# 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, print_function

import copy

import itertools

import os

import re

from .i18n import _

from . import (

    encoding,

    error,

    pathutil,

    pycompat,

    util,

)

from .utils import (

    stringutil,

)

allpatternkinds = ('re', 'glob', 'path', 'relglob', 'relpath', 'relre',

                   'rootglob',

                   'listfile', 'listfile0', 'set', 'include', 'subinclude',

                   'rootfilesin')

cwdrelativepatternkinds = ('relpath', 'glob')

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

    '''Returns the kindpats list with the 'set' patterns expanded to matchers'''

    matchers = []

    other = []

    for kind, pat, source in kindpats:

        if kind == 'set':

            if ctx is None:

                raise error.ProgrammingError("fileset expression with no "

                                             "context")

            matchers.append(ctx.matchfileset(pat, badfn=badfn))

            if listsubrepos:

                for subpath in ctx.substate:

                    sm = ctx.sub(subpath).matchfileset(pat, badfn=badfn)

                    pm = prefixdirmatcher(subpath, sm, badfn=badfn)

                    matchers.append(pm)

            continue

        other.append((kind, pat, source))

    return matchers, 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 _buildkindpatsmatcher(matchercls, root, kindpats, ctx=None,

                          listsubrepos=False, badfn=None):

    matchers = []

    fms, kindpats = _expandsets(kindpats, ctx=ctx,

                                listsubrepos=listsubrepos, badfn=badfn)

    if kindpats:

        m = matchercls(root, kindpats, badfn=badfn)

        matchers.append(m)

    if fms:

        matchers.extend(fms)

    if not matchers:

        return nevermatcher(badfn=badfn)

    if len(matchers) == 1:

        return matchers[0]

    return unionmatcher(matchers)

def match(root, cwd, patterns=None, include=None, exclude=None, default='glob',

          auditor=None, ctx=None, listsubrepos=False, warn=None,

          badfn=None, icasefs=False):

    r"""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

    auditor - optional path auditor

    ctx - optional changecontext

    listsubrepos - if True, recurse into subrepositories

    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

    Usually a patternmatcher is returned:

    >>> match(b'foo', b'.', [b're:.*\.c$', b'path:foo/a', b'*.py'])

    <patternmatcher patterns='.*\\.c$|foo/a(?:/|$)|[^/]*\\.py$'>

    Combining 'patterns' with 'include' (resp. 'exclude') gives an

    intersectionmatcher (resp. a differencematcher):

    >>> type(match(b'foo', b'.', [b're:.*\.c$'], include=[b'path:lib']))

    <class 'mercurial.match.intersectionmatcher'>

    >>> type(match(b'foo', b'.', [b're:.*\.c$'], exclude=[b'path:build']))

    <class 'mercurial.match.differencematcher'>

    Notice that, if 'patterns' is empty, an alwaysmatcher is returned:

    >>> match(b'foo', b'.', [])

    <alwaysmatcher>

    The 'default' argument determines which kind of pattern is assumed if a

    pattern has no prefix:

    >>> match(b'foo', b'.', [b'.*\.c$'], default=b're')

    <patternmatcher patterns='.*\\.c$'>

    >>> match(b'foo', b'.', [b'main.py'], default=b'relpath')

    <patternmatcher patterns='main\\.py(?:/|$)'>

    >>> match(b'foo', b'.', [b'main.py'], default=b're')

    <patternmatcher patterns='main.py'>

    The primary use of matchers is to check whether a value (usually a file

    name) matches againset one of the patterns given at initialization. There

    are two ways of doing this check.

    >>> m = match(b'foo', b'', [b're:.*\.c$', b'relpath:a'])

    1. Calling the matcher with a file name returns True if any pattern

    matches that file name:

    >>> m(b'a')

    True

    >>> m(b'main.c')

    True

    >>> m(b'test.py')

    False

    2. Using the exact() method only returns True if the file name matches one

    of the exact patterns (i.e. not re: or glob: patterns):

    >>> m.exact(b'a')

    True

    >>> m.exact(b'main.c')

    False

    """

    normalize = _donormalize

    if icasefs:

        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 patterns:

        kindpats = normalize(patterns, default, root, cwd, auditor, warn)

        if _kindpatsalwaysmatch(kindpats):

            m = alwaysmatcher(badfn)

        else:

            m = _buildkindpatsmatcher(patternmatcher, root, 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(badfn)

    if include:

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

        im = _buildkindpatsmatcher(includematcher, root, kindpats, ctx=ctx,

                                   listsubrepos=listsubrepos, badfn=None)

        m = intersectmatchers(m, im)

    if exclude:

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

        em = _buildkindpatsmatcher(includematcher, root, kindpats, ctx=ctx,

                                   listsubrepos=listsubrepos, badfn=None)

        m = differencematcher(m, em)

    return m

def exact(files, badfn=None):

    return exactmatcher(files, badfn=badfn)

def always(badfn=None):

    return alwaysmatcher(badfn)

def never(badfn=None):

    return nevermatcher(badfn)

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=None, warn=None):

    '''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 cwdrelativepatternkinds:

            pat = pathutil.canonpath(root, cwd, pat, auditor=auditor)

        elif kind in ('relglob', 'path', 'rootfilesin', 'rootglob'):

            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, stringutil.forcebytestr(inst.strerror)))

            continue

        # else: re or relre - which cannot be normalized

        kindpats.append((kind, pat, ''))

    return kindpats

class basematcher(object):

    def __init__(self, badfn=None):

        if badfn is not None:

            self.bad = badfn

    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.'''

    # 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

    @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.

        '''

        return True

    def visitchildrenset(self, dir):

        '''Decides whether a directory should be visited based on whether it

        has potential matches in it or one of its subdirectories, and

        potentially lists which subdirectories of that directory should be

        visited. This is based on the match's primary, included, and excluded

        patterns.

        This function is very similar to 'visitdir', and the following mapping

        can be applied:

             visitdir | visitchildrenlist

            ----------+-------------------

             False    | set()

             'all'    | 'all'

             True     | 'this' OR non-empty set of subdirs -or files- to visit

        Example:

          Assume matchers ['path:foo/bar', 'rootfilesin:qux'], we would return

          the following values (assuming the implementation of visitchildrenset

          is capable of recognizing this; some implementations are not).

          '.' -> {'foo', 'qux'}

          'baz' -> set()

          'foo' -> {'bar'}

          # Ideally this would be 'all', but since the prefix nature of matchers

          # is applied to the entire matcher, we have to downgrade this to

          # 'this' due to the non-prefix 'rootfilesin'-kind matcher being mixed

          # in.

          'foo/bar' -> 'this'

          'qux' -> 'this'

        Important:

          Most matchers do not know if they're representing files or

          directories. They see ['path:dir/f'] and don't know whether 'f' is a

          file or a directory, so visitchildrenset('dir') for most matchers will

          return {'f'}, but if the matcher knows it's a file (like exactmatcher