match: handle includes using new intersectionmatcher

mercurial/match.py

@@ -1,875 +1,949
 # match.py - filename matching
 #
 #  Copyright 2008, 2009 Matt Mackall <[e-mail]> and others
 #
 # 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 copy
 import os
 import re
 
 from .i18n import _
 from . import (
     error,
     pathutil,
     util,
 )
 
 propertycache = util.propertycache
 
 def _rematcher(regex):
     '''compile the regexp with the best available regexp engine and return a
     matcher function'''
     m = util.re.compile(regex)
     try:
         # slightly faster, provided by facebook's re2 bindings
         return m.test_match
     except AttributeError:
         return m.match
 
 def _expandsets(kindpats, ctx, listsubrepos):
     '''Returns the kindpats list with the 'set' patterns expanded.'''
     fset = set()
     other = []
 
     for kind, pat, source in kindpats:
         if kind == 'set':
             if not ctx:
                 raise error.ProgrammingError("fileset expression with no "
                                              "context")
             s = ctx.getfileset(pat)
             fset.update(s)
 
             if listsubrepos:
                 for subpath in ctx.substate:
                     s = ctx.sub(subpath).getfileset(pat)
                     fset.update(subpath + '/' + f for f in s)
 
             continue
         other.append((kind, pat, source))
     return fset, other
 
 def _expandsubinclude(kindpats, root):
     '''Returns the list of subinclude matcher args and the kindpats without the
     subincludes in it.'''
     relmatchers = []
     other = []
 
     for kind, pat, source in kindpats:
         if kind == 'subinclude':
             sourceroot = pathutil.dirname(util.normpath(source))
             pat = util.pconvert(pat)
             path = pathutil.join(sourceroot, pat)
 
             newroot = pathutil.dirname(path)
             matcherargs = (newroot, '', [], ['include:%s' % path])
 
             prefix = pathutil.canonpath(root, root, newroot)
             if prefix:
                 prefix += '/'
             relmatchers.append((prefix, matcherargs))
         else:
             other.append((kind, pat, source))
 
     return relmatchers, other
 
 def _kindpatsalwaysmatch(kindpats):
     """"Checks whether the kindspats match everything, as e.g.
     'relpath:.' does.
     """
     for kind, pat, source in kindpats:
         if pat != '' or kind not in ['relpath', 'glob']:
             return False
     return True
 
 def match(root, cwd, patterns, include=None, exclude=None, default='glob',
           exact=False, auditor=None, ctx=None, listsubrepos=False, warn=None,
           badfn=None, icasefs=False):
     """build an object to match a set of file patterns
 
     arguments:
     root - the canonical root of the tree you're matching against
     cwd - the current working directory, if relevant
     patterns - patterns to find
     include - patterns to include (unless they are excluded)
     exclude - patterns to exclude (even if they are included)
     default - if a pattern in patterns has no explicit type, assume this one
     exact - patterns are actually filenames (include/exclude still apply)
     warn - optional function used for printing warnings
     badfn - optional bad() callback for this matcher instead of the default
     icasefs - make a matcher for wdir on case insensitive filesystems, which
         normalizes the given patterns to the case in the filesystem
 
     a pattern is one of:
     'glob:<glob>' - a glob relative to cwd
     're:<regexp>' - a regular expression
     'path:<path>' - a path relative to repository root, which is matched
                     recursively
     'rootfilesin:<path>' - a path relative to repository root, which is
                     matched non-recursively (will not match subdirectories)
     'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs)
     'relpath:<path>' - a path relative to cwd
     'relre:<regexp>' - a regexp that needn't match the start of a name
     'set:<fileset>' - a fileset expression
     'include:<path>' - a file of patterns to read and include
     'subinclude:<path>' - a file of patterns to match against files under
                           the same directory
     '<something>' - a pattern of the specified default type
     """
     normalize = _donormalize
     if icasefs:
         if exact:
             raise error.ProgrammingError("a case-insensitive exact matcher "
                                          "doesn't make sense")
         dirstate = ctx.repo().dirstate
         dsnormalize = dirstate.normalize
 
         def normalize(patterns, default, root, cwd, auditor, warn):
             kp = _donormalize(patterns, default, root, cwd, auditor, warn)
             kindpats = []
             for kind, pats, source in kp:
                 if kind not in ('re', 'relre'):  # regex can't be normalized
                     p = pats
                     pats = dsnormalize(pats)
 
                     # Preserve the original to handle a case only rename.
                     if p != pats and p in dirstate:
                         kindpats.append((kind, p, source))
 
                 kindpats.append((kind, pats, source))
             return kindpats
 
-    m = matcher(root, cwd, normalize, patterns, include=include,
+    m = matcher(root, cwd, normalize, patterns, include=None,
                 default=default, exact=exact, auditor=auditor, ctx=ctx,
                 listsubrepos=listsubrepos, warn=warn, badfn=badfn)
+    if include:
+        im = matcher(root, cwd, normalize, [], include=include, default=default,
+                     exact=False, auditor=auditor, ctx=ctx,
+                     listsubrepos=listsubrepos, warn=warn, badfn=None)
+        m = intersectmatchers(m, im)
     if exclude:
         em = matcher(root, cwd, normalize, [], include=exclude, default=default,
                      exact=False, auditor=auditor, ctx=ctx,
                      listsubrepos=listsubrepos, warn=warn, badfn=None)
         m = differencematcher(m, em)
     return m
 
 def exact(root, cwd, files, badfn=None):
     return match(root, cwd, files, exact=True, badfn=badfn)
 
 def always(root, cwd):
     return match(root, cwd, [])
 
 def badmatch(match, badfn):
     """Make a copy of the given matcher, replacing its bad method with the given
     one.
     """
     m = copy.copy(match)
     m.bad = badfn
     return m
 
 def _donormalize(patterns, default, root, cwd, auditor, warn):
     '''Convert 'kind:pat' from the patterns list to tuples with kind and
     normalized and rooted patterns and with listfiles expanded.'''
     kindpats = []
     for kind, pat in [_patsplit(p, default) for p in patterns]:
         if kind in ('glob', 'relpath'):
             pat = pathutil.canonpath(root, cwd, pat, auditor)
         elif kind in ('relglob', 'path', 'rootfilesin'):
             pat = util.normpath(pat)
         elif kind in ('listfile', 'listfile0'):
             try:
                 files = util.readfile(pat)
                 if kind == 'listfile0':
                     files = files.split('\0')
                 else:
                     files = files.splitlines()
                 files = [f for f in files if f]
             except EnvironmentError:
                 raise error.Abort(_("unable to read file list (%s)") % pat)
             for k, p, source in _donormalize(files, default, root, cwd,
                                              auditor, warn):
                 kindpats.append((k, p, pat))
             continue
         elif kind == 'include':
             try:
                 fullpath = os.path.join(root, util.localpath(pat))
                 includepats = readpatternfile(fullpath, warn)
                 for k, p, source in _donormalize(includepats, default,
                                                  root, cwd, auditor, warn):
                     kindpats.append((k, p, source or pat))
             except error.Abort as inst:
                 raise error.Abort('%s: %s' % (pat, inst[0]))
             except IOError as inst:
                 if warn:
                     warn(_("skipping unreadable pattern file '%s': %s\n") %
                          (pat, inst.strerror))
             continue
         # else: re or relre - which cannot be normalized
         kindpats.append((kind, pat, ''))
     return kindpats
 
 class basematcher(object):
 
     def __init__(self, root, cwd, badfn=None, relativeuipath=True):
         self._root = root
         self._cwd = cwd
         if badfn is not None:
             self.bad = badfn
         self._relativeuipath = relativeuipath
 
     def __call__(self, fn):
         return self.matchfn(fn)
     def __iter__(self):
         for f in self._files:
             yield f
     # Callbacks related to how the matcher is used by dirstate.walk.
     # Subscribers to these events must monkeypatch the matcher object.
     def bad(self, f, msg):
         '''Callback from dirstate.walk for each explicit file that can't be
         found/accessed, with an error message.'''
         pass
 
     # If an explicitdir is set, it will be called when an explicitly listed
     # directory is visited.
     explicitdir = None
 
     # If an traversedir is set, it will be called when a directory discovered
     # by recursive traversal is visited.
     traversedir = None
 
     def abs(self, f):
         '''Convert a repo path back to path that is relative to the root of the
         matcher.'''
         return f
 
     def rel(self, f):
         '''Convert repo path back to path that is relative to cwd of matcher.'''
         return util.pathto(self._root, self._cwd, f)
 
     def uipath(self, f):
         '''Convert repo path to a display path.  If patterns or -I/-X were used
         to create this matcher, the display path will be relative to cwd.
         Otherwise it is relative to the root of the repo.'''
         return (self._relativeuipath and self.rel(f)) or self.abs(f)
 
     @propertycache
     def _files(self):
         return []
 
     def files(self):
         '''Explicitly listed files or patterns or roots:
         if no patterns or .always(): empty list,
         if exact: list exact files,
         if not .anypats(): list all files and dirs,
         else: optimal roots'''
         return self._files
 
     @propertycache
     def _fileset(self):
         return set(self._files)
 
     def exact(self, f):
         '''Returns True if f is in .files().'''
         return f in self._fileset
 
     def matchfn(self, f):
         return False
 
     def visitdir(self, dir):
         '''Decides whether a directory should be visited based on whether it
         has potential matches in it or one of its subdirectories. This is
         based on the match's primary, included, and excluded patterns.
 
         Returns the string 'all' if the given directory and all subdirectories
         should be visited. Otherwise returns True or False indicating whether
         the given directory should be visited.
 
         This function's behavior is undefined if it has returned False for
         one of the dir's parent directories.
         '''
         return False
 
     def anypats(self):
         '''Matcher uses patterns or include/exclude.'''
         return False
 
     def always(self):
         '''Matcher will match everything and .files() will be empty
         - optimization might be possible and necessary.'''
         return False
 
     def isexact(self):
         return False
 
     def prefix(self):
         return not self.always() and not self.isexact() and not self.anypats()
 
 class matcher(basematcher):
 
     def __init__(self, root, cwd, normalize, patterns, include=None,
                  default='glob', exact=False, auditor=None, ctx=None,
                  listsubrepos=False, warn=None, badfn=None):
         super(matcher, self).__init__(root, cwd, badfn,
                                       relativeuipath=bool(include or patterns))
         if include is None:
             include = []
 
         self._anypats = bool(include)
         self._anyincludepats = False
         self._always = False
         self.patternspat = None
         self.includepat = None
 
         # roots are directories which are recursively included.
         self._includeroots = set()
         # dirs are directories which are non-recursively included.
         self._includedirs = set()
 
         matchfns = []
         if include:
             kindpats = normalize(include, 'glob', root, cwd, auditor, warn)
             self.includepat, im = _buildmatch(ctx, kindpats, '(?:/|$)',
                                               listsubrepos, root)
             self._anyincludepats = _anypats(kindpats)
             roots, dirs = _rootsanddirs(kindpats)
             self._includeroots.update(roots)
             self._includedirs.update(dirs)
             matchfns.append(im)
         if exact:
             if isinstance(patterns, list):
                 self._files = patterns
             else:
                 self._files = list(patterns)
             matchfns.append(self.exact)
         elif patterns:
             kindpats = normalize(patterns, default, root, cwd, auditor, warn)
             if not _kindpatsalwaysmatch(kindpats):
                 self._files = _explicitfiles(kindpats)
                 self._anypats = self._anypats or _anypats(kindpats)
                 self.patternspat, pm = _buildmatch(ctx, kindpats, '$',
                                                    listsubrepos, root)
                 matchfns.append(pm)
 
         if not matchfns:
             m = util.always
             self._always = True
         elif len(matchfns) == 1:
             m = matchfns[0]
         else:
             def m(f):
                 for matchfn in matchfns:
                     if not matchfn(f):
                         return False
                 return True
 
         self.matchfn = m
 
     @propertycache
     def _dirs(self):
         return set(util.dirs(self._fileset)) | {'.'}
 
     def visitdir(self, dir):
         if self.prefix() and dir in self._fileset:
             return 'all'
         if self._includeroots or self._includedirs:
             if (not self._anyincludepats and
                 dir in self._includeroots):
                 # The condition above is essentially self.prefix() for includes
                 return 'all'
             if ('.' not in self._includeroots and
                 dir not in self._includeroots and
                 dir not in self._includedirs and
                 not any(parent in self._includeroots
                         for parent in util.finddirs(dir))):
                 return False
         return (not self._fileset or
                 '.' in self._fileset or
                 dir in self._fileset or
                 dir in self._dirs or
                 any(parentdir in self._fileset
                     for parentdir in util.finddirs(dir)))
 
     def anypats(self):
         return self._anypats
 
     def always(self):
         return self._always
 
     def isexact(self):
         return self.matchfn == self.exact
 
     def __repr__(self):
         return ('<matcher files=%r, patterns=%r, includes=%r>' %
                 (self._files, self.patternspat, self.includepat))
 
 class differencematcher(basematcher):
     '''Composes two matchers by matching if the first matches and the second
     does not. Well, almost... If the user provides a pattern like "-X foo foo",
     Mercurial actually does match "foo" against that. That's because exact
     matches are treated specially. So, since this differencematcher is used for
     excludes, it needs to special-case exact matching.
 
     The second matcher's non-matching-attributes (root, cwd, bad, explicitdir,
     traversedir) are ignored.
 
     TODO: If we want to keep the behavior described above for exact matches, we
     should consider instead treating the above case something like this:
     union(exact(foo), difference(pattern(foo), include(foo)))
     '''
     def __init__(self, m1, m2):
         super(differencematcher, self).__init__(m1._root, m1._cwd)
         self._m1 = m1
         self._m2 = m2
         self.bad = m1.bad
         self.explicitdir = m1.explicitdir
         self.traversedir = m1.traversedir
 
     def matchfn(self, f):
         return self._m1(f) and (not self._m2(f) or self._m1.exact(f))
 
     @propertycache
     def _files(self):
         if self.isexact():
             return [f for f in self._m1.files() if self(f)]
         # If m1 is not an exact matcher, we can't easily figure out the set of
         # files, because its files() are not always files. For example, if
         # m1 is "path:dir" and m2 is "rootfileins:.", we don't
         # want to remove "dir" from the set even though it would match m2,
         # because the "dir" in m1 may not be a file.
         return self._m1.files()
 
     def visitdir(self, dir):
         if self._m2.visitdir(dir) == 'all':
             # There's a bug here: If m1 matches file 'dir/file' and m2 excludes
             # 'dir' (recursively), we should still visit 'dir' due to the
             # exception we have for exact matches.
             return False
         return bool(self._m1.visitdir(dir))
 
     def isexact(self):
         return self._m1.isexact()
 
     def anypats(self):
         return self._m1.anypats() or self._m2.anypats()
 
     def prefix(self):
         return not self.always() and not self.isexact() and not self.anypats()
 
     def __repr__(self):
         return ('<differencematcher m1=%r, m2=%r>' % (self._m1, self._m2))
 
+def intersectmatchers(m1, m2):
+    '''Composes two matchers by matching if both of them match.
+
+    The second matcher's non-matching-attributes (root, cwd, bad, explicitdir,
+    traversedir) are ignored.
+    '''
+    if m1 is None or m2 is None:
+        return m1 or m2
+    if m1.always():
+        m = copy.copy(m2)
+        # TODO: Consider encapsulating these things in a class so there's only
+        # one thing to copy from m1.
+        m.bad = m1.bad
+        m.explicitdir = m1.explicitdir
+        m.traversedir = m1.traversedir
+        m.abs = m1.abs
+        m.rel = m1.rel
+        m._relativeuipath |= m1._relativeuipath
+        return m
+    if m2.always():
+        m = copy.copy(m1)
+        m._relativeuipath |= m2._relativeuipath
+        return m
+    return intersectionmatcher(m1, m2)
+
+class intersectionmatcher(basematcher):
+    def __init__(self, m1, m2):
+        super(intersectionmatcher, self).__init__(m1._root, m1._cwd)
+        self._m1 = m1
+        self._m2 = m2
+        self.bad = m1.bad
+        self.explicitdir = m1.explicitdir
+        self.traversedir = m1.traversedir
+
+    @propertycache
+    def _files(self):
+        if self.isexact():
+            m1, m2 = self._m1, self._m2
+            if not m1.isexact():
+                m1, m2 = m2, m1
+            return [f for f in m1.files() if m2(f)]
+        # It neither m1 nor m2 is an exact matcher, we can't easily intersect
+        # the set of files, because their files() are not always files. For
+        # example, if intersecting a matcher "-I glob:foo.txt" with matcher of
+        # "path:dir2", we don't want to remove "dir2" from the set.
+        return self._m1.files() + self._m2.files()
+
+    def matchfn(self, f):
+        return self._m1(f) and self._m2(f)
+
+    def visitdir(self, dir):
+        visit1 = self._m1.visitdir(dir)
+        if visit1 == 'all':
+            return self._m2.visitdir(dir)
+        # bool() because visit1=True + visit2='all' should not be 'all'
+        return bool(visit1 and self._m2.visitdir(dir))
+
+    def always(self):
+        return self._m1.always() and self._m2.always()
+
+    def isexact(self):
+        return self._m1.isexact() or self._m2.isexact()
+
+    def anypats(self):
+        return self._m1.anypats() or self._m2.anypats()
+
+    def __repr__(self):
+        return ('<intersectionmatcher m1=%r, m2=%r>' % (self._m1, self._m2))
+
 class subdirmatcher(basematcher):
     """Adapt a matcher to work on a subdirectory only.
 
     The paths are remapped to remove/insert the path as needed:
 
     >>> m1 = match('root', '', ['a.txt', 'sub/b.txt'])
     >>> m2 = subdirmatcher('sub', m1)
     >>> bool(m2('a.txt'))
     False
     >>> bool(m2('b.txt'))
     True
     >>> bool(m2.matchfn('a.txt'))
     False
     >>> bool(m2.matchfn('b.txt'))
     True
     >>> m2.files()
     ['b.txt']
     >>> m2.exact('b.txt')
     True
     >>> util.pconvert(m2.rel('b.txt'))
     'sub/b.txt'
     >>> def bad(f, msg):
     ...     print "%s: %s" % (f, msg)
     >>> m1.bad = bad
     >>> m2.bad('x.txt', 'No such file')
     sub/x.txt: No such file
     >>> m2.abs('c.txt')
     'sub/c.txt'
     """
 
     def __init__(self, path, matcher):
         super(subdirmatcher, self).__init__(matcher._root, matcher._cwd)
         self._path = path
         self._matcher = matcher
         self._always = matcher.always()
 
         self._files = [f[len(path) + 1:] for f in matcher._files
                        if f.startswith(path + "/")]
 
         # If the parent repo had a path to this subrepo and the matcher is
         # a prefix matcher, this submatcher always matches.
         if matcher.prefix():
             self._always = any(f == path for f in matcher._files)
 
     def bad(self, f, msg):
         self._matcher.bad(self._path + "/" + f, msg)
 
     def abs(self, f):
         return self._matcher.abs(self._path + "/" + f)
 
     def rel(self, f):
         return self._matcher.rel(self._path + "/" + f)
 
     def uipath(self, f):
         return self._matcher.uipath(self._path + "/" + f)
 
     def matchfn(self, f):
         # Some information is lost in the superclass's constructor, so we
         # can not accurately create the matching function for the subdirectory
         # from the inputs. Instead, we override matchfn() and visitdir() to
         # call the original matcher with the subdirectory path prepended.
         return self._matcher.matchfn(self._path + "/" + f)
 
     def visitdir(self, dir):
         if dir == '.':
             dir = self._path
         else:
             dir = self._path + "/" + dir
         return self._matcher.visitdir(dir)
 
     def always(self):
         return self._always
 
     def anypats(self):
         return self._matcher.anypats()
 
 def patkind(pattern, default=None):
     '''If pattern is 'kind:pat' with a known kind, return kind.'''
     return _patsplit(pattern, default)[0]
 
 def _patsplit(pattern, default):
     """Split a string into the optional pattern kind prefix and the actual
     pattern."""
     if ':' in pattern:
         kind, pat = pattern.split(':', 1)
         if kind in ('re', 'glob', 'path', 'relglob', 'relpath', 'relre',
                     'listfile', 'listfile0', 'set', 'include', 'subinclude',
                     'rootfilesin'):
             return kind, pat
     return default, pattern
 
 def _globre(pat):
     r'''Convert an extended glob string to a regexp string.
 
     >>> print _globre(r'?')
     .
     >>> print _globre(r'*')
     [^/]*
     >>> print _globre(r'**')
     .*
     >>> print _globre(r'**/a')
     (?:.*/)?a
     >>> print _globre(r'a/**/b')
     a\/(?:.*/)?b
     >>> print _globre(r'[a*?!^][^b][!c]')
     [a*?!^][\^b][^c]
     >>> print _globre(r'{a,b}')
     (?:a|b)
     >>> print _globre(r'.\*\?')
     \.\*\?
     '''
     i, n = 0, len(pat)
     res = ''
     group = 0
     escape = util.re.escape
     def peek():
         return i < n and pat[i:i + 1]
     while i < n:
         c = pat[i:i + 1]
         i += 1
         if c not in '*?[{},\\':
             res += escape(c)
         elif c == '*':
             if peek() == '*':
                 i += 1
                 if peek() == '/':
                     i += 1
                     res += '(?:.*/)?'
                 else:
                     res += '.*'
             else:
                 res += '[^/]*'
         elif c == '?':
             res += '.'
         elif c == '[':
             j = i
             if j < n and pat[j:j + 1] in '!]':
                 j += 1
             while j < n and pat[j:j + 1] != ']':
                 j += 1
             if j >= n:
                 res += '\\['
             else:
                 stuff = pat[i:j].replace('\\','\\\\')
                 i = j + 1
                 if stuff[0:1] == '!':
                     stuff = '^' + stuff[1:]
                 elif stuff[0:1] == '^':
                     stuff = '\\' + stuff
                 res = '%s[%s]' % (res, stuff)
         elif c == '{':
             group += 1
             res += '(?:'
         elif c == '}' and group:
             res += ')'
             group -= 1
         elif c == ',' and group:
             res += '|'
         elif c == '\\':
             p = peek()
             if p:
                 i += 1
                 res += escape(p)
             else:
                 res += escape(c)
         else:
             res += escape(c)
     return res
 
 def _regex(kind, pat, globsuffix):
     '''Convert a (normalized) pattern of any kind into a regular expression.
     globsuffix is appended to the regexp of globs.'''
     if not pat:
         return ''
     if kind == 're':
         return pat
     if kind == 'path':
         if pat == '.':
             return ''
         return '^' + util.re.escape(pat) + '(?:/|$)'
     if kind == 'rootfilesin':
         if pat == '.':
             escaped = ''
         else:
             # Pattern is a directory name.
             escaped = util.re.escape(pat) + '/'
         # Anything after the pattern must be a non-directory.
         return '^' + escaped + '[^/]+$'
     if kind == 'relglob':
         return '(?:|.*/)' + _globre(pat) + globsuffix
     if kind == 'relpath':
         return util.re.escape(pat) + '(?:/|$)'
     if kind == 'relre':
         if pat.startswith('^'):
             return pat
         return '.*' + pat
     return _globre(pat) + globsuffix
 
 def _buildmatch(ctx, kindpats, globsuffix, listsubrepos, root):
     '''Return regexp string and a matcher function for kindpats.
     globsuffix is appended to the regexp of globs.'''
     matchfuncs = []
 
     subincludes, kindpats = _expandsubinclude(kindpats, root)
     if subincludes:
         submatchers = {}
         def matchsubinclude(f):
             for prefix, matcherargs in subincludes:
                 if f.startswith(prefix):
                     mf = submatchers.get(prefix)
                     if mf is None:
                         mf = match(*matcherargs)
                         submatchers[prefix] = mf
 
                     if mf(f[len(prefix):]):
                         return True
             return False
         matchfuncs.append(matchsubinclude)
 
     fset, kindpats = _expandsets(kindpats, ctx, listsubrepos)
     if fset:
         matchfuncs.append(fset.__contains__)
 
     regex = ''
     if kindpats:
         regex, mf = _buildregexmatch(kindpats, globsuffix)
         matchfuncs.append(mf)
 
     if len(matchfuncs) == 1:
         return regex, matchfuncs[0]
     else:
         return regex, lambda f: any(mf(f) for mf in matchfuncs)
 
 def _buildregexmatch(kindpats, globsuffix):
     """Build a match function from a list of kinds and kindpats,
     return regexp string and a matcher function."""
     try:
         regex = '(?:%s)' % '|'.join([_regex(k, p, globsuffix)
                                      for (k, p, s) in kindpats])
         if len(regex) > 20000:
             raise OverflowError
         return regex, _rematcher(regex)
     except OverflowError:
         # We're using a Python with a tiny regex engine and we
         # made it explode, so we'll divide the pattern list in two
         # until it works
         l = len(kindpats)
         if l < 2:
             raise
         regexa, a = _buildregexmatch(kindpats[:l//2], globsuffix)
         regexb, b = _buildregexmatch(kindpats[l//2:], globsuffix)
         return regex, lambda s: a(s) or b(s)
     except re.error:
         for k, p, s in kindpats:
             try:
                 _rematcher('(?:%s)' % _regex(k, p, globsuffix))
             except re.error:
                 if s:
                     raise error.Abort(_("%s: invalid pattern (%s): %s") %
                                      (s, k, p))
                 else:
                     raise error.Abort(_("invalid pattern (%s): %s") % (k, p))
         raise error.Abort(_("invalid pattern"))
 
 def _patternrootsanddirs(kindpats):
     '''Returns roots and directories corresponding to each pattern.
 
     This calculates the roots and directories exactly matching the patterns and
     returns a tuple of (roots, dirs) for each. It does not return other
     directories which may also need to be considered, like the parent
     directories.
     '''
     r = []
     d = []
     for kind, pat, source in kindpats:
         if kind == 'glob': # find the non-glob prefix
             root = []
             for p in pat.split('/'):
                 if '[' in p or '{' in p or '*' in p or '?' in p:
                     break
                 root.append(p)
             r.append('/'.join(root) or '.')
         elif kind in ('relpath', 'path'):
             r.append(pat or '.')
         elif kind in ('rootfilesin',):
             d.append(pat or '.')
         else: # relglob, re, relre
             r.append('.')
     return r, d
 
 def _roots(kindpats):
     '''Returns root directories to match recursively from the given patterns.'''
     roots, dirs = _patternrootsanddirs(kindpats)
     return roots
 
 def _rootsanddirs(kindpats):
     '''Returns roots and exact directories from patterns.
 
     roots are directories to match recursively, whereas exact directories should
     be matched non-recursively. The returned (roots, dirs) tuple will also
     include directories that need to be implicitly considered as either, such as
     parent directories.
 
     >>> _rootsanddirs(\
         [('glob', 'g/h/*', ''), ('glob', 'g/h', ''), ('glob', 'g*', '')])
     (['g/h', 'g/h', '.'], ['g', '.'])
     >>> _rootsanddirs(\
         [('rootfilesin', 'g/h', ''), ('rootfilesin', '', '')])
     ([], ['g/h', '.', 'g', '.'])
     >>> _rootsanddirs(\
         [('relpath', 'r', ''), ('path', 'p/p', ''), ('path', '', '')])
     (['r', 'p/p', '.'], ['p', '.'])
     >>> _rootsanddirs(\
         [('relglob', 'rg*', ''), ('re', 're/', ''), ('relre', 'rr', '')])
     (['.', '.', '.'], ['.'])
     '''
     r, d = _patternrootsanddirs(kindpats)
 
     # Append the parents as non-recursive/exact directories, since they must be
     # scanned to get to either the roots or the other exact directories.
     d.extend(util.dirs(d))
     d.extend(util.dirs(r))
     # util.dirs() does not include the root directory, so add it manually
     d.append('.')
 
     return r, d
 
 def _explicitfiles(kindpats):
     '''Returns the potential explicit filenames from the patterns.
 
     >>> _explicitfiles([('path', 'foo/bar', '')])
     ['foo/bar']
     >>> _explicitfiles([('rootfilesin', 'foo/bar', '')])
     []
     '''
     # Keep only the pattern kinds where one can specify filenames (vs only
     # directory names).
     filable = [kp for kp in kindpats if kp[0] not in ('rootfilesin',)]
     return _roots(filable)
 
 def _anypats(kindpats):
     for kind, pat, source in kindpats:
         if kind in ('glob', 're', 'relglob', 'relre', 'set', 'rootfilesin'):
             return True
 
 _commentre = None
 
 def readpatternfile(filepath, warn, sourceinfo=False):
     '''parse a pattern file, returning a list of
     patterns. These patterns should be given to compile()
     to be validated and converted into a match function.
 
     trailing white space is dropped.
     the escape character is backslash.
     comments start with #.
     empty lines are skipped.
 
     lines can be of the following formats:
 
     syntax: regexp # defaults following lines to non-rooted regexps
     syntax: glob   # defaults following lines to non-rooted globs
     re:pattern     # non-rooted regular expression
     glob:pattern   # non-rooted glob
     pattern        # pattern of the current default type
 
     if sourceinfo is set, returns a list of tuples:
     (pattern, lineno, originalline). This is useful to debug ignore patterns.
     '''
 
     syntaxes = {'re': 'relre:', 'regexp': 'relre:', 'glob': 'relglob:',
                 'include': 'include', 'subinclude': 'subinclude'}
     syntax = 'relre:'
     patterns = []
 
     fp = open(filepath, 'rb')
     for lineno, line in enumerate(util.iterfile(fp), start=1):
         if "#" in line:
             global _commentre
             if not _commentre:
                 _commentre = util.re.compile(br'((?:^|[^\\])(?:\\\\)*)#.*')
             # remove comments prefixed by an even number of escapes
             m = _commentre.search(line)
             if m:
                 line = line[:m.end(1)]
             # fixup properly escaped comments that survived the above
             line = line.replace("\\#", "#")
         line = line.rstrip()
         if not line:
             continue
 
         if line.startswith('syntax:'):
             s = line[7:].strip()
             try:
                 syntax = syntaxes[s]
             except KeyError:
                 if warn:
                     warn(_("%s: ignoring invalid syntax '%s'\n") %
                          (filepath, s))
             continue
 
         linesyntax = syntax
         for s, rels in syntaxes.iteritems():
             if line.startswith(rels):
                 linesyntax = rels
                 line = line[len(rels):]
                 break
             elif line.startswith(s+':'):
                 linesyntax = rels
                 line = line[len(s) + 1:]
                 break
         if sourceinfo:
             patterns.append((linesyntax + line, lineno, line))
         else:
             patterns.append(linesyntax + line)
     fp.close()
     return patterns

tests/test-walk.t

@@ -1,518 +1,518
   $ hg init t
   $ cd t
   $ mkdir -p beans
   $ for b in kidney navy turtle borlotti black pinto; do
   >     echo $b > beans/$b
   > done
   $ mkdir -p mammals/Procyonidae
   $ for m in cacomistle coatimundi raccoon; do
   >     echo $m > mammals/Procyonidae/$m
   > done
   $ echo skunk > mammals/skunk
   $ echo fennel > fennel
   $ echo fenugreek > fenugreek
   $ echo fiddlehead > fiddlehead
   $ hg addremove
   adding beans/black
   adding beans/borlotti
   adding beans/kidney
   adding beans/navy
   adding beans/pinto
   adding beans/turtle
   adding fennel
   adding fenugreek
   adding fiddlehead
   adding mammals/Procyonidae/cacomistle
   adding mammals/Procyonidae/coatimundi
   adding mammals/Procyonidae/raccoon
   adding mammals/skunk
   $ hg commit -m "commit #0"
 
   $ hg debugwalk
   matcher: <matcher files=[], patterns=None, includes=None>
   f  beans/black                     beans/black
   f  beans/borlotti                  beans/borlotti
   f  beans/kidney                    beans/kidney
   f  beans/navy                      beans/navy
   f  beans/pinto                     beans/pinto
   f  beans/turtle                    beans/turtle
   f  fennel                          fennel
   f  fenugreek                       fenugreek
   f  fiddlehead                      fiddlehead
   f  mammals/Procyonidae/cacomistle  mammals/Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  mammals/Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     mammals/Procyonidae/raccoon
   f  mammals/skunk                   mammals/skunk
   $ hg debugwalk -I.
   matcher: <matcher files=[], patterns=None, includes='(?:)'>
   f  beans/black                     beans/black
   f  beans/borlotti                  beans/borlotti
   f  beans/kidney                    beans/kidney
   f  beans/navy                      beans/navy
   f  beans/pinto                     beans/pinto
   f  beans/turtle                    beans/turtle
   f  fennel                          fennel
   f  fenugreek                       fenugreek
   f  fiddlehead                      fiddlehead
   f  mammals/Procyonidae/cacomistle  mammals/Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  mammals/Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     mammals/Procyonidae/raccoon
   f  mammals/skunk                   mammals/skunk
 
   $ cd mammals
   $ hg debugwalk
   matcher: <matcher files=[], patterns=None, includes=None>
   f  beans/black                     ../beans/black
   f  beans/borlotti                  ../beans/borlotti
   f  beans/kidney                    ../beans/kidney
   f  beans/navy                      ../beans/navy
   f  beans/pinto                     ../beans/pinto
   f  beans/turtle                    ../beans/turtle
   f  fennel                          ../fennel
   f  fenugreek                       ../fenugreek
   f  fiddlehead                      ../fiddlehead
   f  mammals/Procyonidae/cacomistle  Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     Procyonidae/raccoon
   f  mammals/skunk                   skunk
   $ hg debugwalk -X ../beans
   matcher: <differencematcher m1=<matcher files=[], patterns=None, includes=None>, m2=<matcher files=[], patterns=None, includes='(?:beans(?:/|$))'>>
   f  fennel                          ../fennel
   f  fenugreek                       ../fenugreek
   f  fiddlehead                      ../fiddlehead
   f  mammals/Procyonidae/cacomistle  Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     Procyonidae/raccoon
   f  mammals/skunk                   skunk
   $ hg debugwalk -I '*k'
   matcher: <matcher files=[], patterns=None, includes='(?:mammals\\/[^/]*k(?:/|$))'>
   f  mammals/skunk  skunk
   $ hg debugwalk -I 'glob:*k'
   matcher: <matcher files=[], patterns=None, includes='(?:mammals\\/[^/]*k(?:/|$))'>
   f  mammals/skunk  skunk
   $ hg debugwalk -I 'relglob:*k'
   matcher: <matcher files=[], patterns=None, includes='(?:(?:|.*/)[^/]*k(?:/|$))'>
   f  beans/black    ../beans/black
   f  fenugreek      ../fenugreek
   f  mammals/skunk  skunk
   $ hg debugwalk -I 'relglob:*k' .
-  matcher: <matcher files=['mammals'], patterns='(?:mammals(?:/|$))', includes='(?:(?:|.*/)[^/]*k(?:/|$))'>
+  matcher: <intersectionmatcher m1=<matcher files=['mammals'], patterns='(?:mammals(?:/|$))', includes=None>, m2=<matcher files=[], patterns=None, includes='(?:(?:|.*/)[^/]*k(?:/|$))'>>
   f  mammals/skunk  skunk
   $ hg debugwalk -I 're:.*k$'
   matcher: <matcher files=[], patterns=None, includes='(?:.*k$)'>
   f  beans/black    ../beans/black
   f  fenugreek      ../fenugreek
   f  mammals/skunk  skunk
   $ hg debugwalk -I 'relre:.*k$'
   matcher: <matcher files=[], patterns=None, includes='(?:.*.*k$)'>
   f  beans/black    ../beans/black
   f  fenugreek      ../fenugreek
   f  mammals/skunk  skunk
   $ hg debugwalk -I 'path:beans'
   matcher: <matcher files=[], patterns=None, includes='(?:^beans(?:/|$))'>
   f  beans/black     ../beans/black
   f  beans/borlotti  ../beans/borlotti
   f  beans/kidney    ../beans/kidney
   f  beans/navy      ../beans/navy
   f  beans/pinto     ../beans/pinto
   f  beans/turtle    ../beans/turtle
   $ hg debugwalk -I 'relpath:detour/../../beans'
   matcher: <matcher files=[], patterns=None, includes='(?:beans(?:/|$))'>
   f  beans/black     ../beans/black
   f  beans/borlotti  ../beans/borlotti
   f  beans/kidney    ../beans/kidney
   f  beans/navy      ../beans/navy
   f  beans/pinto     ../beans/pinto
   f  beans/turtle    ../beans/turtle
 
   $ hg debugwalk 'rootfilesin:'
   matcher: <matcher files=[], patterns='(?:^[^/]+$)', includes=None>
   f  fennel      ../fennel
   f  fenugreek   ../fenugreek
   f  fiddlehead  ../fiddlehead
   $ hg debugwalk -I 'rootfilesin:'
   matcher: <matcher files=[], patterns=None, includes='(?:^[^/]+$)'>
   f  fennel      ../fennel
   f  fenugreek   ../fenugreek
   f  fiddlehead  ../fiddlehead
   $ hg debugwalk 'rootfilesin:.'
   matcher: <matcher files=[], patterns='(?:^[^/]+$)', includes=None>
   f  fennel      ../fennel
   f  fenugreek   ../fenugreek
   f  fiddlehead  ../fiddlehead
   $ hg debugwalk -I 'rootfilesin:.'
   matcher: <matcher files=[], patterns=None, includes='(?:^[^/]+$)'>
   f  fennel      ../fennel
   f  fenugreek   ../fenugreek
   f  fiddlehead  ../fiddlehead
   $ hg debugwalk -X 'rootfilesin:'
   matcher: <differencematcher m1=<matcher files=[], patterns=None, includes=None>, m2=<matcher files=[], patterns=None, includes='(?:^[^/]+$)'>>
   f  beans/black                     ../beans/black
   f  beans/borlotti                  ../beans/borlotti
   f  beans/kidney                    ../beans/kidney
   f  beans/navy                      ../beans/navy
   f  beans/pinto                     ../beans/pinto
   f  beans/turtle                    ../beans/turtle
   f  mammals/Procyonidae/cacomistle  Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     Procyonidae/raccoon
   f  mammals/skunk                   skunk
   $ hg debugwalk 'rootfilesin:fennel'
   matcher: <matcher files=[], patterns='(?:^fennel/[^/]+$)', includes=None>
   $ hg debugwalk -I 'rootfilesin:fennel'
   matcher: <matcher files=[], patterns=None, includes='(?:^fennel/[^/]+$)'>
   $ hg debugwalk 'rootfilesin:skunk'
   matcher: <matcher files=[], patterns='(?:^skunk/[^/]+$)', includes=None>
   $ hg debugwalk -I 'rootfilesin:skunk'
   matcher: <matcher files=[], patterns=None, includes='(?:^skunk/[^/]+$)'>
   $ hg debugwalk 'rootfilesin:beans'
   matcher: <matcher files=[], patterns='(?:^beans/[^/]+$)', includes=None>
   f  beans/black     ../beans/black
   f  beans/borlotti  ../beans/borlotti
   f  beans/kidney    ../beans/kidney
   f  beans/navy      ../beans/navy
   f  beans/pinto     ../beans/pinto
   f  beans/turtle    ../beans/turtle
   $ hg debugwalk -I 'rootfilesin:beans'
   matcher: <matcher files=[], patterns=None, includes='(?:^beans/[^/]+$)'>
   f  beans/black     ../beans/black
   f  beans/borlotti  ../beans/borlotti
   f  beans/kidney    ../beans/kidney
   f  beans/navy      ../beans/navy
   f  beans/pinto     ../beans/pinto
   f  beans/turtle    ../beans/turtle
   $ hg debugwalk 'rootfilesin:mammals'
   matcher: <matcher files=[], patterns='(?:^mammals/[^/]+$)', includes=None>
   f  mammals/skunk  skunk
   $ hg debugwalk -I 'rootfilesin:mammals'
   matcher: <matcher files=[], patterns=None, includes='(?:^mammals/[^/]+$)'>
   f  mammals/skunk  skunk
   $ hg debugwalk 'rootfilesin:mammals/'
   matcher: <matcher files=[], patterns='(?:^mammals/[^/]+$)', includes=None>
   f  mammals/skunk  skunk
   $ hg debugwalk -I 'rootfilesin:mammals/'
   matcher: <matcher files=[], patterns=None, includes='(?:^mammals/[^/]+$)'>
   f  mammals/skunk  skunk
   $ hg debugwalk -X 'rootfilesin:mammals'
   matcher: <differencematcher m1=<matcher files=[], patterns=None, includes=None>, m2=<matcher files=[], patterns=None, includes='(?:^mammals/[^/]+$)'>>
   f  beans/black                     ../beans/black
   f  beans/borlotti                  ../beans/borlotti
   f  beans/kidney                    ../beans/kidney
   f  beans/navy                      ../beans/navy
   f  beans/pinto                     ../beans/pinto
   f  beans/turtle                    ../beans/turtle
   f  fennel                          ../fennel
   f  fenugreek                       ../fenugreek
   f  fiddlehead                      ../fiddlehead
   f  mammals/Procyonidae/cacomistle  Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     Procyonidae/raccoon
 
   $ hg debugwalk .
   matcher: <matcher files=['mammals'], patterns='(?:mammals(?:/|$))', includes=None>
   f  mammals/Procyonidae/cacomistle  Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     Procyonidae/raccoon
   f  mammals/skunk                   skunk
   $ hg debugwalk -I.
   matcher: <matcher files=[], patterns=None, includes='(?:mammals(?:/|$))'>
   f  mammals/Procyonidae/cacomistle  Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     Procyonidae/raccoon
   f  mammals/skunk                   skunk
   $ hg debugwalk Procyonidae
   matcher: <matcher files=['mammals/Procyonidae'], patterns='(?:mammals\\/Procyonidae(?:/|$))', includes=None>
   f  mammals/Procyonidae/cacomistle  Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     Procyonidae/raccoon
 
   $ cd Procyonidae
   $ hg debugwalk .
   matcher: <matcher files=['mammals/Procyonidae'], patterns='(?:mammals\\/Procyonidae(?:/|$))', includes=None>
   f  mammals/Procyonidae/cacomistle  cacomistle
   f  mammals/Procyonidae/coatimundi  coatimundi
   f  mammals/Procyonidae/raccoon     raccoon
   $ hg debugwalk ..
   matcher: <matcher files=['mammals'], patterns='(?:mammals(?:/|$))', includes=None>
   f  mammals/Procyonidae/cacomistle  cacomistle
   f  mammals/Procyonidae/coatimundi  coatimundi
   f  mammals/Procyonidae/raccoon     raccoon
   f  mammals/skunk                   ../skunk
   $ cd ..
 
   $ hg debugwalk ../beans
   matcher: <matcher files=['beans'], patterns='(?:beans(?:/|$))', includes=None>
   f  beans/black     ../beans/black
   f  beans/borlotti  ../beans/borlotti
   f  beans/kidney    ../beans/kidney
   f  beans/navy      ../beans/navy
   f  beans/pinto     ../beans/pinto
   f  beans/turtle    ../beans/turtle
   $ hg debugwalk .
   matcher: <matcher files=['mammals'], patterns='(?:mammals(?:/|$))', includes=None>
   f  mammals/Procyonidae/cacomistle  Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     Procyonidae/raccoon
   f  mammals/skunk                   skunk
   $ hg debugwalk .hg
   abort: path 'mammals/.hg' is inside nested repo 'mammals' (glob)
   [255]
   $ hg debugwalk ../.hg
   abort: path contains illegal component: .hg
   [255]
   $ cd ..
 
   $ hg debugwalk -Ibeans
   matcher: <matcher files=[], patterns=None, includes='(?:beans(?:/|$))'>
   f  beans/black     beans/black
   f  beans/borlotti  beans/borlotti
   f  beans/kidney    beans/kidney
   f  beans/navy      beans/navy
   f  beans/pinto     beans/pinto
   f  beans/turtle    beans/turtle
   $ hg debugwalk -I '{*,{b,m}*/*}k'
   matcher: <matcher files=[], patterns=None, includes='(?:(?:[^/]*|(?:b|m)[^/]*\\/[^/]*)k(?:/|$))'>
   f  beans/black    beans/black
   f  fenugreek      fenugreek
   f  mammals/skunk  mammals/skunk
   $ hg debugwalk -Ibeans mammals
-  matcher: <matcher files=['mammals'], patterns='(?:mammals(?:/|$))', includes='(?:beans(?:/|$))'>
+  matcher: <intersectionmatcher m1=<matcher files=['mammals'], patterns='(?:mammals(?:/|$))', includes=None>, m2=<matcher files=[], patterns=None, includes='(?:beans(?:/|$))'>>
   $ hg debugwalk -Inon-existent
   matcher: <matcher files=[], patterns=None, includes='(?:non\\-existent(?:/|$))'>
   $ hg debugwalk -Inon-existent -Ibeans/black
   matcher: <matcher files=[], patterns=None, includes='(?:non\\-existent(?:/|$)|beans\\/black(?:/|$))'>
   f  beans/black  beans/black
   $ hg debugwalk -Ibeans beans/black
-  matcher: <matcher files=['beans/black'], patterns='(?:beans\\/black(?:/|$))', includes='(?:beans(?:/|$))'>
+  matcher: <intersectionmatcher m1=<matcher files=['beans/black'], patterns='(?:beans\\/black(?:/|$))', includes=None>, m2=<matcher files=[], patterns=None, includes='(?:beans(?:/|$))'>>
   f  beans/black  beans/black  exact
   $ hg debugwalk -Ibeans/black beans
-  matcher: <matcher files=['beans'], patterns='(?:beans(?:/|$))', includes='(?:beans\\/black(?:/|$))'>
+  matcher: <intersectionmatcher m1=<matcher files=['beans'], patterns='(?:beans(?:/|$))', includes=None>, m2=<matcher files=[], patterns=None, includes='(?:beans\\/black(?:/|$))'>>
   f  beans/black  beans/black
   $ hg debugwalk -Xbeans/black beans
   matcher: <differencematcher m1=<matcher files=['beans'], patterns='(?:beans(?:/|$))', includes=None>, m2=<matcher files=[], patterns=None, includes='(?:beans\\/black(?:/|$))'>>
   f  beans/borlotti  beans/borlotti
   f  beans/kidney    beans/kidney
   f  beans/navy      beans/navy
   f  beans/pinto     beans/pinto
   f  beans/turtle    beans/turtle
   $ hg debugwalk -Xbeans/black -Ibeans
   matcher: <differencematcher m1=<matcher files=[], patterns=None, includes='(?:beans(?:/|$))'>, m2=<matcher files=[], patterns=None, includes='(?:beans\\/black(?:/|$))'>>
   f  beans/borlotti  beans/borlotti
   f  beans/kidney    beans/kidney
   f  beans/navy      beans/navy
   f  beans/pinto     beans/pinto
   f  beans/turtle    beans/turtle
   $ hg debugwalk -Xbeans/black beans/black
   matcher: <differencematcher m1=<matcher files=['beans/black'], patterns='(?:beans\\/black(?:/|$))', includes=None>, m2=<matcher files=[], patterns=None, includes='(?:beans\\/black(?:/|$))'>>
   f  beans/black  beans/black  exact
   $ hg debugwalk -Xbeans/black -Ibeans/black
   matcher: <differencematcher m1=<matcher files=[], patterns=None, includes='(?:beans\\/black(?:/|$))'>, m2=<matcher files=[], patterns=None, includes='(?:beans\\/black(?:/|$))'>>
   $ hg debugwalk -Xbeans beans/black
   matcher: <differencematcher m1=<matcher files=['beans/black'], patterns='(?:beans\\/black(?:/|$))', includes=None>, m2=<matcher files=[], patterns=None, includes='(?:beans(?:/|$))'>>
   f  beans/black  beans/black  exact
   $ hg debugwalk -Xbeans -Ibeans/black
   matcher: <differencematcher m1=<matcher files=[], patterns=None, includes='(?:beans\\/black(?:/|$))'>, m2=<matcher files=[], patterns=None, includes='(?:beans(?:/|$))'>>
   $ hg debugwalk 'glob:mammals/../beans/b*'
   matcher: <matcher files=['beans'], patterns='(?:beans\\/b[^/]*$)', includes=None>
   f  beans/black     beans/black
   f  beans/borlotti  beans/borlotti
   $ hg debugwalk '-X*/Procyonidae' mammals
   matcher: <differencematcher m1=<matcher files=['mammals'], patterns='(?:mammals(?:/|$))', includes=None>, m2=<matcher files=[], patterns=None, includes='(?:[^/]*\\/Procyonidae(?:/|$))'>>
   f  mammals/skunk  mammals/skunk
   $ hg debugwalk path:mammals
   matcher: <matcher files=['mammals'], patterns='(?:^mammals(?:/|$))', includes=None>
   f  mammals/Procyonidae/cacomistle  mammals/Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  mammals/Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     mammals/Procyonidae/raccoon
   f  mammals/skunk                   mammals/skunk
   $ hg debugwalk ..
   abort: .. not under root '$TESTTMP/t' (glob)
   [255]
   $ hg debugwalk beans/../..
   abort: beans/../.. not under root '$TESTTMP/t' (glob)
   [255]
   $ hg debugwalk .hg
   abort: path contains illegal component: .hg
   [255]
   $ hg debugwalk beans/../.hg
   abort: path contains illegal component: .hg
   [255]
   $ hg debugwalk beans/../.hg/data
   abort: path contains illegal component: .hg/data (glob)
   [255]
   $ hg debugwalk beans/.hg
   abort: path 'beans/.hg' is inside nested repo 'beans' (glob)
   [255]
 
 Test absolute paths:
 
   $ hg debugwalk `pwd`/beans
   matcher: <matcher files=['beans'], patterns='(?:beans(?:/|$))', includes=None>
   f  beans/black     beans/black
   f  beans/borlotti  beans/borlotti
   f  beans/kidney    beans/kidney
   f  beans/navy      beans/navy
   f  beans/pinto     beans/pinto
   f  beans/turtle    beans/turtle
   $ hg debugwalk `pwd`/..
   abort: $TESTTMP/t/.. not under root '$TESTTMP/t' (glob)
   [255]
 
 Test patterns:
 
   $ hg debugwalk glob:\*
   matcher: <matcher files=['.'], patterns='(?:[^/]*$)', includes=None>
   f  fennel      fennel
   f  fenugreek   fenugreek
   f  fiddlehead  fiddlehead
 #if eol-in-paths
   $ echo glob:glob > glob:glob
   $ hg addremove
   adding glob:glob
   warning: filename contains ':', which is reserved on Windows: 'glob:glob'
   $ hg debugwalk glob:\*
   matcher: <matcher files=['.'], patterns='(?:[^/]*$)', includes=None>
   f  fennel      fennel
   f  fenugreek   fenugreek
   f  fiddlehead  fiddlehead
   f  glob:glob   glob:glob
   $ hg debugwalk glob:glob
   matcher: <matcher files=['glob'], patterns='(?:glob$)', includes=None>
   glob: No such file or directory
   $ hg debugwalk glob:glob:glob
   matcher: <matcher files=['glob:glob'], patterns='(?:glob\\:glob$)', includes=None>
   f  glob:glob  glob:glob  exact
   $ hg debugwalk path:glob:glob
   matcher: <matcher files=['glob:glob'], patterns='(?:^glob\\:glob(?:/|$))', includes=None>
   f  glob:glob  glob:glob  exact
   $ rm glob:glob
   $ hg addremove
   removing glob:glob
 #endif
 
   $ hg debugwalk 'glob:**e'
   matcher: <matcher files=['.'], patterns='(?:.*e$)', includes=None>
   f  beans/turtle                    beans/turtle
   f  mammals/Procyonidae/cacomistle  mammals/Procyonidae/cacomistle
 
   $ hg debugwalk 're:.*[kb]$'
   matcher: <matcher files=['.'], patterns='(?:.*[kb]$)', includes=None>
   f  beans/black    beans/black
   f  fenugreek      fenugreek
   f  mammals/skunk  mammals/skunk
 
   $ hg debugwalk path:beans/black
   matcher: <matcher files=['beans/black'], patterns='(?:^beans\\/black(?:/|$))', includes=None>
   f  beans/black  beans/black  exact
   $ hg debugwalk path:beans//black
   matcher: <matcher files=['beans/black'], patterns='(?:^beans\\/black(?:/|$))', includes=None>
   f  beans/black  beans/black  exact
 
   $ hg debugwalk relglob:Procyonidae
   matcher: <matcher files=['.'], patterns='(?:(?:|.*/)Procyonidae$)', includes=None>
   $ hg debugwalk 'relglob:Procyonidae/**'
   matcher: <matcher files=['.'], patterns='(?:(?:|.*/)Procyonidae\\/.*$)', includes=None>
   f  mammals/Procyonidae/cacomistle  mammals/Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  mammals/Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     mammals/Procyonidae/raccoon
   $ hg debugwalk 'relglob:Procyonidae/**' fennel
   matcher: <matcher files=['.', 'fennel'], patterns='(?:(?:|.*/)Procyonidae\\/.*$|fennel(?:/|$))', includes=None>
   f  fennel                          fennel                          exact
   f  mammals/Procyonidae/cacomistle  mammals/Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  mammals/Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     mammals/Procyonidae/raccoon
   $ hg debugwalk beans 'glob:beans/*'
   matcher: <matcher files=['beans', 'beans'], patterns='(?:beans(?:/|$)|beans\\/[^/]*$)', includes=None>
   f  beans/black     beans/black
   f  beans/borlotti  beans/borlotti
   f  beans/kidney    beans/kidney
   f  beans/navy      beans/navy
   f  beans/pinto     beans/pinto
   f  beans/turtle    beans/turtle
   $ hg debugwalk 'glob:mamm**'
   matcher: <matcher files=['.'], patterns='(?:mamm.*$)', includes=None>
   f  mammals/Procyonidae/cacomistle  mammals/Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  mammals/Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     mammals/Procyonidae/raccoon
   f  mammals/skunk                   mammals/skunk
   $ hg debugwalk 'glob:mamm**' fennel
   matcher: <matcher files=['.', 'fennel'], patterns='(?:mamm.*$|fennel(?:/|$))', includes=None>
   f  fennel                          fennel                          exact
   f  mammals/Procyonidae/cacomistle  mammals/Procyonidae/cacomistle
   f  mammals/Procyonidae/coatimundi  mammals/Procyonidae/coatimundi
   f  mammals/Procyonidae/raccoon     mammals/Procyonidae/raccoon
   f  mammals/skunk                   mammals/skunk
   $ hg debugwalk 'glob:j*'
   matcher: <matcher files=['.'], patterns='(?:j[^/]*$)', includes=None>
   $ hg debugwalk NOEXIST
   matcher: <matcher files=['NOEXIST'], patterns='(?:NOEXIST(?:/|$))', includes=None>
   NOEXIST: * (glob)
 
 #if fifo
   $ mkfifo fifo
   $ hg debugwalk fifo
   matcher: <matcher files=['fifo'], patterns='(?:fifo(?:/|$))', includes=None>
   fifo: unsupported file type (type is fifo)
 #endif
 
   $ rm fenugreek
   $ hg debugwalk fenugreek
   matcher: <matcher files=['fenugreek'], patterns='(?:fenugreek(?:/|$))', includes=None>
   f  fenugreek  fenugreek  exact
   $ hg rm fenugreek
   $ hg debugwalk fenugreek
   matcher: <matcher files=['fenugreek'], patterns='(?:fenugreek(?:/|$))', includes=None>
   f  fenugreek  fenugreek  exact
   $ touch new
   $ hg debugwalk new
   matcher: <matcher files=['new'], patterns='(?:new(?:/|$))', includes=None>
   f  new  new  exact
 
   $ mkdir ignored
   $ touch ignored/file
   $ echo '^ignored$' > .hgignore
   $ hg debugwalk ignored
   matcher: <matcher files=['ignored'], patterns='(?:ignored(?:/|$))', includes=None>
   $ hg debugwalk ignored/file
   matcher: <matcher files=['ignored/file'], patterns='(?:ignored\\/file(?:/|$))', includes=None>
   f  ignored/file  ignored/file  exact
 
 Test listfile and listfile0
 
   $ $PYTHON -c "file('listfile0', 'wb').write('fenugreek\0new\0')"
   $ hg debugwalk -I 'listfile0:listfile0'
   matcher: <matcher files=[], patterns=None, includes='(?:fenugreek(?:/|$)|new(?:/|$))'>
   f  fenugreek  fenugreek
   f  new        new
   $ $PYTHON -c "file('listfile', 'wb').write('fenugreek\nnew\r\nmammals/skunk\n')"
   $ hg debugwalk -I 'listfile:listfile'
   matcher: <matcher files=[], patterns=None, includes='(?:fenugreek(?:/|$)|new(?:/|$)|mammals\\/skunk(?:/|$))'>
   f  fenugreek      fenugreek
   f  mammals/skunk  mammals/skunk
   f  new            new
 
   $ cd ..
   $ hg debugwalk -R t t/mammals/skunk
   matcher: <matcher files=['mammals/skunk'], patterns='(?:mammals\\/skunk(?:/|$))', includes=None>
   f  mammals/skunk  t/mammals/skunk  exact
   $ mkdir t2
   $ cd t2
   $ hg debugwalk -R ../t ../t/mammals/skunk
   matcher: <matcher files=['mammals/skunk'], patterns='(?:mammals\\/skunk(?:/|$))', includes=None>
   f  mammals/skunk  ../t/mammals/skunk  exact
   $ hg debugwalk --cwd ../t mammals/skunk
   matcher: <matcher files=['mammals/skunk'], patterns='(?:mammals\\/skunk(?:/|$))', includes=None>
   f  mammals/skunk  mammals/skunk  exact
 
   $ cd ..
 
 Test split patterns on overflow
 
   $ cd t
   $ echo fennel > overflow.list
   $ $PYTHON -c "for i in xrange(20000 / 100): print 'x' * 100" >> overflow.list
   $ echo fenugreek >> overflow.list
   $ hg debugwalk 'listfile:overflow.list' 2>&1 | egrep -v '(^matcher: |^xxx)'
   f  fennel     fennel     exact
   f  fenugreek  fenugreek  exact
   $ cd ..