# match.py - filename matching

#

#  Copyright 2008, 2009 Matt Mackall <mpm@selenic.com> 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.Abort(_("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):

    """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

    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

    """

    return matcher(root, cwd, _donormalize, patterns, include=include,

                   exclude=exclude, default=default, exact=exact,

                   auditor=auditor, ctx=ctx, listsubrepos=listsubrepos,

                   warn=warn, badfn=badfn)

def icasefsmatch(root, cwd, patterns, include=None, exclude=None,

                 default='glob', auditor=None, ctx=None,

                 listsubrepos=False, badfn=None):

    """A matcher for wdir on case insensitive filesystems, which normalizes the

    given patterns to the case in the filesystem.

    """

    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

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 matcher(object):

    def __init__(self, root, cwd, normalize, patterns, include=None,

                 exclude=None, default='glob', exact=False, auditor=None,

                 ctx=None, listsubrepos=False, warn=None, badfn=None):

        if include is None:

            include = []

        if exclude is None:

            exclude = []

        self._root = root

        self._cwd = cwd

        self._files = [] # exact files and roots of patterns

        self._anypats = bool(include or exclude)

        self._always = False

        self._pathrestricted = bool(include or exclude or patterns)

        # roots are directories which are recursively included/excluded.

        self._includeroots = set()

        self._excluderoots = set()

        # dirs are directories which are non-recursively included.

        self._includedirs = set()

        if badfn is not None:

            self.bad = badfn

        matchfns = []

        if include:

            kindpats = normalize(include, 'glob', root, cwd, auditor, warn)

            self.includepat, im = _buildmatch(ctx, kindpats, '(?:/|$)',

                                              listsubrepos, root)

            roots, dirs = _rootsanddirs(kindpats)

            self._includeroots.update(roots)

            self._includedirs.update(dirs)

            matchfns.append(im)

        if exclude:

            kindpats = normalize(exclude, 'glob', root, cwd, auditor, warn)

            self.excludepat, em = _buildmatch(ctx, kindpats, '(?:/|$)',

                                              listsubrepos, root)

            if not _anypats(kindpats):

                # Only consider recursive excludes as such - if a non-recursive

                # exclude is used, we must still recurse into the excluded

                # directory, at least to find subdirectories. In such a case,

                # the regex still won't match the non-recursively-excluded

                # files.

                self._excluderoots.update(_roots(kindpats))

            matchfns.append(lambda f: not em(f))

        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

    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._pathrestricted and self.rel(f)) or self.abs(f)

    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)

    @propertycache

    def _dirs(self):

        return set(util.dirs(self._fileset)) | {'.'}

    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.

        '''

        if self.prefix() and dir in self._fileset:

            return 'all'

        if dir in self._excluderoots:

            return False

        if ((self._includeroots or self._includedirs) and

            '.' 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 exact(self, f):

        '''Returns True if f is in .files().'''

        return f in self._fileset

    def anypats(self):

        '''Matcher uses patterns or include/exclude.'''

        return self._anypats

    def always(self):

        '''Matcher will match everything and .files() will be empty

        - optimization might be possible and necessary.'''