i18n: use UTF-8 string to lower filename for case collision check

Some character sets, cp932 (known as Shift-JIS for Japanese) for
example, use 0x41('A') - 0x5A('Z') and 0x61('a') - 0x7A('z') as second
or later character.

In such character set, case collision checking recognizes different
files as CASEFOLDED same file, if filenames are treated as byte
sequence.

win32mbcs extension is not appropriate to handle this problem, because
this problem can occur on other than Windows platform only if
problematic character set is used.

Callers of util.checkcase() use known ASCII filenames as last
component of path, and string.lower() is not applied to directory part
of path. So, util.checkcase() is kept intact, even though it applies
string.lower() to filenames.

tests/test-casecollision-i18n.t

@@ -0,0 +1,43
+run only on case-insensitive filesystems, because collision check at
+"hg update" is done only on case-insensitive filesystems
+
+  $ "$TESTDIR/hghave" icasefs || exit 80
+
+setup repository, and target files
+
+  $ HGENCODING=cp932
+  $ export HGENCODING
+  $ hg init t
+  $ cd t
+  $ python << EOF
+  > names = ["\x83\x41", # cp932(0x83, 0x41='A'), UNICODE(0x30a2)
+  >          "\x83\x5A", # cp932(0x83, 0x5A='Z'), UNICODE(0x30bb)
+  >          "\x83\x61", # cp932(0x83, 0x61='a'), UNICODE(0x30c2)
+  >          "\x83\x7A", # cp932(0x83, 0x7A='z'), UNICODE(0x30db)
+  >         ]
+  > for num, name in zip(range(len(names)), names):
+  >     # file for getting target filename of "hg add"
+  >     f = file(str(num), 'w'); f.write(name); f.close()
+  >     # target file of "hg add"
+  >     f = file(name, 'w'); f.write(name); f.close()
+  > EOF
+
+test filename collison check at "hg add"
+
+  $ hg add --config ui.portablefilenames=abort `cat 0`
+  $ hg add --config ui.portablefilenames=abort `cat 1`
+  $ hg add --config ui.portablefilenames=abort `cat 2`
+  $ hg add --config ui.portablefilenames=abort `cat 3`
+  $ hg status -a
+  A \x83A (esc)
+  A \x83Z (esc)
+  A \x83a (esc)
+  A \x83z (esc)
+
+test filename collision check at "hg update"
+
+  $ hg commit -m 'revision 0'
+  $ hg update null
+  0 files updated, 0 files merged, 4 files removed, 0 files unresolved
+  $ hg update tip
+  4 files updated, 0 files merged, 0 files removed, 0 files unresolved

mercurial/merge.py

@@ -1,566 +1,566
 # merge.py - directory-level update/merge handling for Mercurial
 #
 # Copyright 2006, 2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from node import nullid, nullrev, hex, bin
 from i18n import _
-import scmutil, util, filemerge, copies, subrepo
+import scmutil, util, filemerge, copies, subrepo, encoding
 import errno, os, shutil
 
 class mergestate(object):
     '''track 3-way merge state of individual files'''
     def __init__(self, repo):
         self._repo = repo
         self._dirty = False
         self._read()
     def reset(self, node=None):
         self._state = {}
         if node:
             self._local = node
         shutil.rmtree(self._repo.join("merge"), True)
         self._dirty = False
     def _read(self):
         self._state = {}
         try:
             f = self._repo.opener("merge/state")
             for i, l in enumerate(f):
                 if i == 0:
                     self._local = bin(l[:-1])
                 else:
                     bits = l[:-1].split("\0")
                     self._state[bits[0]] = bits[1:]
             f.close()
         except IOError, err:
             if err.errno != errno.ENOENT:
                 raise
         self._dirty = False
     def commit(self):
         if self._dirty:
             f = self._repo.opener("merge/state", "w")
             f.write(hex(self._local) + "\n")
             for d, v in self._state.iteritems():
                 f.write("\0".join([d] + v) + "\n")
             f.close()
             self._dirty = False
     def add(self, fcl, fco, fca, fd, flags):
         hash = util.sha1(fcl.path()).hexdigest()
         self._repo.opener.write("merge/" + hash, fcl.data())
         self._state[fd] = ['u', hash, fcl.path(), fca.path(),
                            hex(fca.filenode()), fco.path(), flags]
         self._dirty = True
     def __contains__(self, dfile):
         return dfile in self._state
     def __getitem__(self, dfile):
         return self._state[dfile][0]
     def __iter__(self):
         l = self._state.keys()
         l.sort()
         for f in l:
             yield f
     def mark(self, dfile, state):
         self._state[dfile][0] = state
         self._dirty = True
     def resolve(self, dfile, wctx, octx):
         if self[dfile] == 'r':
             return 0
         state, hash, lfile, afile, anode, ofile, flags = self._state[dfile]
         f = self._repo.opener("merge/" + hash)
         self._repo.wwrite(dfile, f.read(), flags)
         f.close()
         fcd = wctx[dfile]
         fco = octx[ofile]
         fca = self._repo.filectx(afile, fileid=anode)
         r = filemerge.filemerge(self._repo, self._local, lfile, fcd, fco, fca)
         if r is None:
             # no real conflict
             del self._state[dfile]
         elif not r:
             self.mark(dfile, 'r')
         return r
 
 def _checkunknown(wctx, mctx):
     "check for collisions between unknown files and files in mctx"
     for f in wctx.unknown():
         if f in mctx and mctx[f].cmp(wctx[f]):
             raise util.Abort(_("untracked file in working directory differs"
                                " from file in requested revision: '%s'") % f)
 
 def _checkcollision(mctx):
     "check for case folding collisions in the destination context"
     folded = {}
     for fn in mctx:
-        fold = fn.lower()
+        fold = encoding.lower(fn)
         if fold in folded:
             raise util.Abort(_("case-folding collision between %s and %s")
                              % (fn, folded[fold]))
         folded[fold] = fn
 
 def _forgetremoved(wctx, mctx, branchmerge):
     """
     Forget removed files
 
     If we're jumping between revisions (as opposed to merging), and if
     neither the working directory nor the target rev has the file,
     then we need to remove it from the dirstate, to prevent the
     dirstate from listing the file when it is no longer in the
     manifest.
 
     If we're merging, and the other revision has removed a file
     that is not present in the working directory, we need to mark it
     as removed.
     """
 
     action = []
     state = branchmerge and 'r' or 'f'
     for f in wctx.deleted():
         if f not in mctx:
             action.append((f, state))
 
     if not branchmerge:
         for f in wctx.removed():
             if f not in mctx:
                 action.append((f, "f"))
 
     return action
 
 def manifestmerge(repo, p1, p2, pa, overwrite, partial):
     """
     Merge p1 and p2 with ancestor pa and generate merge action list
 
     overwrite = whether we clobber working files
     partial = function to filter file lists
     """
 
     def fmerge(f, f2, fa):
         """merge flags"""
         a, m, n = ma.flags(fa), m1.flags(f), m2.flags(f2)
         if m == n: # flags agree
             return m # unchanged
         if m and n and not a: # flags set, don't agree, differ from parent
             r = repo.ui.promptchoice(
                 _(" conflicting flags for %s\n"
                   "(n)one, e(x)ec or sym(l)ink?") % f,
                 (_("&None"), _("E&xec"), _("Sym&link")), 0)
             if r == 1:
                 return "x" # Exec
             if r == 2:
                 return "l" # Symlink
             return ""
         if m and m != a: # changed from a to m
             return m
         if n and n != a: # changed from a to n
             return n
         return '' # flag was cleared
 
     def act(msg, m, f, *args):
         repo.ui.debug(" %s: %s -> %s\n" % (f, msg, m))
         action.append((f, m) + args)
 
     action, copy = [], {}
 
     if overwrite:
         pa = p1
     elif pa == p2: # backwards
         pa = p1.p1()
     elif pa and repo.ui.configbool("merge", "followcopies", True):
         dirs = repo.ui.configbool("merge", "followdirs", True)
         copy, diverge = copies.copies(repo, p1, p2, pa, dirs)
         for of, fl in diverge.iteritems():
             act("divergent renames", "dr", of, fl)
 
     repo.ui.note(_("resolving manifests\n"))
     repo.ui.debug(" overwrite %s partial %s\n" % (overwrite, bool(partial)))
     repo.ui.debug(" ancestor %s local %s remote %s\n" % (pa, p1, p2))
 
     m1, m2, ma = p1.manifest(), p2.manifest(), pa.manifest()
     copied = set(copy.values())
 
     if '.hgsubstate' in m1:
         # check whether sub state is modified
         for s in p1.substate:
             if p1.sub(s).dirty():
                 m1['.hgsubstate'] += "+"
                 break
 
     # Compare manifests
     for f, n in m1.iteritems():
         if partial and not partial(f):
             continue
         if f in m2:
             rflags = fmerge(f, f, f)
             a = ma.get(f, nullid)
             if n == m2[f] or m2[f] == a: # same or local newer
                 # is file locally modified or flags need changing?
                 # dirstate flags may need to be made current
                 if m1.flags(f) != rflags or n[20:]:
                     act("update permissions", "e", f, rflags)
             elif n == a: # remote newer
                 act("remote is newer", "g", f, rflags)
             else: # both changed
                 act("versions differ", "m", f, f, f, rflags, False)
         elif f in copied: # files we'll deal with on m2 side
             pass
         elif f in copy:
             f2 = copy[f]
             if f2 not in m2: # directory rename
                 act("remote renamed directory to " + f2, "d",
                     f, None, f2, m1.flags(f))
             else: # case 2 A,B/B/B or case 4,21 A/B/B
                 act("local copied/moved to " + f2, "m",
                     f, f2, f, fmerge(f, f2, f2), False)
         elif f in ma: # clean, a different, no remote
             if n != ma[f]:
                 if repo.ui.promptchoice(
                     _(" local changed %s which remote deleted\n"
                       "use (c)hanged version or (d)elete?") % f,
                     (_("&Changed"), _("&Delete")), 0):
                     act("prompt delete", "r", f)
                 else:
                     act("prompt keep", "a", f)
             elif n[20:] == "a": # added, no remote
                 act("remote deleted", "f", f)
             elif n[20:] != "u":
                 act("other deleted", "r", f)
 
     for f, n in m2.iteritems():
         if partial and not partial(f):
             continue
         if f in m1 or f in copied: # files already visited
             continue
         if f in copy:
             f2 = copy[f]
             if f2 not in m1: # directory rename
                 act("local renamed directory to " + f2, "d",
                     None, f, f2, m2.flags(f))
             elif f2 in m2: # rename case 1, A/A,B/A
                 act("remote copied to " + f, "m",
                     f2, f, f, fmerge(f2, f, f2), False)
             else: # case 3,20 A/B/A
                 act("remote moved to " + f, "m",
                     f2, f, f, fmerge(f2, f, f2), True)
         elif f not in ma:
             act("remote created", "g", f, m2.flags(f))
         elif n != ma[f]:
             if repo.ui.promptchoice(
                 _("remote changed %s which local deleted\n"
                   "use (c)hanged version or leave (d)eleted?") % f,
                 (_("&Changed"), _("&Deleted")), 0) == 0:
                 act("prompt recreating", "g", f, m2.flags(f))
 
     return action
 
 def actionkey(a):
     return a[1] == 'r' and -1 or 0, a
 
 def applyupdates(repo, action, wctx, mctx, actx, overwrite):
     """apply the merge action list to the working directory
 
     wctx is the working copy context
     mctx is the context to be merged into the working copy
     actx is the context of the common ancestor
 
     Return a tuple of counts (updated, merged, removed, unresolved) that
     describes how many files were affected by the update.
     """
 
     updated, merged, removed, unresolved = 0, 0, 0, 0
     ms = mergestate(repo)
     ms.reset(wctx.p1().node())
     moves = []
     action.sort(key=actionkey)
 
     # prescan for merges
     u = repo.ui
     for a in action:
         f, m = a[:2]
         if m == 'm': # merge
             f2, fd, flags, move = a[2:]
             if f == '.hgsubstate': # merged internally
                 continue
             repo.ui.debug("preserving %s for resolve of %s\n" % (f, fd))
             fcl = wctx[f]
             fco = mctx[f2]
             if mctx == actx: # backwards, use working dir parent as ancestor
                 if fcl.parents():
                     fca = fcl.p1()
                 else:
                     fca = repo.filectx(f, fileid=nullrev)
             else:
                 fca = fcl.ancestor(fco, actx)
             if not fca:
                 fca = repo.filectx(f, fileid=nullrev)
             ms.add(fcl, fco, fca, fd, flags)
             if f != fd and move:
                 moves.append(f)
 
     audit = scmutil.pathauditor(repo.root)
 
     # remove renamed files after safely stored
     for f in moves:
         if os.path.lexists(repo.wjoin(f)):
             repo.ui.debug("removing %s\n" % f)
             audit(f)
             os.unlink(repo.wjoin(f))
 
     numupdates = len(action)
     for i, a in enumerate(action):
         f, m = a[:2]
         u.progress(_('updating'), i + 1, item=f, total=numupdates,
                    unit=_('files'))
         if f and f[0] == "/":
             continue
         if m == "r": # remove
             repo.ui.note(_("removing %s\n") % f)
             audit(f)
             if f == '.hgsubstate': # subrepo states need updating
                 subrepo.submerge(repo, wctx, mctx, wctx, overwrite)
             try:
                 util.unlinkpath(repo.wjoin(f))
             except OSError, inst:
                 if inst.errno != errno.ENOENT:
                     repo.ui.warn(_("update failed to remove %s: %s!\n") %
                                  (f, inst.strerror))
             removed += 1
         elif m == "m": # merge
             if f == '.hgsubstate': # subrepo states need updating
                 subrepo.submerge(repo, wctx, mctx, wctx.ancestor(mctx), overwrite)
                 continue
             f2, fd, flags, move = a[2:]
             repo.wopener.audit(fd)
             r = ms.resolve(fd, wctx, mctx)
             if r is not None and r > 0:
                 unresolved += 1
             else:
                 if r is None:
                     updated += 1
                 else:
                     merged += 1
             util.setflags(repo.wjoin(fd), 'l' in flags, 'x' in flags)
             if (move and repo.dirstate.normalize(fd) != f
                 and os.path.lexists(repo.wjoin(f))):
                 repo.ui.debug("removing %s\n" % f)
                 audit(f)
                 os.unlink(repo.wjoin(f))
         elif m == "g": # get
             flags = a[2]
             repo.ui.note(_("getting %s\n") % f)
             t = mctx.filectx(f).data()
             repo.wwrite(f, t, flags)
             t = None
             updated += 1
             if f == '.hgsubstate': # subrepo states need updating
                 subrepo.submerge(repo, wctx, mctx, wctx, overwrite)
         elif m == "d": # directory rename
             f2, fd, flags = a[2:]
             if f:
                 repo.ui.note(_("moving %s to %s\n") % (f, fd))
                 audit(f)
                 t = wctx.filectx(f).data()
                 repo.wwrite(fd, t, flags)
                 util.unlinkpath(repo.wjoin(f))
             if f2:
                 repo.ui.note(_("getting %s to %s\n") % (f2, fd))
                 t = mctx.filectx(f2).data()
                 repo.wwrite(fd, t, flags)
             updated += 1
         elif m == "dr": # divergent renames
             fl = a[2]
             repo.ui.warn(_("note: possible conflict - %s was renamed "
                            "multiple times to:\n") % f)
             for nf in fl:
                 repo.ui.warn(" %s\n" % nf)
         elif m == "e": # exec
             flags = a[2]
             repo.wopener.audit(f)
             util.setflags(repo.wjoin(f), 'l' in flags, 'x' in flags)
     ms.commit()
     u.progress(_('updating'), None, total=numupdates, unit=_('files'))
 
     return updated, merged, removed, unresolved
 
 def recordupdates(repo, action, branchmerge):
     "record merge actions to the dirstate"
 
     for a in action:
         f, m = a[:2]
         if m == "r": # remove
             if branchmerge:
                 repo.dirstate.remove(f)
             else:
                 repo.dirstate.drop(f)
         elif m == "a": # re-add
             if not branchmerge:
                 repo.dirstate.add(f)
         elif m == "f": # forget
             repo.dirstate.drop(f)
         elif m == "e": # exec change
             repo.dirstate.normallookup(f)
         elif m == "g": # get
             if branchmerge:
                 repo.dirstate.otherparent(f)
             else:
                 repo.dirstate.normal(f)
         elif m == "m": # merge
             f2, fd, flag, move = a[2:]
             if branchmerge:
                 # We've done a branch merge, mark this file as merged
                 # so that we properly record the merger later
                 repo.dirstate.merge(fd)
                 if f != f2: # copy/rename
                     if move:
                         repo.dirstate.remove(f)
                     if f != fd:
                         repo.dirstate.copy(f, fd)
                     else:
                         repo.dirstate.copy(f2, fd)
             else:
                 # We've update-merged a locally modified file, so
                 # we set the dirstate to emulate a normal checkout
                 # of that file some time in the past. Thus our
                 # merge will appear as a normal local file
                 # modification.
                 if f2 == fd: # file not locally copied/moved
                     repo.dirstate.normallookup(fd)
                 if move:
                     repo.dirstate.drop(f)
         elif m == "d": # directory rename
             f2, fd, flag = a[2:]
             if not f2 and f not in repo.dirstate:
                 # untracked file moved
                 continue
             if branchmerge:
                 repo.dirstate.add(fd)
                 if f:
                     repo.dirstate.remove(f)
                     repo.dirstate.copy(f, fd)
                 if f2:
                     repo.dirstate.copy(f2, fd)
             else:
                 repo.dirstate.normal(fd)
                 if f:
                     repo.dirstate.drop(f)
 
 def update(repo, node, branchmerge, force, partial, ancestor=None):
     """
     Perform a merge between the working directory and the given node
 
     node = the node to update to, or None if unspecified
     branchmerge = whether to merge between branches
     force = whether to force branch merging or file overwriting
     partial = a function to filter file lists (dirstate not updated)
 
     The table below shows all the behaviors of the update command
     given the -c and -C or no options, whether the working directory
     is dirty, whether a revision is specified, and the relationship of
     the parent rev to the target rev (linear, on the same named
     branch, or on another named branch).
 
     This logic is tested by test-update-branches.t.
 
     -c  -C  dirty  rev  |  linear   same  cross
      n   n    n     n   |    ok     (1)     x
      n   n    n     y   |    ok     ok     ok
      n   n    y     *   |   merge   (2)    (2)
      n   y    *     *   |    ---  discard  ---
      y   n    y     *   |    ---    (3)    ---
      y   n    n     *   |    ---    ok     ---
      y   y    *     *   |    ---    (4)    ---
 
     x = can't happen
     * = don't-care
     1 = abort: crosses branches (use 'hg merge' or 'hg update -c')
     2 = abort: crosses branches (use 'hg merge' to merge or
                  use 'hg update -C' to discard changes)
     3 = abort: uncommitted local changes
     4 = incompatible options (checked in commands.py)
 
     Return the same tuple as applyupdates().
     """
 
     onode = node
     wlock = repo.wlock()
     try:
         wc = repo[None]
         if node is None:
             # tip of current branch
             try:
                 node = repo.branchtags()[wc.branch()]
             except KeyError:
                 if wc.branch() == "default": # no default branch!
                     node = repo.lookup("tip") # update to tip
                 else:
                     raise util.Abort(_("branch %s not found") % wc.branch())
         overwrite = force and not branchmerge
         pl = wc.parents()
         p1, p2 = pl[0], repo[node]
         if ancestor:
             pa = repo[ancestor]
         else:
             pa = p1.ancestor(p2)
 
         fp1, fp2, xp1, xp2 = p1.node(), p2.node(), str(p1), str(p2)
 
         ### check phase
         if not overwrite and len(pl) > 1:
             raise util.Abort(_("outstanding uncommitted merges"))
         if branchmerge:
             if pa == p2:
                 raise util.Abort(_("merging with a working directory ancestor"
                                    " has no effect"))
             elif pa == p1:
                 if p1.branch() == p2.branch():
                     raise util.Abort(_("nothing to merge (use 'hg update'"
                                        " or check 'hg heads')"))
             if not force and (wc.files() or wc.deleted()):
                 raise util.Abort(_("outstanding uncommitted changes "
                                    "(use 'hg status' to list changes)"))
             for s in wc.substate:
                 if wc.sub(s).dirty():
                     raise util.Abort(_("outstanding uncommitted changes in "
                                        "subrepository '%s'") % s)
 
         elif not overwrite:
             if pa == p1 or pa == p2: # linear
                 pass # all good
             elif wc.dirty(missing=True):
                 raise util.Abort(_("crosses branches (merge branches or use"
                                    " --clean to discard changes)"))
             elif onode is None:
                 raise util.Abort(_("crosses branches (merge branches or update"
                                    " --check to force update)"))
             else:
                 # Allow jumping branches if clean and specific rev given
                 overwrite = True
 
         ### calculate phase
         action = []
         wc.status(unknown=True) # prime cache
         if not force:
             _checkunknown(wc, p2)
         if not util.checkcase(repo.path):
             _checkcollision(p2)
         action += _forgetremoved(wc, p2, branchmerge)
         action += manifestmerge(repo, wc, p2, pa, overwrite, partial)
 
         ### apply phase
         if not branchmerge: # just jump to the new rev
             fp1, fp2, xp1, xp2 = fp2, nullid, xp2, ''
         if not partial:
             repo.hook('preupdate', throw=True, parent1=xp1, parent2=xp2)
 
         stats = applyupdates(repo, action, wc, p2, pa, overwrite)
 
         if not partial:
             repo.dirstate.setparents(fp1, fp2)
             recordupdates(repo, action, branchmerge)
             if not branchmerge:
                 repo.dirstate.setbranch(p2.branch())
     finally:
         wlock.release()
 
     if not partial:
         repo.hook('update', parent1=xp1, parent2=xp2, error=stats[3])
     return stats

mercurial/scmutil.py

@@ -1,711 +1,711
 # scmutil.py - Mercurial core utility functions
 #
 #  Copyright Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from i18n import _
-import util, error, osutil, revset, similar
+import util, error, osutil, revset, similar, encoding
 import match as matchmod
 import os, errno, re, stat, sys, glob
 
 def checkfilename(f):
     '''Check that the filename f is an acceptable filename for a tracked file'''
     if '\r' in f or '\n' in f:
         raise util.Abort(_("'\\n' and '\\r' disallowed in filenames: %r") % f)
 
 def checkportable(ui, f):
     '''Check if filename f is portable and warn or abort depending on config'''
     checkfilename(f)
     abort, warn = checkportabilityalert(ui)
     if abort or warn:
         msg = util.checkwinfilename(f)
         if msg:
             msg = "%s: %r" % (msg, f)
             if abort:
                 raise util.Abort(msg)
             ui.warn(_("warning: %s\n") % msg)
 
 def checkportabilityalert(ui):
     '''check if the user's config requests nothing, a warning, or abort for
     non-portable filenames'''
     val = ui.config('ui', 'portablefilenames', 'warn')
     lval = val.lower()
     bval = util.parsebool(val)
     abort = os.name == 'nt' or lval == 'abort'
     warn = bval or lval == 'warn'
     if bval is None and not (warn or abort or lval == 'ignore'):
         raise error.ConfigError(
             _("ui.portablefilenames value is invalid ('%s')") % val)
     return abort, warn
 
 class casecollisionauditor(object):
     def __init__(self, ui, abort, existingiter):
         self._ui = ui
         self._abort = abort
         self._map = {}
         for f in existingiter:
-            self._map[f.lower()] = f
+            self._map[encoding.lower(f)] = f
 
     def __call__(self, f):
-        fl = f.lower()
+        fl = encoding.lower(f)
         map = self._map
         if fl in map and map[fl] != f:
             msg = _('possible case-folding collision for %s') % f
             if self._abort:
                 raise util.Abort(msg)
             self._ui.warn(_("warning: %s\n") % msg)
         map[fl] = f
 
 class pathauditor(object):
     '''ensure that a filesystem path contains no banned components.
     the following properties of a path are checked:
 
     - ends with a directory separator
     - under top-level .hg
     - starts at the root of a windows drive
     - contains ".."
     - traverses a symlink (e.g. a/symlink_here/b)
     - inside a nested repository (a callback can be used to approve
       some nested repositories, e.g., subrepositories)
     '''
 
     def __init__(self, root, callback=None):
         self.audited = set()
         self.auditeddir = set()
         self.root = root
         self.callback = callback
 
     def __call__(self, path):
         '''Check the relative path.
         path may contain a pattern (e.g. foodir/**.txt)'''
 
         if path in self.audited:
             return
         # AIX ignores "/" at end of path, others raise EISDIR.
         if util.endswithsep(path):
             raise util.Abort(_("path ends in directory separator: %s") % path)
         normpath = os.path.normcase(path)
         parts = util.splitpath(normpath)
         if (os.path.splitdrive(path)[0]
             or parts[0].lower() in ('.hg', '.hg.', '')
             or os.pardir in parts):
             raise util.Abort(_("path contains illegal component: %s") % path)
         if '.hg' in path.lower():
             lparts = [p.lower() for p in parts]
             for p in '.hg', '.hg.':
                 if p in lparts[1:]:
                     pos = lparts.index(p)
                     base = os.path.join(*parts[:pos])
                     raise util.Abort(_('path %r is inside nested repo %r')
                                      % (path, base))
 
         parts.pop()
         prefixes = []
         while parts:
             prefix = os.sep.join(parts)
             if prefix in self.auditeddir:
                 break
             curpath = os.path.join(self.root, prefix)
             try:
                 st = os.lstat(curpath)
             except OSError, err:
                 # EINVAL can be raised as invalid path syntax under win32.
                 # They must be ignored for patterns can be checked too.
                 if err.errno not in (errno.ENOENT, errno.ENOTDIR, errno.EINVAL):
                     raise
             else:
                 if stat.S_ISLNK(st.st_mode):
                     raise util.Abort(
                         _('path %r traverses symbolic link %r')
                         % (path, prefix))
                 elif (stat.S_ISDIR(st.st_mode) and
                       os.path.isdir(os.path.join(curpath, '.hg'))):
                     if not self.callback or not self.callback(curpath):
                         raise util.Abort(_('path %r is inside nested repo %r') %
                                          (path, prefix))
             prefixes.append(prefix)
             parts.pop()
 
         self.audited.add(path)
         # only add prefixes to the cache after checking everything: we don't
         # want to add "foo/bar/baz" before checking if there's a "foo/.hg"
         self.auditeddir.update(prefixes)
 
 class abstractopener(object):
     """Abstract base class; cannot be instantiated"""
 
     def __init__(self, *args, **kwargs):
         '''Prevent instantiation; don't call this from subclasses.'''
         raise NotImplementedError('attempted instantiating ' + str(type(self)))
 
     def read(self, path):
         fp = self(path, 'rb')
         try:
             return fp.read()
         finally:
             fp.close()
 
     def write(self, path, data):
         fp = self(path, 'wb')
         try:
             return fp.write(data)
         finally:
             fp.close()
 
     def append(self, path, data):
         fp = self(path, 'ab')
         try:
             return fp.write(data)
         finally:
             fp.close()
 
 class opener(abstractopener):
     '''Open files relative to a base directory
 
     This class is used to hide the details of COW semantics and
     remote file access from higher level code.
     '''
     def __init__(self, base, audit=True):
         self.base = base
         self._audit = audit
         if audit:
             self.auditor = pathauditor(base)
         else:
             self.auditor = util.always
         self.createmode = None
         self._trustnlink = None
 
     @util.propertycache
     def _cansymlink(self):
         return util.checklink(self.base)
 
     def _fixfilemode(self, name):
         if self.createmode is None:
             return
         os.chmod(name, self.createmode & 0666)
 
     def __call__(self, path, mode="r", text=False, atomictemp=False):
         if self._audit:
             r = util.checkosfilename(path)
             if r:
                 raise util.Abort("%s: %r" % (r, path))
         self.auditor(path)
         f = os.path.join(self.base, path)
 
         if not text and "b" not in mode:
             mode += "b" # for that other OS
 
         nlink = -1
         dirname, basename = os.path.split(f)
         # If basename is empty, then the path is malformed because it points
         # to a directory. Let the posixfile() call below raise IOError.
         if basename and mode not in ('r', 'rb'):
             if atomictemp:
                 if not os.path.isdir(dirname):
                     util.makedirs(dirname, self.createmode)
                 return util.atomictempfile(f, mode, self.createmode)
             try:
                 if 'w' in mode:
                     util.unlink(f)
                     nlink = 0
                 else:
                     # nlinks() may behave differently for files on Windows
                     # shares if the file is open.
                     fd = util.posixfile(f)
                     nlink = util.nlinks(f)
                     if nlink < 1:
                         nlink = 2 # force mktempcopy (issue1922)
                     fd.close()
             except (OSError, IOError), e:
                 if e.errno != errno.ENOENT:
                     raise
                 nlink = 0
                 if not os.path.isdir(dirname):
                     util.makedirs(dirname, self.createmode)
             if nlink > 0:
                 if self._trustnlink is None:
                     self._trustnlink = nlink > 1 or util.checknlink(f)
                 if nlink > 1 or not self._trustnlink:
                     util.rename(util.mktempcopy(f), f)
         fp = util.posixfile(f, mode)
         if nlink == 0:
             self._fixfilemode(f)
         return fp
 
     def symlink(self, src, dst):
         self.auditor(dst)
         linkname = os.path.join(self.base, dst)
         try:
             os.unlink(linkname)
         except OSError:
             pass
 
         dirname = os.path.dirname(linkname)
         if not os.path.exists(dirname):
             util.makedirs(dirname, self.createmode)
 
         if self._cansymlink:
             try:
                 os.symlink(src, linkname)
             except OSError, err:
                 raise OSError(err.errno, _('could not symlink to %r: %s') %
                               (src, err.strerror), linkname)
         else:
             f = self(dst, "w")
             f.write(src)
             f.close()
             self._fixfilemode(dst)
 
     def audit(self, path):
         self.auditor(path)
 
 class filteropener(abstractopener):
     '''Wrapper opener for filtering filenames with a function.'''
 
     def __init__(self, opener, filter):
         self._filter = filter
         self._orig = opener
 
     def __call__(self, path, *args, **kwargs):
         return self._orig(self._filter(path), *args, **kwargs)
 
 def canonpath(root, cwd, myname, auditor=None):
     '''return the canonical path of myname, given cwd and root'''
     if util.endswithsep(root):
         rootsep = root
     else:
         rootsep = root + os.sep
     name = myname
     if not os.path.isabs(name):
         name = os.path.join(root, cwd, name)
     name = os.path.normpath(name)
     if auditor is None:
         auditor = pathauditor(root)
     if name != rootsep and name.startswith(rootsep):
         name = name[len(rootsep):]
         auditor(name)
         return util.pconvert(name)
     elif name == root:
         return ''
     else:
         # Determine whether `name' is in the hierarchy at or beneath `root',
         # by iterating name=dirname(name) until that causes no change (can't
         # check name == '/', because that doesn't work on windows).  For each
         # `name', compare dev/inode numbers.  If they match, the list `rel'
         # holds the reversed list of components making up the relative file
         # name we want.
         root_st = os.stat(root)
         rel = []
         while True:
             try:
                 name_st = os.stat(name)
             except OSError:
                 break
             if util.samestat(name_st, root_st):
                 if not rel:
                     # name was actually the same as root (maybe a symlink)
                     return ''
                 rel.reverse()
                 name = os.path.join(*rel)
                 auditor(name)
                 return util.pconvert(name)
             dirname, basename = os.path.split(name)
             rel.append(basename)
             if dirname == name:
                 break
             name = dirname
 
         raise util.Abort('%s not under root' % myname)
 
 def walkrepos(path, followsym=False, seen_dirs=None, recurse=False):
     '''yield every hg repository under path, recursively.'''
     def errhandler(err):
         if err.filename == path:
             raise err
     if followsym and hasattr(os.path, 'samestat'):
         def adddir(dirlst, dirname):
             match = False
             samestat = os.path.samestat
             dirstat = os.stat(dirname)
             for lstdirstat in dirlst:
                 if samestat(dirstat, lstdirstat):
                     match = True
                     break
             if not match:
                 dirlst.append(dirstat)
             return not match
     else:
         followsym = False
 
     if (seen_dirs is None) and followsym:
         seen_dirs = []
         adddir(seen_dirs, path)
     for root, dirs, files in os.walk(path, topdown=True, onerror=errhandler):
         dirs.sort()
         if '.hg' in dirs:
             yield root # found a repository
             qroot = os.path.join(root, '.hg', 'patches')
             if os.path.isdir(os.path.join(qroot, '.hg')):
                 yield qroot # we have a patch queue repo here
             if recurse:
                 # avoid recursing inside the .hg directory
                 dirs.remove('.hg')
             else:
                 dirs[:] = [] # don't descend further
         elif followsym:
             newdirs = []
             for d in dirs:
                 fname = os.path.join(root, d)
                 if adddir(seen_dirs, fname):
                     if os.path.islink(fname):
                         for hgname in walkrepos(fname, True, seen_dirs):
                             yield hgname
                     else:
                         newdirs.append(d)
             dirs[:] = newdirs
 
 def osrcpath():
     '''return default os-specific hgrc search path'''
     path = systemrcpath()
     path.extend(userrcpath())
     path = [os.path.normpath(f) for f in path]
     return path
 
 _rcpath = None
 
 def rcpath():
     '''return hgrc search path. if env var HGRCPATH is set, use it.
     for each item in path, if directory, use files ending in .rc,
     else use item.
     make HGRCPATH empty to only look in .hg/hgrc of current repo.
     if no HGRCPATH, use default os-specific path.'''
     global _rcpath
     if _rcpath is None:
         if 'HGRCPATH' in os.environ:
             _rcpath = []
             for p in os.environ['HGRCPATH'].split(os.pathsep):
                 if not p:
                     continue
                 p = util.expandpath(p)
                 if os.path.isdir(p):
                     for f, kind in osutil.listdir(p):
                         if f.endswith('.rc'):
                             _rcpath.append(os.path.join(p, f))
                 else:
                     _rcpath.append(p)
         else:
             _rcpath = osrcpath()
     return _rcpath
 
 if os.name != 'nt':
 
     def rcfiles(path):
         rcs = [os.path.join(path, 'hgrc')]
         rcdir = os.path.join(path, 'hgrc.d')
         try:
             rcs.extend([os.path.join(rcdir, f)
                         for f, kind in osutil.listdir(rcdir)
                         if f.endswith(".rc")])
         except OSError:
             pass
         return rcs
 
     def systemrcpath():
         path = []
         # old mod_python does not set sys.argv
         if len(getattr(sys, 'argv', [])) > 0:
             p = os.path.dirname(os.path.dirname(sys.argv[0]))
             path.extend(rcfiles(os.path.join(p, 'etc/mercurial')))
         path.extend(rcfiles('/etc/mercurial'))
         return path
 
     def userrcpath():
         return [os.path.expanduser('~/.hgrc')]
 
 else:
 
     _HKEY_LOCAL_MACHINE = 0x80000002L
 
     def systemrcpath():
         '''return default os-specific hgrc search path'''
         rcpath = []
         filename = util.executablepath()
         # Use mercurial.ini found in directory with hg.exe
         progrc = os.path.join(os.path.dirname(filename), 'mercurial.ini')
         if os.path.isfile(progrc):
             rcpath.append(progrc)
             return rcpath
         # Use hgrc.d found in directory with hg.exe
         progrcd = os.path.join(os.path.dirname(filename), 'hgrc.d')
         if os.path.isdir(progrcd):
             for f, kind in osutil.listdir(progrcd):
                 if f.endswith('.rc'):
                     rcpath.append(os.path.join(progrcd, f))
             return rcpath
         # else look for a system rcpath in the registry
         value = util.lookupreg('SOFTWARE\\Mercurial', None,
                                _HKEY_LOCAL_MACHINE)
         if not isinstance(value, str) or not value:
             return rcpath
         value = value.replace('/', os.sep)
         for p in value.split(os.pathsep):
             if p.lower().endswith('mercurial.ini'):
                 rcpath.append(p)
             elif os.path.isdir(p):
                 for f, kind in osutil.listdir(p):
                     if f.endswith('.rc'):
                         rcpath.append(os.path.join(p, f))
         return rcpath
 
     def userrcpath():
         '''return os-specific hgrc search path to the user dir'''
         home = os.path.expanduser('~')
         path = [os.path.join(home, 'mercurial.ini'),
                 os.path.join(home, '.hgrc')]
         userprofile = os.environ.get('USERPROFILE')
         if userprofile:
             path.append(os.path.join(userprofile, 'mercurial.ini'))
             path.append(os.path.join(userprofile, '.hgrc'))
         return path
 
 def revsingle(repo, revspec, default='.'):
     if not revspec:
         return repo[default]
 
     l = revrange(repo, [revspec])
     if len(l) < 1:
         raise util.Abort(_('empty revision set'))
     return repo[l[-1]]
 
 def revpair(repo, revs):
     if not revs:
         return repo.dirstate.p1(), None
 
     l = revrange(repo, revs)
 
     if len(l) == 0:
         return repo.dirstate.p1(), None
 
     if len(l) == 1:
         return repo.lookup(l[0]), None
 
     return repo.lookup(l[0]), repo.lookup(l[-1])
 
 _revrangesep = ':'
 
 def revrange(repo, revs):
     """Yield revision as strings from a list of revision specifications."""
 
     def revfix(repo, val, defval):
         if not val and val != 0 and defval is not None:
             return defval
         return repo.changelog.rev(repo.lookup(val))
 
     seen, l = set(), []
     for spec in revs:
         # attempt to parse old-style ranges first to deal with
         # things like old-tag which contain query metacharacters
         try:
             if isinstance(spec, int):
                 seen.add(spec)
                 l.append(spec)
                 continue
 
             if _revrangesep in spec:
                 start, end = spec.split(_revrangesep, 1)
                 start = revfix(repo, start, 0)
                 end = revfix(repo, end, len(repo) - 1)
                 step = start > end and -1 or 1
                 for rev in xrange(start, end + step, step):
                     if rev in seen:
                         continue
                     seen.add(rev)
                     l.append(rev)
                 continue
             elif spec and spec in repo: # single unquoted rev
                 rev = revfix(repo, spec, None)
                 if rev in seen:
                     continue
                 seen.add(rev)
                 l.append(rev)
                 continue
         except error.RepoLookupError:
             pass
 
         # fall through to new-style queries if old-style fails
         m = revset.match(repo.ui, spec)
         for r in m(repo, range(len(repo))):
             if r not in seen:
                 l.append(r)
         seen.update(l)
 
     return l
 
 def expandpats(pats):
     if not util.expandglobs:
         return list(pats)
     ret = []
     for p in pats:
         kind, name = matchmod._patsplit(p, None)
         if kind is None:
             try:
                 globbed = glob.glob(name)
             except re.error:
                 globbed = [name]
             if globbed:
                 ret.extend(globbed)
                 continue
         ret.append(p)
     return ret
 
 def match(ctx, pats=[], opts={}, globbed=False, default='relpath'):
     if pats == ("",):
         pats = []
     if not globbed and default == 'relpath':
         pats = expandpats(pats or [])
 
     m = ctx.match(pats, opts.get('include'), opts.get('exclude'),
                          default)
     def badfn(f, msg):
         ctx._repo.ui.warn("%s: %s\n" % (m.rel(f), msg))
     m.bad = badfn
     return m
 
 def matchall(repo):
     return matchmod.always(repo.root, repo.getcwd())
 
 def matchfiles(repo, files):
     return matchmod.exact(repo.root, repo.getcwd(), files)
 
 def addremove(repo, pats=[], opts={}, dry_run=None, similarity=None):
     if dry_run is None:
         dry_run = opts.get('dry_run')
     if similarity is None:
         similarity = float(opts.get('similarity') or 0)
     # we'd use status here, except handling of symlinks and ignore is tricky
     added, unknown, deleted, removed = [], [], [], []
     audit_path = pathauditor(repo.root)
     m = match(repo[None], pats, opts)
     for abs in repo.walk(m):
         target = repo.wjoin(abs)
         good = True
         try:
             audit_path(abs)
         except (OSError, util.Abort):
             good = False
         rel = m.rel(abs)
         exact = m.exact(abs)
         if good and abs not in repo.dirstate:
             unknown.append(abs)
             if repo.ui.verbose or not exact:
                 repo.ui.status(_('adding %s\n') % ((pats and rel) or abs))
         elif repo.dirstate[abs] != 'r' and (not good or not os.path.lexists(target)
             or (os.path.isdir(target) and not os.path.islink(target))):
             deleted.append(abs)
             if repo.ui.verbose or not exact:
                 repo.ui.status(_('removing %s\n') % ((pats and rel) or abs))
         # for finding renames
         elif repo.dirstate[abs] == 'r':
             removed.append(abs)
         elif repo.dirstate[abs] == 'a':
             added.append(abs)
     copies = {}
     if similarity > 0:
         for old, new, score in similar.findrenames(repo,
                 added + unknown, removed + deleted, similarity):
             if repo.ui.verbose or not m.exact(old) or not m.exact(new):
                 repo.ui.status(_('recording removal of %s as rename to %s '
                                  '(%d%% similar)\n') %
                                (m.rel(old), m.rel(new), score * 100))
             copies[new] = old
 
     if not dry_run:
         wctx = repo[None]
         wlock = repo.wlock()
         try:
             wctx.forget(deleted)
             wctx.add(unknown)
             for new, old in copies.iteritems():
                 wctx.copy(old, new)
         finally:
             wlock.release()
 
 def updatedir(ui, repo, patches, similarity=0):
     '''Update dirstate after patch application according to metadata'''
     if not patches:
         return []
     copies = []
     removes = set()
     cfiles = patches.keys()
     cwd = repo.getcwd()
     if cwd:
         cfiles = [util.pathto(repo.root, cwd, f) for f in patches.keys()]
     for f in patches:
         gp = patches[f]
         if not gp:
             continue
         if gp.op == 'RENAME':
             copies.append((gp.oldpath, gp.path))
             removes.add(gp.oldpath)
         elif gp.op == 'COPY':
             copies.append((gp.oldpath, gp.path))
         elif gp.op == 'DELETE':
             removes.add(gp.path)
 
     wctx = repo[None]
     for src, dst in copies:
         dirstatecopy(ui, repo, wctx, src, dst, cwd=cwd)
     if (not similarity) and removes:
         wctx.remove(sorted(removes), True)
 
     for f in patches:
         gp = patches[f]
         if gp and gp.mode:
             islink, isexec = gp.mode
             dst = repo.wjoin(gp.path)
             # patch won't create empty files
             if gp.op == 'ADD' and not os.path.lexists(dst):
                 flags = (isexec and 'x' or '') + (islink and 'l' or '')
                 repo.wwrite(gp.path, '', flags)
             util.setflags(dst, islink, isexec)
     addremove(repo, cfiles, similarity=similarity)
     files = patches.keys()
     files.extend([r for r in removes if r not in files])
     return sorted(files)
 
 def dirstatecopy(ui, repo, wctx, src, dst, dryrun=False, cwd=None):
     """Update the dirstate to reflect the intent of copying src to dst. For
     different reasons it might not end with dst being marked as copied from src.
     """
     origsrc = repo.dirstate.copied(src) or src
     if dst == origsrc: # copying back a copy?
         if repo.dirstate[dst] not in 'mn' and not dryrun:
             repo.dirstate.normallookup(dst)
     else:
         if repo.dirstate[origsrc] == 'a' and origsrc == src:
             if not ui.quiet:
                 ui.warn(_("%s has not been committed yet, so no copy "
                           "data will be stored for %s.\n")
                         % (repo.pathto(origsrc, cwd), repo.pathto(dst, cwd)))
             if repo.dirstate[dst] in '?r' and not dryrun:
                 wctx.add([dst])
         elif not dryrun:
             wctx.copy(origsrc, dst)
 
 def readrequires(opener, supported):
     '''Reads and parses .hg/requires and checks if all entries found
     are in the list of supported features.'''
     requirements = set(opener.read("requires").splitlines())
     missings = []
     for r in requirements:
         if r not in supported:
             if not r or not r[0].isalnum():
                 raise error.RequirementError(_(".hg/requires file is corrupt"))
             missings.append(r)
     missings.sort()
     if missings:
         raise error.RequirementError(_("unknown repository format: "
             "requires features '%s' (upgrade Mercurial)") % "', '".join(missings))
     return requirements