match: write forceincludematcher using unionmatcher

The forceincludematcher is simply a unionmatcher of a includematcher
(matching paths recursively) with the given matcher. Since the
forceincludematcher is only used by sparse, move it there.

I don't have a good sparse repo setup to test performance impact on.

Differential Revision: https://phab.mercurial-scm.org/D57

mercurial/match.py

@@ -1,1027 +1,1013
 # 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=None, 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
 
     if exact:
         m = exactmatcher(root, cwd, patterns, badfn)
     elif patterns:
         kindpats = normalize(patterns, default, root, cwd, auditor, warn)
         if _kindpatsalwaysmatch(kindpats):
             m = alwaysmatcher(root, cwd, badfn, relativeuipath=True)
         else:
             m = patternmatcher(root, cwd, kindpats, ctx=ctx,
                                listsubrepos=listsubrepos, badfn=badfn)
     else:
         # It's a little strange that no patterns means to match everything.
         # Consider changing this to match nothing (probably using nevermatcher).
         m = alwaysmatcher(root, cwd, badfn)
 
     if include:
         kindpats = normalize(include, 'glob', root, cwd, auditor, warn)
         im = includematcher(root, cwd, kindpats, ctx=ctx,
                             listsubrepos=listsubrepos, badfn=None)
         m = intersectmatchers(m, im)
     if exclude:
         kindpats = normalize(exclude, 'glob', root, cwd, auditor, warn)
         em = includematcher(root, cwd, kindpats, ctx=ctx,
                             listsubrepos=listsubrepos, badfn=None)
         m = differencematcher(m, em)
     return m
 
 def exact(root, cwd, files, badfn=None):
     return exactmatcher(root, cwd, files, badfn=badfn)
 
 def always(root, cwd):
     return alwaysmatcher(root, cwd)
 
 def never(root, cwd):
     return nevermatcher(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 always(self):
         '''Matcher will match everything and .files() will be empty --
         optimization might be possible.'''
         return False
 
     def isexact(self):
         '''Matcher will match exactly the list of files in .files() --
         optimization might be possible.'''
         return False
 
     def prefix(self):
         '''Matcher will match the paths in .files() recursively --
         optimization might be possible.'''
         return False
 
     def anypats(self):
         '''None of .always(), .isexact(), and .prefix() is true --
         optimizations will be difficult.'''
         return not self.always() and not self.isexact() and not self.prefix()
 
 class alwaysmatcher(basematcher):
     '''Matches everything.'''
 
     def __init__(self, root, cwd, badfn=None, relativeuipath=False):
         super(alwaysmatcher, self).__init__(root, cwd, badfn,
                                             relativeuipath=relativeuipath)
 
     def always(self):
         return True
 
     def matchfn(self, f):
         return True
 
     def visitdir(self, dir):
         return 'all'
 
     def __repr__(self):
         return '<alwaysmatcher>'
 
 class nevermatcher(basematcher):
     '''Matches nothing.'''
 
     def __init__(self, root, cwd, badfn=None):
         super(nevermatcher, self).__init__(root, cwd, badfn)
 
     # It's a little weird to say that the nevermatcher is an exact matcher
     # or a prefix matcher, but it seems to make sense to let callers take
     # fast paths based on either. There will be no exact matches, nor any
     # prefixes (files() returns []), so fast paths iterating over them should
     # be efficient (and correct).
     def isexact(self):
         return True
 
     def prefix(self):
         return True
 
     def __repr__(self):
         return '<nevermatcher>'
 
 class patternmatcher(basematcher):
 
     def __init__(self, root, cwd, kindpats, ctx=None, listsubrepos=False,
                  badfn=None):
         super(patternmatcher, self).__init__(root, cwd, badfn)
 
         self._files = _explicitfiles(kindpats)
         self._prefix = _prefix(kindpats)
         self._pats, self.matchfn = _buildmatch(ctx, kindpats, '$', listsubrepos,
                                                root)
 
     @propertycache
     def _dirs(self):
         return set(util.dirs(self._fileset)) | {'.'}
 
     def visitdir(self, dir):
         if self._prefix and dir in self._fileset:
             return 'all'
         return ('.' 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 prefix(self):
         return self._prefix
 
     def __repr__(self):
         return ('<patternmatcher patterns=%r>' % self._pats)
 
 class includematcher(basematcher):
 
     def __init__(self, root, cwd, kindpats, ctx=None, listsubrepos=False,
                  badfn=None):
         super(includematcher, self).__init__(root, cwd, badfn)
 
         self._pats, self.matchfn = _buildmatch(ctx, kindpats, '(?:/|$)',
                                                listsubrepos, root)
         self._prefix = _prefix(kindpats)
         roots, dirs = _rootsanddirs(kindpats)
         # roots are directories which are recursively included.
         self._roots = set(roots)
         # dirs are directories which are non-recursively included.
         self._dirs = set(dirs)
 
     def visitdir(self, dir):
         if self._prefix and dir in self._roots:
             return 'all'
         return ('.' in self._roots or
                 dir in self._roots or
                 dir in self._dirs or
                 any(parentdir in self._roots
                     for parentdir in util.finddirs(dir)))
 
     def __repr__(self):
         return ('<includematcher includes=%r>' % self._pats)
 
 class exactmatcher(basematcher):
     '''Matches the input files exactly. They are interpreted as paths, not
     patterns (so no kind-prefixes).
     '''
 
     def __init__(self, root, cwd, files, badfn=None):
         super(exactmatcher, self).__init__(root, cwd, badfn)
 
         if isinstance(files, list):
             self._files = files
         else:
             self._files = list(files)
 
     matchfn = basematcher.exact
 
     @propertycache
     def _dirs(self):
         return set(util.dirs(self._fileset)) | {'.'}
 
     def visitdir(self, dir):
         return dir in self._dirs
 
     def isexact(self):
         return True
 
     def __repr__(self):
         return ('<exactmatcher files=%r>' % self._files)
 
 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 __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 __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 prefix(self):
         return self._matcher.prefix() and not self._always
 
     def __repr__(self):
         return ('<subdirmatcher path=%r, matcher=%r>' %
                 (self._path, self._matcher))
 
-class forceincludematcher(basematcher):
-    """A matcher that returns true for any of the forced includes before testing
-    against the actual matcher."""
-    def __init__(self, matcher, includes):
-        self._matcher = matcher
-        self._includes = includes
-
-    def matchfn(self, f):
-        return f in self._includes or self._matcher(f)
-
-    def __repr__(self):
-        return ('<forceincludematcher matcher=%r, includes=%r>' %
-                (self._matcher, sorted(self._includes)))
-
 class unionmatcher(basematcher):
     """A matcher that is the union of several matchers."""
     def __init__(self, matchers):
         self._matchers = matchers
 
     def matchfn(self, f):
         for match in self._matchers:
             if match(f):
                 return True
         return False
 
     def __repr__(self):
         return ('<unionmatcher matchers=%r>' % self._matchers)
 
 class negatematcher(basematcher):
     def __init__(self, matcher):
         self._matcher = matcher
 
     def matchfn(self, f):
         return not self._matcher(f)
 
     def __repr__(self):
         return ('<negatematcher matcher=%r>' % self._matcher)
 
 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 in ('path', 'relpath'):
         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 == '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 _prefix(kindpats):
     '''Whether all the patterns match a prefix (i.e. recursively)'''
     for kind, pat, source in kindpats:
         if kind not in ('path', 'relpath'):
             return False
     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

mercurial/sparse.py

@@ -1,671 +1,678
 # sparse.py - functionality for sparse checkouts
 #
 # Copyright 2014 Facebook, Inc.
 #
 # 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 collections
 import hashlib
 import os
 
 from .i18n import _
 from .node import nullid
 from . import (
     error,
     match as matchmod,
     merge as mergemod,
     pycompat,
     util,
 )
 
 # Whether sparse features are enabled. This variable is intended to be
 # temporary to facilitate porting sparse to core. It should eventually be
 # a per-repo option, possibly a repo requirement.
 enabled = False
 
 def parseconfig(ui, raw):
     """Parse sparse config file content.
 
     Returns a tuple of includes, excludes, and profiles.
     """
     includes = set()
     excludes = set()
     current = includes
     profiles = []
     for line in raw.split('\n'):
         line = line.strip()
         if not line or line.startswith('#'):
             # empty or comment line, skip
             continue
         elif line.startswith('%include '):
             line = line[9:].strip()
             if line:
                 profiles.append(line)
         elif line == '[include]':
             if current != includes:
                 # TODO pass filename into this API so we can report it.
                 raise error.Abort(_('sparse config cannot have includes ' +
                                     'after excludes'))
             continue
         elif line == '[exclude]':
             current = excludes
         elif line:
             if line.strip().startswith('/'):
                 ui.warn(_('warning: sparse profile cannot use' +
                           ' paths starting with /, ignoring %s\n') % line)
                 continue
             current.add(line)
 
     return includes, excludes, profiles
 
 # Exists as separate function to facilitate monkeypatching.
 def readprofile(repo, profile, changeid):
     """Resolve the raw content of a sparse profile file."""
     # TODO add some kind of cache here because this incurs a manifest
     # resolve and can be slow.
     return repo.filectx(profile, changeid=changeid).data()
 
 def patternsforrev(repo, rev):
     """Obtain sparse checkout patterns for the given rev.
 
     Returns a tuple of iterables representing includes, excludes, and
     patterns.
     """
     # Feature isn't enabled. No-op.
     if not enabled:
         return set(), set(), []
 
     raw = repo.vfs.tryread('sparse')
     if not raw:
         return set(), set(), []
 
     if rev is None:
         raise error.Abort(_('cannot parse sparse patterns from working '
                             'directory'))
 
     includes, excludes, profiles = parseconfig(repo.ui, raw)
     ctx = repo[rev]
 
     if profiles:
         visited = set()
         while profiles:
             profile = profiles.pop()
             if profile in visited:
                 continue
 
             visited.add(profile)
 
             try:
                 raw = readprofile(repo, profile, rev)
             except error.ManifestLookupError:
                 msg = (
                     "warning: sparse profile '%s' not found "
                     "in rev %s - ignoring it\n" % (profile, ctx))
                 # experimental config: sparse.missingwarning
                 if repo.ui.configbool(
                         'sparse', 'missingwarning', True):
                     repo.ui.warn(msg)
                 else:
                     repo.ui.debug(msg)
                 continue
 
             pincludes, pexcludes, subprofs = parseconfig(repo.ui, raw)
             includes.update(pincludes)
             excludes.update(pexcludes)
             for subprofile in subprofs:
                 profiles.append(subprofile)
 
         profiles = visited
 
     if includes:
         includes.add('.hg*')
 
     return includes, excludes, profiles
 
 def activeconfig(repo):
     """Determine the active sparse config rules.
 
     Rules are constructed by reading the current sparse config and bringing in
     referenced profiles from parents of the working directory.
     """
     revs = [repo.changelog.rev(node) for node in
             repo.dirstate.parents() if node != nullid]
 
     allincludes = set()
     allexcludes = set()
     allprofiles = set()
 
     for rev in revs:
         includes, excludes, profiles = patternsforrev(repo, rev)
         allincludes |= includes
         allexcludes |= excludes
         allprofiles |= set(profiles)
 
     return allincludes, allexcludes, allprofiles
 
 def configsignature(repo, includetemp=True):
     """Obtain the signature string for the current sparse configuration.
 
     This is used to construct a cache key for matchers.
     """
     cache = repo._sparsesignaturecache
 
     signature = cache.get('signature')
 
     if includetemp:
         tempsignature = cache.get('tempsignature')
     else:
         tempsignature = '0'
 
     if signature is None or (includetemp and tempsignature is None):
         signature = hashlib.sha1(repo.vfs.tryread('sparse')).hexdigest()
         cache['signature'] = signature
 
         if includetemp:
             raw = repo.vfs.tryread('tempsparse')
             tempsignature = hashlib.sha1(raw).hexdigest()
             cache['tempsignature'] = tempsignature
 
     return '%s %s' % (signature, tempsignature)
 
 def writeconfig(repo, includes, excludes, profiles):
     """Write the sparse config file given a sparse configuration."""
     with repo.vfs('sparse', 'wb') as fh:
         for p in sorted(profiles):
             fh.write('%%include %s\n' % p)
 
         if includes:
             fh.write('[include]\n')
             for i in sorted(includes):
                 fh.write(i)
                 fh.write('\n')
 
         if excludes:
             fh.write('[exclude]\n')
             for e in sorted(excludes):
                 fh.write(e)
                 fh.write('\n')
 
     repo._sparsesignaturecache.clear()
 
 def readtemporaryincludes(repo):
     raw = repo.vfs.tryread('tempsparse')
     if not raw:
         return set()
 
     return set(raw.split('\n'))
 
 def writetemporaryincludes(repo, includes):
     repo.vfs.write('tempsparse', '\n'.join(sorted(includes)))
     repo._sparsesignaturecache.clear()
 
 def addtemporaryincludes(repo, additional):
     includes = readtemporaryincludes(repo)
     for i in additional:
         includes.add(i)
     writetemporaryincludes(repo, includes)
 
 def prunetemporaryincludes(repo):
     if not enabled or not repo.vfs.exists('tempsparse'):
         return
 
     s = repo.status()
     if s.modified or s.added or s.removed or s.deleted:
         # Still have pending changes. Don't bother trying to prune.
         return
 
     sparsematch = matcher(repo, includetemp=False)
     dirstate = repo.dirstate
     actions = []
     dropped = []
     tempincludes = readtemporaryincludes(repo)
     for file in tempincludes:
         if file in dirstate and not sparsematch(file):
             message = _('dropping temporarily included sparse files')
             actions.append((file, None, message))
             dropped.append(file)
 
     typeactions = collections.defaultdict(list)
     typeactions['r'] = actions
     mergemod.applyupdates(repo, typeactions, repo[None], repo['.'], False)
 
     # Fix dirstate
     for file in dropped:
         dirstate.drop(file)
 
     repo.vfs.unlink('tempsparse')
     repo._sparsesignaturecache.clear()
     msg = _('cleaned up %d temporarily added file(s) from the '
             'sparse checkout\n')
     repo.ui.status(msg % len(tempincludes))
 
+def forceincludematcher(matcher, includes):
+    """Returns a matcher that returns true for any of the forced includes
+    before testing against the actual matcher."""
+    kindpats = [('path', include, '') for include in includes]
+    includematcher = matchmod.includematcher('', '', kindpats)
+    return matchmod.unionmatcher([includematcher, matcher])
+
 def matcher(repo, revs=None, includetemp=True):
     """Obtain a matcher for sparse working directories for the given revs.
 
     If multiple revisions are specified, the matcher is the union of all
     revs.
 
     ``includetemp`` indicates whether to use the temporary sparse profile.
     """
     # If sparse isn't enabled, sparse matcher matches everything.
     if not enabled:
         return matchmod.always(repo.root, '')
 
     if not revs or revs == [None]:
         revs = [repo.changelog.rev(node)
                 for node in repo.dirstate.parents() if node != nullid]
 
     signature = configsignature(repo, includetemp=includetemp)
 
     key = '%s %s' % (signature, ' '.join(map(pycompat.bytestr, revs)))
 
     result = repo._sparsematchercache.get(key)
     if result:
         return result
 
     matchers = []
     for rev in revs:
         try:
             includes, excludes, profiles = patternsforrev(repo, rev)
 
             if includes or excludes:
                 # Explicitly include subdirectories of includes so
                 # status will walk them down to the actual include.
                 subdirs = set()
                 for include in includes:
                     # TODO consider using posix path functions here so Windows
                     # \ directory separators don't come into play.
                     dirname = os.path.dirname(include)
                     # basename is used to avoid issues with absolute
                     # paths (which on Windows can include the drive).
                     while os.path.basename(dirname):
                         subdirs.add(dirname)
                         dirname = os.path.dirname(dirname)
 
                 matcher = matchmod.match(repo.root, '', [],
                                          include=includes, exclude=excludes,
                                          default='relpath')
                 if subdirs:
-                    matcher = matchmod.forceincludematcher(matcher, subdirs)
+                    matcher = forceincludematcher(matcher, subdirs)
                 matchers.append(matcher)
         except IOError:
             pass
 
     if not matchers:
         result = matchmod.always(repo.root, '')
     elif len(matchers) == 1:
         result = matchers[0]
     else:
         result = matchmod.unionmatcher(matchers)
 
     if includetemp:
         tempincludes = readtemporaryincludes(repo)
-        result = matchmod.forceincludematcher(result, tempincludes)
+        result = forceincludematcher(result, tempincludes)
 
     repo._sparsematchercache[key] = result
 
     return result
 
 def filterupdatesactions(repo, wctx, mctx, branchmerge, actions):
     """Filter updates to only lay out files that match the sparse rules."""
     if not enabled:
         return actions
 
     oldrevs = [pctx.rev() for pctx in wctx.parents()]
     oldsparsematch = matcher(repo, oldrevs)
 
     if oldsparsematch.always():
         return actions
 
     files = set()
     prunedactions = {}
 
     if branchmerge:
         # If we're merging, use the wctx filter, since we're merging into
         # the wctx.
         sparsematch = matcher(repo, [wctx.parents()[0].rev()])
     else:
         # If we're updating, use the target context's filter, since we're
         # moving to the target context.
         sparsematch = matcher(repo, [mctx.rev()])
 
     temporaryfiles = []
     for file, action in actions.iteritems():
         type, args, msg = action
         files.add(file)
         if sparsematch(file):
             prunedactions[file] = action
         elif type == 'm':
             temporaryfiles.append(file)
             prunedactions[file] = action
         elif branchmerge:
             if type != 'k':
                 temporaryfiles.append(file)
                 prunedactions[file] = action
         elif type == 'f':
             prunedactions[file] = action
         elif file in wctx:
             prunedactions[file] = ('r', args, msg)
 
     if len(temporaryfiles) > 0:
         repo.ui.status(_('temporarily included %d file(s) in the sparse '
                          'checkout for merging\n') % len(temporaryfiles))
         addtemporaryincludes(repo, temporaryfiles)
 
         # Add the new files to the working copy so they can be merged, etc
         actions = []
         message = 'temporarily adding to sparse checkout'
         wctxmanifest = repo[None].manifest()
         for file in temporaryfiles:
             if file in wctxmanifest:
                 fctx = repo[None][file]
                 actions.append((file, (fctx.flags(), False), message))
 
         typeactions = collections.defaultdict(list)
         typeactions['g'] = actions
         mergemod.applyupdates(repo, typeactions, repo[None], repo['.'],
                               False)
 
         dirstate = repo.dirstate
         for file, flags, msg in actions:
             dirstate.normal(file)
 
     profiles = activeconfig(repo)[2]
     changedprofiles = profiles & files
     # If an active profile changed during the update, refresh the checkout.
     # Don't do this during a branch merge, since all incoming changes should
     # have been handled by the temporary includes above.
     if changedprofiles and not branchmerge:
         mf = mctx.manifest()
         for file in mf:
             old = oldsparsematch(file)
             new = sparsematch(file)
             if not old and new:
                 flags = mf.flags(file)
                 prunedactions[file] = ('g', (flags, False), '')
             elif old and not new:
                 prunedactions[file] = ('r', [], '')
 
     return prunedactions
 
 def refreshwdir(repo, origstatus, origsparsematch, force=False):
     """Refreshes working directory by taking sparse config into account.
 
     The old status and sparse matcher is compared against the current sparse
     matcher.
 
     Will abort if a file with pending changes is being excluded or included
     unless ``force`` is True.
     """
     # Verify there are no pending changes
     pending = set()
     pending.update(origstatus.modified)
     pending.update(origstatus.added)
     pending.update(origstatus.removed)
     sparsematch = matcher(repo)
     abort = False
 
     for f in pending:
         if not sparsematch(f):
             repo.ui.warn(_("pending changes to '%s'\n") % f)
             abort = not force
 
     if abort:
         raise error.Abort(_('could not update sparseness due to pending '
                             'changes'))
 
     # Calculate actions
     dirstate = repo.dirstate
     ctx = repo['.']
     added = []
     lookup = []
     dropped = []
     mf = ctx.manifest()
     files = set(mf)
 
     actions = {}
 
     for file in files:
         old = origsparsematch(file)
         new = sparsematch(file)
         # Add files that are newly included, or that don't exist in
         # the dirstate yet.
         if (new and not old) or (old and new and not file in dirstate):
             fl = mf.flags(file)
             if repo.wvfs.exists(file):
                 actions[file] = ('e', (fl,), '')
                 lookup.append(file)
             else:
                 actions[file] = ('g', (fl, False), '')
                 added.append(file)
         # Drop files that are newly excluded, or that still exist in
         # the dirstate.
         elif (old and not new) or (not old and not new and file in dirstate):
             dropped.append(file)
             if file not in pending:
                 actions[file] = ('r', [], '')
 
     # Verify there are no pending changes in newly included files
     abort = False
     for file in lookup:
         repo.ui.warn(_("pending changes to '%s'\n") % file)
         abort = not force
     if abort:
         raise error.Abort(_('cannot change sparseness due to pending '
                             'changes (delete the files or use '
                             '--force to bring them back dirty)'))
 
     # Check for files that were only in the dirstate.
     for file, state in dirstate.iteritems():
         if not file in files:
             old = origsparsematch(file)
             new = sparsematch(file)
             if old and not new:
                 dropped.append(file)
 
     # Apply changes to disk
     typeactions = dict((m, []) for m in 'a f g am cd dc r dm dg m e k'.split())
     for f, (m, args, msg) in actions.iteritems():
         if m not in typeactions:
             typeactions[m] = []
         typeactions[m].append((f, args, msg))
 
     mergemod.applyupdates(repo, typeactions, repo[None], repo['.'], False)
 
     # Fix dirstate
     for file in added:
         dirstate.normal(file)
 
     for file in dropped:
         dirstate.drop(file)
 
     for file in lookup:
         # File exists on disk, and we're bringing it back in an unknown state.
         dirstate.normallookup(file)
 
     return added, dropped, lookup
 
 def aftercommit(repo, node):
     """Perform actions after a working directory commit."""
     # This function is called unconditionally, even if sparse isn't
     # enabled.
     ctx = repo[node]
 
     profiles = patternsforrev(repo, ctx.rev())[2]
 
     # profiles will only have data if sparse is enabled.
     if set(profiles) & set(ctx.files()):
         origstatus = repo.status()
         origsparsematch = matcher(repo)
         refreshwdir(repo, origstatus, origsparsematch, force=True)
 
     prunetemporaryincludes(repo)
 
 def clearrules(repo, force=False):
     """Clears include/exclude rules from the sparse config.
 
     The remaining sparse config only has profiles, if defined. The working
     directory is refreshed, as needed.
     """
     with repo.wlock():
         raw = repo.vfs.tryread('sparse')
         includes, excludes, profiles = parseconfig(repo.ui, raw)
 
         if not includes and not excludes:
             return
 
         oldstatus = repo.status()
         oldmatch = matcher(repo)
         writeconfig(repo, set(), set(), profiles)
         refreshwdir(repo, oldstatus, oldmatch, force=force)
 
 def importfromfiles(repo, opts, paths, force=False):
     """Import sparse config rules from files.
 
     The updated sparse config is written out and the working directory
     is refreshed, as needed.
     """
     with repo.wlock():
         # read current configuration
         raw = repo.vfs.tryread('sparse')
         oincludes, oexcludes, oprofiles = parseconfig(repo.ui, raw)
         includes, excludes, profiles = map(
                 set, (oincludes, oexcludes, oprofiles))
 
         aincludes, aexcludes, aprofiles = activeconfig(repo)
 
         # Import rules on top; only take in rules that are not yet
         # part of the active rules.
         changed = False
         for p in paths:
             with util.posixfile(util.expandpath(p)) as fh:
                 raw = fh.read()
 
             iincludes, iexcludes, iprofiles = parseconfig(repo.ui, raw)
             oldsize = len(includes) + len(excludes) + len(profiles)
             includes.update(iincludes - aincludes)
             excludes.update(iexcludes - aexcludes)
             profiles.update(set(iprofiles) - aprofiles)
             if len(includes) + len(excludes) + len(profiles) > oldsize:
                 changed = True
 
         profilecount = includecount = excludecount = 0
         fcounts = (0, 0, 0)
 
         if changed:
             profilecount = len(profiles - aprofiles)
             includecount = len(includes - aincludes)
             excludecount = len(excludes - aexcludes)
 
             oldstatus = repo.status()
             oldsparsematch = matcher(repo)
 
             # TODO remove this try..except once the matcher integrates better
             # with dirstate. We currently have to write the updated config
             # because that will invalidate the matcher cache and force a
             # re-read. We ideally want to update the cached matcher on the
             # repo instance then flush the new config to disk once wdir is
             # updated. But this requires massive rework to matcher() and its
             # consumers.
             writeconfig(repo, includes, excludes, profiles)
 
             try:
                 fcounts = map(
                     len,
                     refreshwdir(repo, oldstatus, oldsparsematch, force=force))
             except Exception:
                 writeconfig(repo, oincludes, oexcludes, oprofiles)
                 raise
 
         printchanges(repo.ui, opts, profilecount, includecount, excludecount,
                      *fcounts)
 
 def updateconfig(repo, pats, opts, include=False, exclude=False, reset=False,
                  delete=False, enableprofile=False, disableprofile=False,
                  force=False):
     """Perform a sparse config update.
 
     Only one of the actions may be performed.
 
     The new config is written out and a working directory refresh is performed.
     """
     with repo.wlock():
         oldmatcher = matcher(repo)
 
         raw = repo.vfs.tryread('sparse')
         oldinclude, oldexclude, oldprofiles = parseconfig(repo.ui, raw)
         oldprofiles = set(oldprofiles)
 
         if reset:
             newinclude = set()
             newexclude = set()
             newprofiles = set()
         else:
             newinclude = set(oldinclude)
             newexclude = set(oldexclude)
             newprofiles = set(oldprofiles)
 
         oldstatus = repo.status()
 
         if any(pat.startswith('/') for pat in pats):
             repo.ui.warn(_('warning: paths cannot start with /, ignoring: %s\n')
                          % ([pat for pat in pats if pat.startswith('/')]))
         elif include:
             newinclude.update(pats)
         elif exclude:
             newexclude.update(pats)
         elif enableprofile:
             newprofiles.update(pats)
         elif disableprofile:
             newprofiles.difference_update(pats)
         elif delete:
             newinclude.difference_update(pats)
             newexclude.difference_update(pats)
 
         profilecount = (len(newprofiles - oldprofiles) -
                         len(oldprofiles - newprofiles))
         includecount = (len(newinclude - oldinclude) -
                         len(oldinclude - newinclude))
         excludecount = (len(newexclude - oldexclude) -
                         len(oldexclude - newexclude))
 
         # TODO clean up this writeconfig() + try..except pattern once we can.
         # See comment in importfromfiles() explaining it.
         writeconfig(repo, newinclude, newexclude, newprofiles)
 
         try:
             fcounts = map(
                 len,
                 refreshwdir(repo, oldstatus, oldmatcher, force=force))
 
             printchanges(repo.ui, opts, profilecount, includecount,
                          excludecount, *fcounts)
         except Exception:
             writeconfig(repo, oldinclude, oldexclude, oldprofiles)
             raise
 
 def printchanges(ui, opts, profilecount=0, includecount=0, excludecount=0,
                  added=0, dropped=0, conflicting=0):
     """Print output summarizing sparse config changes."""
     with ui.formatter('sparse', opts) as fm:
         fm.startitem()
         fm.condwrite(ui.verbose, 'profiles_added', _('Profiles changed: %d\n'),
                      profilecount)
         fm.condwrite(ui.verbose, 'include_rules_added',
                      _('Include rules changed: %d\n'), includecount)
         fm.condwrite(ui.verbose, 'exclude_rules_added',
                      _('Exclude rules changed: %d\n'), excludecount)
 
         # In 'plain' verbose mode, mergemod.applyupdates already outputs what
         # files are added or removed outside of the templating formatter
         # framework. No point in repeating ourselves in that case.
         if not fm.isplain():
             fm.condwrite(ui.verbose, 'files_added', _('Files added: %d\n'),
                          added)
             fm.condwrite(ui.verbose, 'files_dropped', _('Files dropped: %d\n'),
                          dropped)
             fm.condwrite(ui.verbose, 'files_conflicting',
                          _('Files conflicting: %d\n'), conflicting)