"""

util.py - Mercurial utility functions and platform specfic implementations

 Copyright 2005 K. Thananchayan <thananck@yahoo.com>

This software may be used and distributed according to the terms

of the GNU General Public License, incorporated herein by reference.

This contains helper routines that are independent of the SCM core and hide

platform-specific details from the core.

"""

import os, errno

from i18n import gettext as _

from demandload import *

demandload(globals(), "cStringIO errno popen2 re shutil sys tempfile")

demandload(globals(), "threading time")

class SignalInterrupt(Exception):

    """Exception raised on SIGTERM and SIGHUP."""

def pipefilter(s, cmd):

    '''filter string S through command CMD, returning its output'''

    (pout, pin) = popen2.popen2(cmd, -1, 'b')

    def writer():

        try:

            pin.write(s)

            pin.close()

        except IOError, inst:

            if inst.errno != errno.EPIPE:

                raise

    # we should use select instead on UNIX, but this will work on most

    # systems, including Windows

    w = threading.Thread(target=writer)

    w.start()

    f = pout.read()

    pout.close()

    w.join()

    return f

def tempfilter(s, cmd):

    '''filter string S through a pair of temporary files with CMD.

    CMD is used as a template to create the real command to be run,

    with the strings INFILE and OUTFILE replaced by the real names of

    the temporary files generated.'''

    inname, outname = None, None

    try:

        infd, inname = tempfile.mkstemp(prefix='hg-filter-in-')

        fp = os.fdopen(infd, 'wb')

        fp.write(s)

        fp.close()

        outfd, outname = tempfile.mkstemp(prefix='hg-filter-out-')

        os.close(outfd)

        cmd = cmd.replace('INFILE', inname)

        cmd = cmd.replace('OUTFILE', outname)

        code = os.system(cmd)

        if code: raise Abort(_("command '%s' failed: %s") %

                             (cmd, explain_exit(code)))

        return open(outname, 'rb').read()

    finally:

        try:

            if inname: os.unlink(inname)

        except: pass

        try:

            if outname: os.unlink(outname)

        except: pass

filtertable = {

    'tempfile:': tempfilter,

    'pipe:': pipefilter,

    }

def filter(s, cmd):

    "filter a string through a command that transforms its input to its output"

    for name, fn in filtertable.iteritems():

        if cmd.startswith(name):

            return fn(s, cmd[len(name):].lstrip())

    return pipefilter(s, cmd)

def find_in_path(name, path, default=None):

    '''find name in search path. path can be string (will be split

    with os.pathsep), or iterable thing that returns strings.  if name

    found, return path to name. else return default.'''

    if isinstance(path, str):

        path = path.split(os.pathsep)

    for p in path:

        p_name = os.path.join(p, name)

        if os.path.exists(p_name):

            return p_name

    return default

def patch(strip, patchname, ui):

    """apply the patch <patchname> to the working directory.

    a list of patched files is returned"""

    patcher = find_in_path('gpatch', os.environ.get('PATH', ''), 'patch')

    fp = os.popen('"%s" -p%d < "%s"' % (patcher, strip, patchname))

    files = {}

    for line in fp:

        line = line.rstrip()

        ui.status("%s\n" % line)

        if line.startswith('patching file '):

            pf = parse_patch_output(line)

            files.setdefault(pf, 1)

    code = fp.close()

    if code:

        raise Abort(_("patch command failed: %s") % explain_exit(code)[0])

    return files.keys()

def binary(s):

    """return true if a string is binary data using diff's heuristic"""

    if s and '\0' in s[:4096]:

        return True

    return False

def unique(g):

    """return the uniq elements of iterable g"""

    seen = {}

    for f in g:

        if f not in seen:

            seen[f] = 1

            yield f

class Abort(Exception):

    """Raised if a command needs to print an error and exit."""

def always(fn): return True

def never(fn): return False

def patkind(name, dflt_pat='glob'):

    """Split a string into an optional pattern kind prefix and the

    actual pattern."""

    for prefix in 're', 'glob', 'path', 'relglob', 'relpath', 'relre':

        if name.startswith(prefix + ':'): return name.split(':', 1)

    return dflt_pat, name

def globre(pat, head='^', tail='$'):

    "convert a glob pattern into a regexp"

    i, n = 0, len(pat)

    res = ''

    group = False

    def peek(): return i < n and pat[i]

    while i < n:

        c = pat[i]

        i = i+1

        if c == '*':

            if peek() == '*':

                i += 1

                res += '.*'

            else:

                res += '[^/]*'

        elif c == '?':

            res += '.'

        elif c == '[':

            j = i

            if j < n and pat[j] in '!]':

                j += 1

            while j < n and pat[j] != ']':

                j += 1

            if j >= n:

                res += '\\['

            else:

                stuff = pat[i:j].replace('\\','\\\\')

                i = j + 1

                if stuff[0] == '!':

                    stuff = '^' + stuff[1:]

                elif stuff[0] == '^':

                    stuff = '\\' + stuff

                res = '%s[%s]' % (res, stuff)

        elif c == '{':

            group = True

            res += '(?:'

        elif c == '}' and group:

            res += ')'

            group = False

        elif c == ',' and group:

            res += '|'

        elif c == '\\':

            p = peek()

            if p:

                i += 1

                res += re.escape(p)

            else:

                res += re.escape(c)

        else:

            res += re.escape(c)

    return head + res + tail

_globchars = {'[': 1, '{': 1, '*': 1, '?': 1}

def pathto(n1, n2):

    '''return the relative path from one place to another.

    this returns a path in the form used by the local filesystem, not hg.'''

    if not n1: return localpath(n2)

    a, b = n1.split('/'), n2.split('/')

    a.reverse()

    b.reverse()

    while a and b and a[-1] == b[-1]:

        a.pop()

        b.pop()

    b.reverse()

    return os.sep.join((['..'] * len(a)) + b)

def canonpath(root, cwd, myname):

    """return the canonical path of myname, given cwd and root"""

    if root == os.sep:

        rootsep = os.sep

    else:

        rootsep = root + os.sep

    name = myname

    if not os.path.isabs(name):

        name = os.path.join(root, cwd, name)

    name = os.path.normpath(name)

    if name.startswith(rootsep):

        name = name[len(rootsep):]

        audit_path(name)

        return pconvert(name)

    elif name == root:

        return ''

    else:

        # Determine whether `name' is in the hierarchy at or beneath `root',

        # by iterating name=dirname(name) until that causes no change (can't

        # check name == '/', because that doesn't work on windows).  For each

        # `name', compare dev/inode numbers.  If they match, the list `rel'

        # holds the reversed list of components making up the relative file

        # name we want.

        root_st = os.stat(root)

        rel = []

        while True:

            try:

                name_st = os.stat(name)

            except OSError:

                break

            if os.path.samestat(name_st, root_st):

                rel.reverse()

                name = os.path.join(*rel)

                audit_path(name)

                return pconvert(name)

            dirname, basename = os.path.split(name)

            rel.append(basename)

            if dirname == name:

                break

            name = dirname

        raise Abort('%s not under root' % myname)

def matcher(canonroot, cwd='', names=['.'], inc=[], exc=[], head='', src=None):

    return _matcher(canonroot, cwd, names, inc, exc, head, 'glob', src)

def cmdmatcher(canonroot, cwd='', names=['.'], inc=[], exc=[], head='', src=None):

    if os.name == 'nt':

        dflt_pat = 'glob'

    else:

        dflt_pat = 'relpath'

    return _matcher(canonroot, cwd, names, inc, exc, head, dflt_pat, src)

def _matcher(canonroot, cwd, names, inc, exc, head, dflt_pat, src):

    """build a function to match a set of file patterns

    arguments:

    canonroot - the canonical root of the tree you're matching against

    cwd - the current working directory, if relevant

    names - patterns to find

    inc - patterns to include

    exc - patterns to exclude

    head - a regex to prepend to patterns to control whether a match is rooted

    a pattern is one of:

    'glob:<rooted glob>'

    're:<rooted regexp>'

    'path:<rooted path>'

    'relglob:<relative glob>'

    'relpath:<relative path>'

    'relre:<relative regexp>'

    '<rooted path or regexp>'

    returns:

    a 3-tuple containing

    - list of explicit non-pattern names passed in

    - a bool match(filename) function

    - a bool indicating if any patterns were passed in

    todo:

    make head regex a rooted bool

    """

    def contains_glob(name):

        for c in name:

            if c in _globchars: return True

        return False

    def regex(kind, name, tail):

        '''convert a pattern into a regular expression'''

        if kind == 're':

            return name

        elif kind == 'path':

            return '^' + re.escape(name) + '(?:/|$)'

        elif kind == 'relglob':

            return head + globre(name, '(?:|.*/)', tail)

        elif kind == 'relpath':

            return head + re.escape(name) + tail

        elif kind == 'relre':

            if name.startswith('^'):

                return name

            return '.*' + name

        return head + globre(name, '', tail)

    def matchfn(pats, tail):

        """build a matching function from a set of patterns"""

        if not pats:

            return

        matches = []

        for k, p in pats:

            try:

                pat = '(?:%s)' % regex(k, p, tail)

                matches.append(re.compile(pat).match)

            except re.error:

                if src: raise Abort("%s: invalid pattern (%s): %s" % (src, k, p))

                else: raise Abort("invalid pattern (%s): %s" % (k, p))

        def buildfn(text):

            for m in matches:

                r = m(text)

                if r:

                    return r

        return buildfn

    def globprefix(pat):

        '''return the non-glob prefix of a path, e.g. foo/* -> foo'''

        root = []

        for p in pat.split(os.sep):

            if contains_glob(p): break

            root.append(p)

        return '/'.join(root)

    pats = []

    files = []

    roots = []

    for kind, name in [patkind(p, dflt_pat) for p in names]:

        if kind in ('glob', 'relpath'):

            name = canonpath(canonroot, cwd, name)

            if name == '':

                kind, name = 'glob', '**'

        if kind in ('glob', 'path', 're'):

            pats.append((kind, name))

        if kind == 'glob':

            root = globprefix(name)

            if root: roots.append(root)

        elif kind == 'relpath':

            files.append((kind, name))

            roots.append(name)

    patmatch = matchfn(pats, '$') or always

    filematch = matchfn(files, '(?:/|$)') or always

    incmatch = always

    if inc:

        incmatch = matchfn(map(patkind, inc), '(?:/|$)')

    excmatch = lambda fn: False

    if exc:

        excmatch = matchfn(map(patkind, exc), '(?:/|$)')

    return (roots,

            lambda fn: (incmatch(fn) and not excmatch(fn) and

                        (fn.endswith('/') or

                         (not pats and not files) or

                         (pats and patmatch(fn)) or

                         (files and filematch(fn)))),

            (inc or exc or (pats and pats != [('glob', '**')])) and True)