12  m111 u112 111 10800
   2 b111 m11  u12  111 3600
 
   $ hg log -r 'sort(all(), "user -branch date rev")'
   3 b112 m111 u11  120 0
   4 b111 m112 u111 110 14400
   1 b11  m12  u111 112 7200
   0 b12  m111 u112 111 10800
   2 b111 m11  u12  111 3600
 
  toposort prioritises graph branches
 
   $ hg up 2
   0 files updated, 0 files merged, 0 files removed, 0 files unresolved
   $ touch a
   $ hg addremove
   adding a
   $ hg ci -m 't1' -u 'tu' -d '130 0'
   created new head
   $ echo 'a' >> a
   $ hg ci -m 't2' -u 'tu' -d '130 0'
   $ hg book book1
   $ hg up 4
   0 files updated, 0 files merged, 1 files removed, 0 files unresolved
   (leaving bookmark book1)
   $ touch a
   $ hg addremove
   adding a
   $ hg ci -m 't3' -u 'tu' -d '130 0'
 
   $ hg log -r 'sort(all(), topo)'
   7 b111 t3   tu   130 0
   4 b111 m112 u111 110 14400
   3 b112 m111 u11  120 0
   6 b111 t2   tu   130 0
   5 b111 t1   tu   130 0
   2 b111 m11  u12  111 3600
   1 b11  m12  u111 112 7200
   0 b12  m111 u112 111 10800
 
   $ hg log -r 'sort(all(), -topo)'
   0 b12  m111 u112 111 10800
   1 b11  m12  u111 112 7200
   2 b111 m11  u12  111 3600
   5 b111 t1   tu   130 0
   6 b111 t2   tu   130 0
   3 b112 m111 u11  120 0
   4 b111 m112 u111 110 14400
   7 b111 t3   tu   130 0
 
   $ hg log -r 'sort(all(), topo, topo.firstbranch=book1)'
   6 b111 t2   tu   130 0
   5 b111 t1   tu   130 0
   7 b111 t3   tu   130 0
   4 b111 m112 u111 110 14400
   3 b112 m111 u11  120 0
   2 b111 m11  u12  111 3600
   1 b11  m12  u111 112 7200
   0 b12  m111 u112 111 10800
 
 topographical sorting can't be combined with other sort keys, and you can't
 use the topo.firstbranch option when topo sort is not active:
 
   $ hg log -r 'sort(all(), "topo user")'
   hg: parse error: topo sort order cannot be combined with other sort keys
   [255]
 
   $ hg log -r 'sort(all(), user, topo.firstbranch=book1)'
   hg: parse error: topo.firstbranch can only be used when using the topo sort key
   [255]
 
 topo.firstbranch should accept any kind of expressions:
 
   $ hg log -r 'sort(0, topo, topo.firstbranch=(book1))'
   0 b12  m111 u112 111 10800
 
   $ cd ..
   $ cd repo
 
 test subtracting something from an addset
 
   $ log '(outgoing() or removes(a)) - removes(a)'
   8
   9
 
 test intersecting something with an addset
 
   $ log 'parents(outgoing() or removes(a))'
   1
   4
   5
   8
 
 test that `or` operation combines elements in the right order:
 
   $ log '3:4 or 2:5'
   3
   4
   2
   5
   $ log '3:4 or 5:2'
   3
   4
   5
   2
   $ log 'sort(3:4 or 2:5)'
   2
   3
   4
   5
   $ log 'sort(3:4 or 5:2)'
   2
   3
   4
   5
 
 test that more than one `-r`s are combined in the right order and deduplicated:
 
   $ hg log -T '{rev}\n' -r 3 -r 3 -r 4 -r 5:2 -r 'ancestors(4)'
   3
   4
   5
   2
   0
   1
 
 test that `or` operation skips duplicated revisions from right-hand side
 
   $ try 'reverse(1::5) or ancestors(4)'
   (or
     (list
       (func
         ('symbol', 'reverse')
         (dagrange
           ('symbol', '1')
           ('symbol', '5')))
       (func
         ('symbol', 'ancestors')
         ('symbol', '4'))))
   * set:
   <addset
     <baseset- [1, 3, 5]>,
     <generatorset+>>
   5
   3
   1
   0
   2
   4
   $ try 'sort(ancestors(4) or reverse(1::5))'
   (func
     ('symbol', 'sort')
     (or
       (list
         (func
           ('symbol', 'ancestors')
           ('symbol', '4'))
         (func
           ('symbol', 'reverse')
           (dagrange
             ('symbol', '1')
             ('symbol', '5'))))))
   * set:
   <addset+
     <generatorset+>,
     <baseset- [1, 3, 5]>>
   0
   1
   2
   3
   4
   5
 
 test optimization of trivial `or` operation
 
   $ try --optimize '0|(1)|"2"|-2|tip|null'
   (or
     (list
       ('symbol', '0')
       (group
         ('symbol', '1'))
       ('string', '2')
       (negate
         ('symbol', '2'))
       ('symbol', 'tip')
       ('symbol', 'null')))
   * optimized:
   (func
     ('symbol', '_list')
     ('string', '0\x001\x002\x00-2\x00tip\x00null')
     define)
   * set:
   <baseset [0, 1, 2, 8, 9, -1]>
   0
   1
   2
   8
   9
   -1
 
   $ try --optimize '0|1|2:3'
   (or
     (list
       ('symbol', '0')
       ('symbol', '1')
       (range
         ('symbol', '2')
         ('symbol', '3'))))
   * optimized:
   (or
     (list
       (func
         ('symbol', '_list')
         ('string', '0\x001')
         define)
       (range
         ('symbol', '2')
         ('symbol', '3')
         define))
     define)
   * set:
   <addset
     <baseset [0, 1]>,
     <spanset+ 2:4>>
   0
   1
   2
   3
 
   $ try --optimize '0:1|2|3:4|5|6'
   (or
     (list
       (range
         ('symbol', '0')
         ('symbol', '1'))
       ('symbol', '2')
       (range
         ('symbol', '3')
         ('symbol', '4'))
       ('symbol', '5')
       ('symbol', '6')))
   * optimized:
   (or
     (list
       (range
         ('symbol', '0')
         ('symbol', '1')
         define)
       ('symbol', '2')
       (range
         ('symbol', '3')
         ('symbol', '4')
         define)
       (func
         ('symbol', '_list')
         ('string', '5\x006')
         define))
     define)
   * set:
   <addset
     <addset
       <spanset+ 0:2>,
       <baseset [2]>>,
     <addset
       <spanset+ 3:5>,
       <baseset [5, 6]>>>
   0
   1
   2
   3
   4
   5
   6
 
 unoptimized `or` looks like this
 
   $ try --no-optimized -p analyzed '0|1|2|3|4'
   * analyzed:
   (or
     (list
       ('symbol', '0')
       ('symbol', '1')
       ('symbol', '2')
       ('symbol', '3')
       ('symbol', '4'))
     define)
   * set:
   <addset
     <addset
       <baseset [0]>,
       <baseset [1]>>,
     <addset
       <baseset [2]>,
       <addset
         <baseset [3]>,
         <baseset [4]>>>>
   0
   1
   2
   3
   4
 
 test that `_list` should be narrowed by provided `subset`
 
   $ log '0:2 and (null|1|2|3)'
   1
   2
 
 test that `_list` should remove duplicates
 
   $ log '0|1|2|1|2|-1|tip'
   0
   1
   2
   9
 
 test unknown revision in `_list`
 
   $ log '0|unknown'
   abort: unknown revision 'unknown'!
   [255]
 
 test integer range in `_list`
 
   $ log '-1|-10'
   9
   0
 
   $ log '-10|-11'
   abort: unknown revision '-11'!
   [255]
 
   $ log '9|10'
   abort: unknown revision '10'!
   [255]
 
 test '0000' != '0' in `_list`
 
   $ log '0|0000'
   0
   -1
 
 test ',' in `_list`
   $ log '0,1'
   hg: parse error: can't use a list in this context
   (see hg help "revsets.x or y")
   [255]
   $ try '0,1,2'
   (list
     ('symbol', '0')
     ('symbol', '1')
     ('symbol', '2'))
   hg: parse error: can't use a list in this context
   (see hg help "revsets.x or y")
   [255]
 
 test that chained `or` operations make balanced addsets
 
   $ try '0:1|1:2|2:3|3:4|4:5'
   (or
     (list
       (range
         ('symbol', '0')
         ('symbol', '1'))
       (range
         ('symbol', '1')
         ('symbol', '2'))
       (range
         ('symbol', '2')
         ('symbol', '3'))
       (range
         ('symbol', '3')
         ('symbol', '4'))
       (range
         ('symbol', '4')
         ('symbol', '5'))))
   * set:
   <addset
     <addset
       <spanset+ 0:2>,
       <spanset+ 1:3>>,
     <addset
       <spanset+ 2:4>,
       <addset
         <spanset+ 3:5>,
         <spanset+ 4:6>>>>
   0
   1
   2
   3
   4
   5
 
 no crash by empty group "()" while optimizing `or` operations
 
   $ try --optimize '0|()'
   (or
     (list
       ('symbol', '0')
       (group
         None)))
   * optimized:
   (or
     (list
       ('symbol', '0')
       None)
     define)
   hg: parse error: missing argument
   [255]
 
 test that chained `or` operations never eat up stack (issue4624)
 (uses `0:1` instead of `0` to avoid future optimization of trivial revisions)
 
   $ hg log -T '{rev}\n' -r `python -c "print '+'.join(['0:1'] * 500)"`
   0
   1
 
 test that repeated `-r` options never eat up stack (issue4565)
 (uses `-r 0::1` to avoid possible optimization at old-style parser)
 
   $ hg log -T '{rev}\n' `python -c "for i in xrange(500): print '-r 0::1 ',"`
   0
   1
 
 check that conversion to only works
   $ try --optimize '::3 - ::1'
   (minus
     (dagrangepre
       ('symbol', '3'))
     (dagrangepre
       ('symbol', '1')))
   * optimized:
   (func
     ('symbol', 'only')
     (list
       ('symbol', '3')
       ('symbol', '1'))
     define)
   * set:
   <baseset+ [3]>
   3
   $ try --optimize 'ancestors(1) - ancestors(3)'
   (minus
     (func
       ('symbol', 'ancestors')
       ('symbol', '1'))
     (func
       ('symbol', 'ancestors')
       ('symbol', '3')))
   * optimized:
   (func
     ('symbol', 'only')
     (list
       ('symbol', '1')
       ('symbol', '3'))
     define)
   * set:
   <baseset+ []>
   $ try --optimize 'not ::2 and ::6'
   (and
     (not
       (dagrangepre
         ('symbol', '2')))
     (dagrangepre
       ('symbol', '6')))
   * optimized:
   (func
     ('symbol', 'only')
     (list
       ('symbol', '6')
       ('symbol', '2'))
     define)
   * set:
   <baseset+ [3, 4, 5, 6]>
   3
   4
   5
   6
   $ try --optimize 'ancestors(6) and not ancestors(4)'
   (and
     (func
       ('symbol', 'ancestors')
       ('symbol', '6'))
     (not
       (func
         ('symbol', 'ancestors')
         ('symbol', '4'))))
   * optimized:
   (func
     ('symbol', 'only')
     (list
       ('symbol', '6')
       ('symbol', '4'))
     define)
   * set:
   <baseset+ [3, 5, 6]>
   3
   5
   6
 
 no crash by empty group "()" while optimizing to "only()"
 
   $ try --optimize '::1 and ()'
   (and
     (dagrangepre
       ('symbol', '1'))
     (group
       None))
   * optimized:
   (and
     None
     (func
       ('symbol', 'ancestors')
       ('symbol', '1')
       define)
     define)
   hg: parse error: missing argument
   [255]
 
 invalid function call should not be optimized to only()
 
   $ log '"ancestors"(6) and not ancestors(4)'
   hg: parse error: not a symbol
   [255]
 
   $ log 'ancestors(6) and not "ancestors"(4)'
   hg: parse error: not a symbol
   [255]
 
 we can use patterns when searching for tags
 
   $ log 'tag("1..*")'
   abort: tag '1..*' does not exist!
   [255]
   $ log 'tag("re:1..*")'
   6
   $ log 'tag("re:[0-9].[0-9]")'
   6
   $ log 'tag("literal:1.0")'
   6
   $ log 'tag("re:0..*")'
 
   $ log 'tag(unknown)'
   abort: tag 'unknown' does not exist!
   [255]
   $ log 'tag("re:unknown")'
   $ log 'present(tag("unknown"))'
   $ log 'present(tag("re:unknown"))'
   $ log 'branch(unknown)'
   abort: unknown revision 'unknown'!
   [255]
   $ log 'branch("literal:unknown")'
   abort: branch 'unknown' does not exist!
   [255]
   $ log 'branch("re:unknown")'
   $ log 'present(branch("unknown"))'
   $ log 'present(branch("re:unknown"))'
   $ log 'user(bob)'
   2
 
   $ log '4::8'
   4
   8
   $ log '4:8'
   4
   5
   6
   7
   8
 
   $ log 'sort(!merge() & (modifies(b) | user(bob) | keyword(bug) | keyword(issue) & 1::9), "-date")'
   4
   2
   5
 
   $ log 'not 0 and 0:2'
   1
   2
   $ log 'not 1 and 0:2'
   0
   2
   $ log 'not 2 and 0:2'
   0
   1
   $ log '(1 and 2)::'
   $ log '(1 and 2):'
   $ log '(1 and 2):3'
   $ log 'sort(head(), -rev)'
   9
   7
   6
   5
   4
   3
   2
   1
   0
   $ log '4::8 - 8'
   4
 
 matching() should preserve the order of the input set:
 
   $ log '(2 or 3 or 1) and matching(1 or 2 or 3)'
   2
   3
   1
 
   $ log 'named("unknown")'
   abort: namespace 'unknown' does not exist!
   [255]
   $ log 'named("re:unknown")'
   abort: no namespace exists that match 'unknown'!
   [255]
   $ log 'present(named("unknown"))'
   $ log 'present(named("re:unknown"))'
 
   $ log 'tag()'
   6
   $ log 'named("tags")'
   6
 
 issue2437
 
   $ log '3 and p1(5)'
   3
   $ log '4 and p2(6)'
   4
   $ log '1 and parents(:2)'
   1
   $ log '2 and children(1:)'
   2
   $ log 'roots(all()) or roots(all())'
   0
   $ hg debugrevspec 'roots(all()) or roots(all())'
   0
   $ log 'heads(branch(é)) or heads(branch(é))'
   9
   $ log 'ancestors(8) and (heads(branch("-a-b-c-")) or heads(branch(é)))'
   4
 
 issue2654: report a parse error if the revset was not completely parsed
 
   $ log '1 OR 2'
   hg: parse error at 2: invalid token
   [255]
 
 or operator should preserve ordering:
   $ log 'reverse(2::4) or tip'
   4
   2
   9
 
 parentrevspec
 
   $ log 'merge()^0'
   6
   $ log 'merge()^'
   5
   $ log 'merge()^1'
   5
   $ log 'merge()^2'
   4
   $ log '(not merge())^2'
   $ log 'merge()^^'
   3
   $ log 'merge()^1^'
   3
   $ log 'merge()^^^'
   1
 
   $ log '(merge() | 0)~-1'
   7
   1
   $ log 'merge()~-1'
   7
   $ log 'tip~-1'
   $ log '(tip | merge())~-1'
   7
   $ log 'merge()~0'
   6
   $ log 'merge()~1'
   5
   $ log 'merge()~2'
   3
   $ log 'merge()~2^1'
   1
   $ log 'merge()~3'
   1
 
   $ log '(-3:tip)^'
   4
   6
   8
 
   $ log 'tip^foo'
   hg: parse error: ^ expects a number 0, 1, or 2
   [255]
 
   $ log 'branchpoint()~-1'
   abort: revision in set has more than one child!
   [255]
 
 Bogus function gets suggestions
   $ log 'add()'
   hg: parse error: unknown identifier: add
   (did you mean adds?)
   [255]
   $ log 'added()'
   hg: parse error: unknown identifier: added
   (did you mean adds?)
   [255]
   $ log 'remo()'
   hg: parse error: unknown identifier: remo
   (did you mean one of remote, removes?)
   [255]
   $ log 'babar()'
   hg: parse error: unknown identifier: babar
   [255]
 
 Bogus function with a similar internal name doesn't suggest the internal name
   $ log 'matches()'
   hg: parse error: unknown identifier: matches
   (did you mean matching?)
   [255]
 
 Undocumented functions aren't suggested as similar either
   $ log 'tagged2()'
   hg: parse error: unknown identifier: tagged2
   [255]
 
 multiple revspecs
 
   $ hg log -r 'tip~1:tip' -r 'tip~2:tip~1' --template '{rev}\n'
   8
   9
   4
   5
   6
   7
 
 test usage in revpair (with "+")
 
 (real pair)
 
   $ hg diff -r 'tip^^' -r 'tip'
   diff -r 2326846efdab -r 24286f4ae135 .hgtags
   --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
   +++ b/.hgtags	Thu Jan 01 00:00:00 1970 +0000
   @@ -0,0 +1,1 @@
   +e0cc66ef77e8b6f711815af4e001a6594fde3ba5 1.0
   $ hg diff -r 'tip^^::tip'
   diff -r 2326846efdab -r 24286f4ae135 .hgtags
   --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
   +++ b/.hgtags	Thu Jan 01 00:00:00 1970 +0000
   @@ -0,0 +1,1 @@
   +e0cc66ef77e8b6f711815af4e001a6594fde3ba5 1.0
 
 (single rev)
 
   $ hg diff -r 'tip^' -r 'tip^'
   $ hg diff -r 'tip^:tip^'
 
 (single rev that does not looks like a range)
 
   $ hg diff -r 'tip^::tip^ or tip^'
   diff -r d5d0dcbdc4d9 .hgtags
   --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
   +++ b/.hgtags	* (glob)
   @@ -0,0 +1,1 @@
   +e0cc66ef77e8b6f711815af4e001a6594fde3ba5 1.0
   $ hg diff -r 'tip^ or tip^'
   diff -r d5d0dcbdc4d9 .hgtags
   --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
   +++ b/.hgtags	* (glob)
   @@ -0,0 +1,1 @@
   +e0cc66ef77e8b6f711815af4e001a6594fde3ba5 1.0
 
 (no rev)
 
   $ hg diff -r 'author("babar") or author("celeste")'
   abort: empty revision range
   [255]
 
 aliases:
 
   $ echo '[revsetalias]' >> .hg/hgrc
   $ echo 'm = merge()' >> .hg/hgrc
 (revset aliases can override builtin revsets)
   $ echo 'p2($1) = p1($1)' >> .hg/hgrc
   $ echo 'sincem = descendants(m)' >> .hg/hgrc
   $ echo 'd($1) = reverse(sort($1, date))' >> .hg/hgrc
   $ echo 'rs(ARG1, ARG2) = reverse(sort(ARG1, ARG2))' >> .hg/hgrc
   $ echo 'rs4(ARG1, ARGA, ARGB, ARG2) = reverse(sort(ARG1, ARG2))' >> .hg/hgrc
 
   $ try m
   ('symbol', 'm')
   * expanded:
   (func
     ('symbol', 'merge')
     None)
   * set:
   <filteredset
     <fullreposet+ 0:10>,
     <merge>>
   6
 
   $ HGPLAIN=1
   $ export HGPLAIN
   $ try m
   ('symbol', 'm')
   abort: unknown revision 'm'!
   [255]
 
   $ HGPLAINEXCEPT=revsetalias
   $ export HGPLAINEXCEPT
   $ try m
   ('symbol', 'm')
   * expanded:
   (func
     ('symbol', 'merge')
     None)
   * set:
   <filteredset
     <fullreposet+ 0:10>,
     <merge>>
   6
 
   $ unset HGPLAIN
   $ unset HGPLAINEXCEPT
 
   $ try 'p2(.)'
   (func
     ('symbol', 'p2')
     ('symbol', '.'))
   * expanded:
   (func
     ('symbol', 'p1')
     ('symbol', '.'))
   * set:
   <baseset+ [8]>
   8
 
   $ HGPLAIN=1
   $ export HGPLAIN
   $ try 'p2(.)'
   (func
     ('symbol', 'p2')
     ('symbol', '.'))
   * set:
   <baseset+ []>
 
   $ HGPLAINEXCEPT=revsetalias
   $ export HGPLAINEXCEPT
   $ try 'p2(.)'
   (func
     ('symbol', 'p2')
     ('symbol', '.'))
   * expanded:
   (func
     ('symbol', 'p1')
     ('symbol', '.'))
   * set:
   <baseset+ [8]>
   8
 
   $ unset HGPLAIN
   $ unset HGPLAINEXCEPT
 
 test alias recursion
 
   $ try sincem
   ('symbol', 'sincem')
   * expanded:
   (func
     ('symbol', 'descendants')
     (func
       ('symbol', 'merge')
       None))
   * set:
   <addset+
     <filteredset
       <fullreposet+ 0:10>,
       <merge>>,
     <generatorset+>>
   6
   7
 
 test infinite recursion
 
   $ echo 'recurse1 = recurse2' >> .hg/hgrc
   $ echo 'recurse2 = recurse1' >> .hg/hgrc
   $ try recurse1
   ('symbol', 'recurse1')
   hg: parse error: infinite expansion of revset alias "recurse1" detected
   [255]
 
   $ echo 'level1($1, $2) = $1 or $2' >> .hg/hgrc
   $ echo 'level2($1, $2) = level1($2, $1)' >> .hg/hgrc
   $ try "level2(level1(1, 2), 3)"
   (func
     ('symbol', 'level2')
     (list
       (func
         ('symbol', 'level1')
         (list
           ('symbol', '1')
           ('symbol', '2')))
       ('symbol', '3')))
   * expanded:
   (or
     (list
       ('symbol', '3')
       (or
         (list
           ('symbol', '1')
           ('symbol', '2')))))
   * set:
   <addset
     <baseset [3]>,
     <baseset [1, 2]>>
   3
   1
   2
 
 test nesting and variable passing
 
   $ echo 'nested($1) = nested2($1)' >> .hg/hgrc
   $ echo 'nested2($1) = nested3($1)' >> .hg/hgrc
   $ echo 'nested3($1) = max($1)' >> .hg/hgrc
   $ try 'nested(2:5)'
   (func
     ('symbol', 'nested')
     (range
       ('symbol', '2')
       ('symbol', '5')))
   * expanded:
   (func
     ('symbol', 'max')
     (range
       ('symbol', '2')
       ('symbol', '5')))
   * set:
   <baseset
     <max
       <fullreposet+ 0:10>,
       <spanset+ 2:6>>>
   5
 
 test chained `or` operations are flattened at parsing phase
 
   $ echo 'chainedorops($1, $2, $3) = $1|$2|$3' >> .hg/hgrc
   $ try 'chainedorops(0:1, 1:2, 2:3)'
   (func
     ('symbol', 'chainedorops')
     (list
       (range
         ('symbol', '0')
         ('symbol', '1'))
       (range
         ('symbol', '1')
         ('symbol', '2'))
       (range
         ('symbol', '2')
         ('symbol', '3'))))
   * expanded:
   (or
     (list
       (range
         ('symbol', '0')
         ('symbol', '1'))
       (range
         ('symbol', '1')
         ('symbol', '2'))
       (range
         ('symbol', '2')
         ('symbol', '3'))))
   * set:
   <addset
     <spanset+ 0:2>,
     <addset
       <spanset+ 1:3>,
       <spanset+ 2:4>>>
   0
   1
   2
   3
 
 test variable isolation, variable placeholders are rewritten as string
 then parsed and matched again as string. Check they do not leak too
 far away.
 
   $ echo 'injectparamasstring = max("$1")' >> .hg/hgrc
   $ echo 'callinjection($1) = descendants(injectparamasstring)' >> .hg/hgrc
   $ try 'callinjection(2:5)'
   (func
     ('symbol', 'callinjection')
     (range
       ('symbol', '2')
       ('symbol', '5')))
   * expanded:
   (func
     ('symbol', 'descendants')
     (func
       ('symbol', 'max')
       ('string', '$1')))
   abort: unknown revision '$1'!
   [255]
 
 test scope of alias expansion: 'universe' is expanded prior to 'shadowall(0)',
 but 'all()' should never be substituted to '0()'.
 
   $ echo 'universe = all()' >> .hg/hgrc
   $ echo 'shadowall(all) = all and universe' >> .hg/hgrc
   $ try 'shadowall(0)'
   (func
     ('symbol', 'shadowall')
     ('symbol', '0'))
   * expanded:
   (and
     ('symbol', '0')
     (func
       ('symbol', 'all')
       None))
   * set:
   <filteredset
     <baseset [0]>,
     <spanset+ 0:10>>
   0
 
 test unknown reference:
 
   $ try "unknownref(0)" --config 'revsetalias.unknownref($1)=$1:$2'
   (func
     ('symbol', 'unknownref')
     ('symbol', '0'))
   abort: bad definition of revset alias "unknownref": invalid symbol '$2'
   [255]
 
   $ hg debugrevspec --debug --config revsetalias.anotherbadone='branch(' "tip"
   ('symbol', 'tip')
   warning: bad definition of revset alias "anotherbadone": at 7: not a prefix: end
   * set:
   <baseset [9]>
   9
 
   $ try 'tip'
   ('symbol', 'tip')
   * set:
   <baseset [9]>
   9
 
   $ hg debugrevspec --debug --config revsetalias.'bad name'='tip' "tip"
   ('symbol', 'tip')
   warning: bad declaration of revset alias "bad name": at 4: invalid token
   * set:
   <baseset [9]>
   9
   $ echo 'strictreplacing($1, $10) = $10 or desc("$1")' >> .hg/hgrc
   $ try 'strictreplacing("foo", tip)'
   (func
     ('symbol', 'strictreplacing')
     (list
       ('string', 'foo')
       ('symbol', 'tip')))
   * expanded:
   (or
     (list
       ('symbol', 'tip')
       (func
         ('symbol', 'desc')
         ('string', '$1'))))
   * set:
   <addset
     <baseset [9]>,
     <filteredset
       <fullreposet+ 0:10>,
       <desc '$1'>>>
   9
 
   $ try 'd(2:5)'
   (func
     ('symbol', 'd')
     (range
       ('symbol', '2')
       ('symbol', '5')))
   * expanded:
   (func
     ('symbol', 'reverse')
     (func
       ('symbol', 'sort')
       (list
         (range
           ('symbol', '2')
           ('symbol', '5'))
         ('symbol', 'date'))))
   * set:
   <baseset [4, 5, 3, 2]>
   4
   5
   3
   2
   $ try 'rs(2 or 3, date)'
   (func
     ('symbol', 'rs')
     (list
       (or
         (list
           ('symbol', '2')
           ('symbol', '3')))
       ('symbol', 'date')))
   * expanded:
   (func
     ('symbol', 'reverse')
     (func
       ('symbol', 'sort')
       (list
         (or
           (list
             ('symbol', '2')
             ('symbol', '3')))
         ('symbol', 'date'))))
   * set:
   <baseset [3, 2]>
   3
   2
   $ try 'rs()'
   (func
     ('symbol', 'rs')
     None)
   hg: parse error: invalid number of arguments: 0
   [255]
   $ try 'rs(2)'
   (func
     ('symbol', 'rs')
     ('symbol', '2'))
   hg: parse error: invalid number of arguments: 1
   [255]
   $ try 'rs(2, data, 7)'
   (func
     ('symbol', 'rs')
     (list
       ('symbol', '2')
       ('symbol', 'data')
       ('symbol', '7')))
   hg: parse error: invalid number of arguments: 3
   [255]
   $ try 'rs4(2 or 3, x, x, date)'
   (func
     ('symbol', 'rs4')
     (list
       (or
         (list
           ('symbol', '2')
           ('symbol', '3')))
       ('symbol', 'x')
       ('symbol', 'x')
       ('symbol', 'date')))
   * expanded:
   (func
     ('symbol', 'reverse')
     (func
       ('symbol', 'sort')
       (list
         (or
           (list
             ('symbol', '2')
             ('symbol', '3')))
         ('symbol', 'date'))))
   * set:
   <baseset [3, 2]>
   3
   2
 
 issue4553: check that revset aliases override existing hash prefix
 
   $ hg log -qr e
   6:e0cc66ef77e8
 
   $ hg log -qr e --config revsetalias.e="all()"
   0:2785f51eece5
   1:d75937da8da0
   2:5ed5505e9f1c
   3:8528aa5637f2
   4:2326846efdab
   5:904fa392b941
   6:e0cc66ef77e8
   7:013af1973af4
   8:d5d0dcbdc4d9
   9:24286f4ae135
 
   $ hg log -qr e: --config revsetalias.e="0"
   0:2785f51eece5
   1:d75937da8da0
   2:5ed5505e9f1c
   3:8528aa5637f2
   4:2326846efdab
   5:904fa392b941
   6:e0cc66ef77e8
   7:013af1973af4
   8:d5d0dcbdc4d9
   9:24286f4ae135
 
   $ hg log -qr :e --config revsetalias.e="9"
   0:2785f51eece5
   1:d75937da8da0
   2:5ed5505e9f1c
   3:8528aa5637f2
   4:2326846efdab
   5:904fa392b941
   6:e0cc66ef77e8
   7:013af1973af4
   8:d5d0dcbdc4d9
   9:24286f4ae135
 
   $ hg log -qr e:
   6:e0cc66ef77e8
   7:013af1973af4
   8:d5d0dcbdc4d9
   9:24286f4ae135
 
   $ hg log -qr :e
   0:2785f51eece5
   1:d75937da8da0
   2:5ed5505e9f1c
   3:8528aa5637f2
   4:2326846efdab
   5:904fa392b941
   6:e0cc66ef77e8
 
 issue2549 - correct optimizations
 
   $ try 'limit(1 or 2 or 3, 2) and not 2'
   (and
     (func
       ('symbol', 'limit')
       (list
         (or
           (list
             ('symbol', '1')
             ('symbol', '2')
             ('symbol', '3')))
         ('symbol', '2')))
     (not
       ('symbol', '2')))
   * set:
   <filteredset
-    <baseset
-      <limit n=2, offset=0,
-        <baseset [1, 2, 3]>>>,
+    <baseset slice=0:2
+      <baseset [1, 2, 3]>>,
     <not
       <baseset [2]>>>
   1
   $ try 'max(1 or 2) and not 2'
   (and
     (func
       ('symbol', 'max')
       (or
         (list
           ('symbol', '1')
           ('symbol', '2'))))
     (not
       ('symbol', '2')))
   * set:
   <filteredset
     <baseset
       <max
         <fullreposet+ 0:10>,
         <baseset [1, 2]>>>,
     <not
       <baseset [2]>>>
   $ try 'min(1 or 2) and not 1'
   (and
     (func
       ('symbol', 'min')
       (or
         (list
           ('symbol', '1')
           ('symbol', '2'))))
     (not
       ('symbol', '1')))
   * set:
   <filteredset
     <baseset
       <min
         <fullreposet+ 0:10>,
         <baseset [1, 2]>>>,
     <not
       <baseset [1]>>>
   $ try 'last(1 or 2, 1) and not 2'
   (and
     (func
       ('symbol', 'last')
       (list
         (or
           (list
             ('symbol', '1')
             ('symbol', '2')))
         ('symbol', '1')))
     (not
       ('symbol', '2')))
   * set:
   <filteredset
-    <baseset
-      <last n=1,
-        <baseset [2, 1]>>>,
+    <baseset slice=0:1
+      <baseset [2, 1]>>,
     <not
       <baseset [2]>>>
 
 issue4289 - ordering of built-ins
   $ hg log -M -q -r 3:2
   3:8528aa5637f2
   2:5ed5505e9f1c
 
 test revsets started with 40-chars hash (issue3669)
 
   $ ISSUE3669_TIP=`hg tip --template '{node}'`
   $ hg log -r "${ISSUE3669_TIP}" --template '{rev}\n'
   9
   $ hg log -r "${ISSUE3669_TIP}^" --template '{rev}\n'
   8
 
 test or-ed indirect predicates (issue3775)
 
   $ log '6 or 6^1' | sort
   5
   6
   $ log '6^1 or 6' | sort
   5
   6
   $ log '4 or 4~1' | sort
   2
   4
   $ log '4~1 or 4' | sort
   2
   4
   $ log '(0 or 2):(4 or 6) or 0 or 6' | sort
   0
   1
   2
   3
   4
   5
   6
   $ log '0 or 6 or (0 or 2):(4 or 6)' | sort
   0
   1
   2
   3
   4
   5
   6
 
 tests for 'remote()' predicate:
 #.  (csets in remote) (id)            (remote)
 1.  less than local   current branch  "default"
 2.  same with local   specified       "default"
 3.  more than local   specified       specified
 
   $ hg clone --quiet -U . ../remote3
   $ cd ../remote3
   $ hg update -q 7
   $ echo r > r
   $ hg ci -Aqm 10
   $ log 'remote()'
   7
   $ log 'remote("a-b-c-")'
   2
   $ cd ../repo
   $ log 'remote(".a.b.c.", "../remote3")'
 
 tests for concatenation of strings/symbols by "##"
 
   $ try "278 ## '5f5' ## 1ee ## 'ce5'"
   (_concat
     (_concat
       (_concat
         ('symbol', '278')
         ('string', '5f5'))
       ('symbol', '1ee'))
     ('string', 'ce5'))
   * concatenated:
   ('string', '2785f51eece5')
   * set:
   <baseset [0]>
   0
 
   $ echo 'cat4($1, $2, $3, $4) = $1 ## $2 ## $3 ## $4' >> .hg/hgrc
   $ try "cat4(278, '5f5', 1ee, 'ce5')"
   (func
     ('symbol', 'cat4')
     (list
       ('symbol', '278')
       ('string', '5f5')
       ('symbol', '1ee')
       ('string', 'ce5')))
   * expanded:
   (_concat
     (_concat
       (_concat
         ('symbol', '278')
         ('string', '5f5'))
       ('symbol', '1ee'))
     ('string', 'ce5'))
   * concatenated:
   ('string', '2785f51eece5')
   * set:
   <baseset [0]>
   0
 
 (check concatenation in alias nesting)
 
   $ echo 'cat2($1, $2) = $1 ## $2' >> .hg/hgrc
   $ echo 'cat2x2($1, $2, $3, $4) = cat2($1 ## $2, $3 ## $4)' >> .hg/hgrc
   $ log "cat2x2(278, '5f5', 1ee, 'ce5')"
   0
 
 (check operator priority)
 
   $ echo 'cat2n2($1, $2, $3, $4) = $1 ## $2 or $3 ## $4~2' >> .hg/hgrc
   $ log "cat2n2(2785f5, 1eece5, 24286f, 4ae135)"
   0
   4
 
   $ cd ..
 
 prepare repository that has "default" branches of multiple roots
 
   $ hg init namedbranch
   $ cd namedbranch
 
   $ echo default0 >> a
   $ hg ci -Aqm0
   $ echo default1 >> a
   $ hg ci -m1
 
   $ hg branch -q stable
   $ echo stable2 >> a
   $ hg ci -m2
   $ echo stable3 >> a
   $ hg ci -m3
 
   $ hg update -q null
   $ echo default4 >> a
   $ hg ci -Aqm4
   $ echo default5 >> a
   $ hg ci -m5
 
 "null" revision belongs to "default" branch (issue4683)
 
   $ log 'branch(null)'
   0
   1
   4
   5
 
 "null" revision belongs to "default" branch, but it shouldn't appear in set
 unless explicitly specified (issue4682)
 
   $ log 'children(branch(default))'
   1
   2
   5
 
   $ cd ..
 
 test author/desc/keyword in problematic encoding
 # unicode: cp932:
 # u30A2    0x83 0x41(= 'A')
 # u30C2    0x83 0x61(= 'a')
 
   $ hg init problematicencoding
   $ cd problematicencoding
 
   $ python > setup.sh <<EOF
   > print u'''
   > echo a > text
   > hg add text
   > hg --encoding utf-8 commit -u '\u30A2' -m none
   > echo b > text
   > hg --encoding utf-8 commit -u '\u30C2' -m none
   > echo c > text
   > hg --encoding utf-8 commit -u none -m '\u30A2'
   > echo d > text
   > hg --encoding utf-8 commit -u none -m '\u30C2'
   > '''.encode('utf-8')
   > EOF
   $ sh < setup.sh
 
 test in problematic encoding
   $ python > test.sh <<EOF
   > print u'''
   > hg --encoding cp932 log --template '{rev}\\n' -r 'author(\u30A2)'
   > echo ====
   > hg --encoding cp932 log --template '{rev}\\n' -r 'author(\u30C2)'
   > echo ====
   > hg --encoding cp932 log --template '{rev}\\n' -r 'desc(\u30A2)'
   > echo ====
   > hg --encoding cp932 log --template '{rev}\\n' -r 'desc(\u30C2)'
   > echo ====
   > hg --encoding cp932 log --template '{rev}\\n' -r 'keyword(\u30A2)'
   > echo ====
   > hg --encoding cp932 log --template '{rev}\\n' -r 'keyword(\u30C2)'
   > '''.encode('cp932')
   > EOF
   $ sh < test.sh
   0
   ====
   1
   ====
   2
   ====
   3
   ====
   0
   2
   ====
   1
   3
 
 test error message of bad revset
   $ hg log -r 'foo\\'
   hg: parse error at 3: syntax error in revset 'foo\\'
   [255]
 
   $ cd ..
 
 Test that revset predicate of extension isn't loaded at failure of
 loading it
 
   $ cd repo
 
   $ cat <<EOF > $TESTTMP/custompredicate.py
   > from mercurial import error, registrar, revset
   > 
   > revsetpredicate = registrar.revsetpredicate()
   > 
   > @revsetpredicate('custom1()')
   > def custom1(repo, subset, x):
   >     return revset.baseset([1])
   > 
   > raise error.Abort('intentional failure of loading extension')
   > EOF
   $ cat <<EOF > .hg/hgrc
   > [extensions]
   > custompredicate = $TESTTMP/custompredicate.py
   > EOF
 
   $ hg debugrevspec "custom1()"
   *** failed to import extension custompredicate from $TESTTMP/custompredicate.py: intentional failure of loading extension
   hg: parse error: unknown identifier: custom1
   [255]
 
   $ cd ..