/mercurial/util.py

Large files files are truncated, but you can click here to view the full file

# util.py - Mercurial utility functions and platform specific implementations
#
#  Copyright 2005 K. Thananchayan <thananck@yahoo.com>
#  Copyright 2005-2007 Matt Mackall <mpm@selenic.com>
#  Copyright 2006 Vadim Gelfer <vadim.gelfer@gmail.com>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.

"""Mercurial utility functions and platform specific implementations.

This contains helper routines that are independent of the SCM core and
hide platform-specific details from the core.
"""

from __future__ import absolute_import

import bz2
import calendar
import collections
import datetime
import errno
import gc
import hashlib
import imp
import os
import platform as pyplatform
import re as remod
import shutil
import signal
import socket
import stat
import string
import subprocess
import sys
import tempfile
import textwrap
import time
import traceback
import zlib

from . import (
    encoding,
    error,
    i18n,
    osutil,
    parsers,
    pycompat,
)

empty = pycompat.empty
httplib = pycompat.httplib
httpserver = pycompat.httpserver
pickle = pycompat.pickle
queue = pycompat.queue
socketserver = pycompat.socketserver
stderr = pycompat.stderr
stdin = pycompat.stdin
stdout = pycompat.stdout
stringio = pycompat.stringio
urlerr = pycompat.urlerr
urlparse = pycompat.urlparse
urlreq = pycompat.urlreq
xmlrpclib = pycompat.xmlrpclib

def isatty(fp):
    try:
        return fp.isatty()
    except AttributeError:
        return False

# glibc determines buffering on first write to stdout - if we replace a TTY
# destined stdout with a pipe destined stdout (e.g. pager), we want line
# buffering
if isatty(stdout):
    stdout = os.fdopen(stdout.fileno(), pycompat.sysstr('wb'), 1)

if pycompat.osname == 'nt':
    from . import windows as platform
    stdout = platform.winstdout(stdout)
else:
    from . import posix as platform

_ = i18n._

bindunixsocket = platform.bindunixsocket
cachestat = platform.cachestat
checkexec = platform.checkexec
checklink = platform.checklink
copymode = platform.copymode
executablepath = platform.executablepath
expandglobs = platform.expandglobs
explainexit = platform.explainexit
findexe = platform.findexe
gethgcmd = platform.gethgcmd
getuser = platform.getuser
getpid = os.getpid
groupmembers = platform.groupmembers
groupname = platform.groupname
hidewindow = platform.hidewindow
isexec = platform.isexec
isowner = platform.isowner
localpath = platform.localpath
lookupreg = platform.lookupreg
makedir = platform.makedir
nlinks = platform.nlinks
normpath = platform.normpath
normcase = platform.normcase
normcasespec = platform.normcasespec
normcasefallback = platform.normcasefallback
openhardlinks = platform.openhardlinks
oslink = platform.oslink
parsepatchoutput = platform.parsepatchoutput
pconvert = platform.pconvert
poll = platform.poll
popen = platform.popen
posixfile = platform.posixfile
quotecommand = platform.quotecommand
readpipe = platform.readpipe
rename = platform.rename
removedirs = platform.removedirs
samedevice = platform.samedevice
samefile = platform.samefile
samestat = platform.samestat
setbinary = platform.setbinary
setflags = platform.setflags
setsignalhandler = platform.setsignalhandler
shellquote = platform.shellquote
spawndetached = platform.spawndetached
split = platform.split
sshargs = platform.sshargs
statfiles = getattr(osutil, 'statfiles', platform.statfiles)
statisexec = platform.statisexec
statislink = platform.statislink
testpid = platform.testpid
umask = platform.umask
unlink = platform.unlink
unlinkpath = platform.unlinkpath
username = platform.username

# Python compatibility

_notset = object()

# disable Python's problematic floating point timestamps (issue4836)
# (Python hypocritically says you shouldn't change this behavior in
# libraries, and sure enough Mercurial is not a library.)
os.stat_float_times(False)

def safehasattr(thing, attr):
    return getattr(thing, attr, _notset) is not _notset

def bitsfrom(container):
    bits = 0
    for bit in container:
        bits |= bit
    return bits

DIGESTS = {
    'md5': hashlib.md5,
    'sha1': hashlib.sha1,
    'sha512': hashlib.sha512,
}
# List of digest types from strongest to weakest
DIGESTS_BY_STRENGTH = ['sha512', 'sha1', 'md5']

for k in DIGESTS_BY_STRENGTH:
    assert k in DIGESTS

class digester(object):
    """helper to compute digests.

    This helper can be used to compute one or more digests given their name.

    >>> d = digester(['md5', 'sha1'])
    >>> d.update('foo')
    >>> [k for k in sorted(d)]
    ['md5', 'sha1']
    >>> d['md5']
    'acbd18db4cc2f85cedef654fccc4a4d8'
    >>> d['sha1']
    '0beec7b5ea3f0fdbc95d0dd47f3c5bc275da8a33'
    >>> digester.preferred(['md5', 'sha1'])
    'sha1'
    """

    def __init__(self, digests, s=''):
        self._hashes = {}
        for k in digests:
            if k not in DIGESTS:
                raise Abort(_('unknown digest type: %s') % k)
            self._hashes[k] = DIGESTS[k]()
        if s:
            self.update(s)

    def update(self, data):
        for h in self._hashes.values():
            h.update(data)

    def __getitem__(self, key):
        if key not in DIGESTS:
            raise Abort(_('unknown digest type: %s') % k)
        return self._hashes[key].hexdigest()

    def __iter__(self):
        return iter(self._hashes)

    @staticmethod
    def preferred(supported):
        """returns the strongest digest type in both supported and DIGESTS."""

        for k in DIGESTS_BY_STRENGTH:
            if k in supported:
                return k
        return None

class digestchecker(object):
    """file handle wrapper that additionally checks content against a given
    size and digests.

        d = digestchecker(fh, size, {'md5': '...'})

    When multiple digests are given, all of them are validated.
    """

    def __init__(self, fh, size, digests):
        self._fh = fh
        self._size = size
        self._got = 0
        self._digests = dict(digests)
        self._digester = digester(self._digests.keys())

    def read(self, length=-1):
        content = self._fh.read(length)
        self._digester.update(content)
        self._got += len(content)
        return content

    def validate(self):
        if self._size != self._got:
            raise Abort(_('size mismatch: expected %d, got %d') %
                (self._size, self._got))
        for k, v in self._digests.items():
            if v != self._digester[k]:
                # i18n: first parameter is a digest name
                raise Abort(_('%s mismatch: expected %s, got %s') %
                    (k, v, self._digester[k]))

try:
    buffer = buffer
except NameError:
    if not pycompat.ispy3:
        def buffer(sliceable, offset=0, length=None):
            if length is not None:
                return sliceable[offset:offset + length]
            return sliceable[offset:]
    else:
        def buffer(sliceable, offset=0, length=None):
            if length is not None:
                return memoryview(sliceable)[offset:offset + length]
            return memoryview(sliceable)[offset:]

closefds = pycompat.osname == 'posix'

_chunksize = 4096

class bufferedinputpipe(object):
    """a manually buffered input pipe

    Python will not let us use buffered IO and lazy reading with 'polling' at
    the same time. We cannot probe the buffer state and select will not detect
    that data are ready to read if they are already buffered.

    This class let us work around that by implementing its own buffering
    (allowing efficient readline) while offering a way to know if the buffer is
    empty from the output (allowing collaboration of the buffer with polling).

    This class lives in the 'util' module because it makes use of the 'os'
    module from the python stdlib.
    """

    def __init__(self, input):
        self._input = input
        self._buffer = []
        self._eof = False
        self._lenbuf = 0

    @property
    def hasbuffer(self):
        """True is any data is currently buffered

        This will be used externally a pre-step for polling IO. If there is
        already data then no polling should be set in place."""
        return bool(self._buffer)

    @property
    def closed(self):
        return self._input.closed

    def fileno(self):
        return self._input.fileno()

    def close(self):
        return self._input.close()

    def read(self, size):
        while (not self._eof) and (self._lenbuf < size):
            self._fillbuffer()
        return self._frombuffer(size)

    def readline(self, *args, **kwargs):
        if 1 < len(self._buffer):
            # this should not happen because both read and readline end with a
            # _frombuffer call that collapse it.
            self._buffer = [''.join(self._buffer)]
            self._lenbuf = len(self._buffer[0])
        lfi = -1
        if self._buffer:
            lfi = self._buffer[-1].find('\n')
        while (not self._eof) and lfi < 0:
            self._fillbuffer()
            if self._buffer:
                lfi = self._buffer[-1].find('\n')
        size = lfi + 1
        if lfi < 0: # end of file
            size = self._lenbuf
        elif 1 < len(self._buffer):
            # we need to take previous chunks into account
            size += self._lenbuf - len(self._buffer[-1])
        return self._frombuffer(size)

    def _frombuffer(self, size):
        """return at most 'size' data from the buffer

        The data are removed from the buffer."""
        if size == 0 or not self._buffer:
            return ''
        buf = self._buffer[0]
        if 1 < len(self._buffer):
            buf = ''.join(self._buffer)

        data = buf[:size]
        buf = buf[len(data):]
        if buf:
            self._buffer = [buf]
            self._lenbuf = len(buf)
        else:
            self._buffer = []
            self._lenbuf = 0
        return data

    def _fillbuffer(self):
        """read data to the buffer"""
        data = os.read(self._input.fileno(), _chunksize)
        if not data:
            self._eof = True
        else:
            self._lenbuf += len(data)
            self._buffer.append(data)

def popen2(cmd, env=None, newlines=False):
    # Setting bufsize to -1 lets the system decide the buffer size.
    # The default for bufsize is 0, meaning unbuffered. This leads to
    # poor performance on Mac OS X: http://bugs.python.org/issue4194
    p = subprocess.Popen(cmd, shell=True, bufsize=-1,
                         close_fds=closefds,
                         stdin=subprocess.PIPE, stdout=subprocess.PIPE,
                         universal_newlines=newlines,
                         env=env)
    return p.stdin, p.stdout

def popen3(cmd, env=None, newlines=False):
    stdin, stdout, stderr, p = popen4(cmd, env, newlines)
    return stdin, stdout, stderr

def popen4(cmd, env=None, newlines=False, bufsize=-1):
    p = subprocess.Popen(cmd, shell=True, bufsize=bufsize,
                         close_fds=closefds,
                         stdin=subprocess.PIPE, stdout=subprocess.PIPE,
                         stderr=subprocess.PIPE,
                         universal_newlines=newlines,
                         env=env)
    return p.stdin, p.stdout, p.stderr, p

def version():
    """Return version information if available."""
    try:
        from . import __version__
        return __version__.version
    except ImportError:
        return 'unknown'

def versiontuple(v=None, n=4):
    """Parses a Mercurial version string into an N-tuple.

    The version string to be parsed is specified with the ``v`` argument.
    If it isn't defined, the current Mercurial version string will be parsed.

    ``n`` can be 2, 3, or 4. Here is how some version strings map to
    returned values:

    >>> v = '3.6.1+190-df9b73d2d444'
    >>> versiontuple(v, 2)
    (3, 6)
    >>> versiontuple(v, 3)
    (3, 6, 1)
    >>> versiontuple(v, 4)
    (3, 6, 1, '190-df9b73d2d444')

    >>> versiontuple('3.6.1+190-df9b73d2d444+20151118')
    (3, 6, 1, '190-df9b73d2d444+20151118')

    >>> v = '3.6'
    >>> versiontuple(v, 2)
    (3, 6)
    >>> versiontuple(v, 3)
    (3, 6, None)
    >>> versiontuple(v, 4)
    (3, 6, None, None)

    >>> v = '3.9-rc'
    >>> versiontuple(v, 2)
    (3, 9)
    >>> versiontuple(v, 3)
    (3, 9, None)
    >>> versiontuple(v, 4)
    (3, 9, None, 'rc')

    >>> v = '3.9-rc+2-02a8fea4289b'
    >>> versiontuple(v, 2)
    (3, 9)
    >>> versiontuple(v, 3)
    (3, 9, None)
    >>> versiontuple(v, 4)
    (3, 9, None, 'rc+2-02a8fea4289b')
    """
    if not v:
        v = version()
    parts = remod.split('[\+-]', v, 1)
    if len(parts) == 1:
        vparts, extra = parts[0], None
    else:
        vparts, extra = parts

    vints = []
    for i in vparts.split('.'):
        try:
            vints.append(int(i))
        except ValueError:
            break
    # (3, 6) -> (3, 6, None)
    while len(vints) < 3:
        vints.append(None)

    if n == 2:
        return (vints[0], vints[1])
    if n == 3:
        return (vints[0], vints[1], vints[2])
    if n == 4:
        return (vints[0], vints[1], vints[2], extra)

# used by parsedate
defaultdateformats = (
    '%Y-%m-%dT%H:%M:%S', # the 'real' ISO8601
    '%Y-%m-%dT%H:%M',    #   without seconds
    '%Y-%m-%dT%H%M%S',   # another awful but legal variant without :
    '%Y-%m-%dT%H%M',     #   without seconds
    '%Y-%m-%d %H:%M:%S', # our common legal variant
    '%Y-%m-%d %H:%M',    #   without seconds
    '%Y-%m-%d %H%M%S',   # without :
    '%Y-%m-%d %H%M',     #   without seconds
    '%Y-%m-%d %I:%M:%S%p',
    '%Y-%m-%d %H:%M',
    '%Y-%m-%d %I:%M%p',
    '%Y-%m-%d',
    '%m-%d',
    '%m/%d',
    '%m/%d/%y',
    '%m/%d/%Y',
    '%a %b %d %H:%M:%S %Y',
    '%a %b %d %I:%M:%S%p %Y',
    '%a, %d %b %Y %H:%M:%S',        #  GNU coreutils "/bin/date --rfc-2822"
    '%b %d %H:%M:%S %Y',
    '%b %d %I:%M:%S%p %Y',
    '%b %d %H:%M:%S',
    '%b %d %I:%M:%S%p',
    '%b %d %H:%M',
    '%b %d %I:%M%p',
    '%b %d %Y',
    '%b %d',
    '%H:%M:%S',
    '%I:%M:%S%p',
    '%H:%M',
    '%I:%M%p',
)

extendeddateformats = defaultdateformats + (
    "%Y",
    "%Y-%m",
    "%b",
    "%b %Y",
    )

def cachefunc(func):
    '''cache the result of function calls'''
    # XXX doesn't handle keywords args
    if func.__code__.co_argcount == 0:
        cache = []
        def f():
            if len(cache) == 0:
                cache.append(func())
            return cache[0]
        return f
    cache = {}
    if func.__code__.co_argcount == 1:
        # we gain a small amount of time because
        # we don't need to pack/unpack the list
        def f(arg):
            if arg not in cache:
                cache[arg] = func(arg)
            return cache[arg]
    else:
        def f(*args):
            if args not in cache:
                cache[args] = func(*args)
            return cache[args]

    return f

class sortdict(dict):
    '''a simple sorted dictionary'''
    def __init__(self, data=None):
        self._list = []
        if data:
            self.update(data)
    def copy(self):
        return sortdict(self)
    def __setitem__(self, key, val):
        if key in self:
            self._list.remove(key)
        self._list.append(key)
        dict.__setitem__(self, key, val)
    def __iter__(self):
        return self._list.__iter__()
    def update(self, src):
        if isinstance(src, dict):
            src = src.iteritems()
        for k, v in src:
            self[k] = v
    def clear(self):
        dict.clear(self)
        self._list = []
    def items(self):
        return [(k, self[k]) for k in self._list]
    def __delitem__(self, key):
        dict.__delitem__(self, key)
        self._list.remove(key)
    def pop(self, key, *args, **kwargs):
        dict.pop(self, key, *args, **kwargs)
        try:
            self._list.remove(key)
        except ValueError:
            pass
    def keys(self):
        return self._list[:]
    def iterkeys(self):
        return self._list.__iter__()
    def iteritems(self):
        for k in self._list:
            yield k, self[k]
    def insert(self, index, key, val):
        self._list.insert(index, key)
        dict.__setitem__(self, key, val)
    def __repr__(self):
        if not self:
            return '%s()' % self.__class__.__name__
        return '%s(%r)' % (self.__class__.__name__, self.items())

class _lrucachenode(object):
    """A node in a doubly linked list.

    Holds a reference to nodes on either side as well as a key-value
    pair for the dictionary entry.
    """
    __slots__ = (u'next', u'prev', u'key', u'value')

    def __init__(self):
        self.next = None
        self.prev = None

        self.key = _notset
        self.value = None

    def markempty(self):
        """Mark the node as emptied."""
        self.key = _notset

class lrucachedict(object):
    """Dict that caches most recent accesses and sets.

    The dict consists of an actual backing dict - indexed by original
    key - and a doubly linked circular list defining the order of entries in
    the cache.

    The head node is the newest entry in the cache. If the cache is full,
    we recycle head.prev and make it the new head. Cache accesses result in
    the node being moved to before the existing head and being marked as the
    new head node.
    """
    def __init__(self, max):
        self._cache = {}

        self._head = head = _lrucachenode()
        head.prev = head
        head.next = head
        self._size = 1
        self._capacity = max

    def __len__(self):
        return len(self._cache)

    def __contains__(self, k):
        return k in self._cache

    def __iter__(self):
        # We don't have to iterate in cache order, but why not.
        n = self._head
        for i in range(len(self._cache)):
            yield n.key
            n = n.next

    def __getitem__(self, k):
        node = self._cache[k]
        self._movetohead(node)
        return node.value

    def __setitem__(self, k, v):
        node = self._cache.get(k)
        # Replace existing value and mark as newest.
        if node is not None:
            node.value = v
            self._movetohead(node)
            return

        if self._size < self._capacity:
            node = self._addcapacity()
        else:
            # Grab the last/oldest item.
            node = self._head.prev

        # At capacity. Kill the old entry.
        if node.key is not _notset:
            del self._cache[node.key]

        node.key = k
        node.value = v
        self._cache[k] = node
        # And mark it as newest entry. No need to adjust order since it
        # is already self._head.prev.
        self._head = node

    def __delitem__(self, k):
        node = self._cache.pop(k)
        node.markempty()

        # Temporarily mark as newest item before re-adjusting head to make
        # this node the oldest item.
        self._movetohead(node)
        self._head = node.next

    # Additional dict methods.

    def get(self, k, default=None):
        try:
            return self._cache[k].value
        except KeyError:
            return default

    def clear(self):
        n = self._head
        while n.key is not _notset:
            n.markempty()
            n = n.next

        self._cache.clear()

    def copy(self):
        result = lrucachedict(self._capacity)
        n = self._head.prev
        # Iterate in oldest-to-newest order, so the copy has the right ordering
        for i in range(len(self._cache)):
            result[n.key] = n.value
            n = n.prev
        return result

    def _movetohead(self, node):
        """Mark a node as the newest, making it the new head.

        When a node is accessed, it becomes the freshest entry in the LRU
        list, which is denoted by self._head.

        Visually, let's make ``N`` the new head node (* denotes head):

            previous/oldest <-> head <-> next/next newest

            ----<->--- A* ---<->-----
            |                       |
            E <-> D <-> N <-> C <-> B

        To:

            ----<->--- N* ---<->-----
            |                       |
            E <-> D <-> C <-> B <-> A

        This requires the following moves:

           C.next = D  (node.prev.next = node.next)
           D.prev = C  (node.next.prev = node.prev)
           E.next = N  (head.prev.next = node)
           N.prev = E  (node.prev = head.prev)
           N.next = A  (node.next = head)
           A.prev = N  (head.prev = node)
        """
        head = self._head
        # C.next = D
        node.prev.next = node.next
        # D.prev = C
        node.next.prev = node.prev
        # N.prev = E
        node.prev = head.prev
        # N.next = A
        # It is tempting to do just "head" here, however if node is
        # adjacent to head, this will do bad things.
        node.next = head.prev.next
        # E.next = N
        node.next.prev = node
        # A.prev = N
        node.prev.next = node

        self._head = node

    def _addcapacity(self):
        """Add a node to the circular linked list.

        The new node is inserted before the head node.
        """
        head = self._head
        node = _lrucachenode()
        head.prev.next = node
        node.prev = head.prev
        node.next = head
        head.prev = node
        self._size += 1
        return node

def lrucachefunc(func):
    '''cache most recent results of function calls'''
    cache = {}
    order = collections.deque()
    if func.__code__.co_argcount == 1:
        def f(arg):
            if arg not in cache:
                if len(cache) > 20:
                    del cache[order.popleft()]
                cache[arg] = func(arg)
            else:
                order.remove(arg)
            order.append(arg)
            return cache[arg]
    else:
        def f(*args):
            if args not in cache:
                if len(cache) > 20:
                    del cache[order.popleft()]
                cache[args] = func(*args)
            else:
                order.remove(args)
            order.append(args)
            return cache[args]

    return f

class propertycache(object):
    def __init__(self, func):
        self.func = func
        self.name = func.__name__
    def __get__(self, obj, type=None):
        result = self.func(obj)
        self.cachevalue(obj, result)
        return result

    def cachevalue(self, obj, value):
        # __dict__ assignment required to bypass __setattr__ (eg: repoview)
        obj.__dict__[self.name] = value

def pipefilter(s, cmd):
    '''filter string S through command CMD, returning its output'''
    p = subprocess.Popen(cmd, shell=True, close_fds=closefds,
                         stdin=subprocess.PIPE, stdout=subprocess.PIPE)
    pout, perr = p.communicate(s)
    return pout

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, pycompat.sysstr('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 pycompat.sysplatform == 'OpenVMS' and code & 1:
            code = 0
        if code:
            raise Abort(_("command '%s' failed: %s") %
                        (cmd, explainexit(code)))
        return readfile(outname)
    finally:
        try:
            if inname:
                os.unlink(inname)
        except OSError:
            pass
        try:
            if outname:
                os.unlink(outname)
        except OSError:
            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 binary(s):
    """return true if a string is binary data"""
    return bool(s and '\0' in s)

def increasingchunks(source, min=1024, max=65536):
    '''return no less than min bytes per chunk while data remains,
    doubling min after each chunk until it reaches max'''
    def log2(x):
        if not x:
            return 0
        i = 0
        while x:
            x >>= 1
            i += 1
        return i - 1

    buf = []
    blen = 0
    for chunk in source:
        buf.append(chunk)
        blen += len(chunk)
        if blen >= min:
            if min < max:
                min = min << 1
                nmin = 1 << log2(blen)
                if nmin > min:
                    min = nmin
                if min > max:
                    min = max
            yield ''.join(buf)
            blen = 0
            buf = []
    if buf:
        yield ''.join(buf)

Abort = error.Abort

def always(fn):
    return True

def never(fn):
    return False

def nogc(func):
    """disable garbage collector

    Python's garbage collector triggers a GC each time a certain number of
    container objects (the number being defined by gc.get_threshold()) are
    allocated even when marked not to be tracked by the collector. Tracking has
    no effect on when GCs are triggered, only on what objects the GC looks
    into. As a workaround, disable GC while building complex (huge)
    containers.

    This garbage collector issue have been fixed in 2.7.
    """
    if sys.version_info >= (2, 7):
        return func
    def wrapper(*args, **kwargs):
        gcenabled = gc.isenabled()
        gc.disable()
        try:
            return func(*args, **kwargs)
        finally:
            if gcenabled:
                gc.enable()
    return wrapper

def pathto(root, n1, n2):
    '''return the relative path from one place to another.
    root should use os.sep to separate directories
    n1 should use os.sep to separate directories
    n2 should use "/" to separate directories
    returns an os.sep-separated path.

    If n1 is a relative path, it's assumed it's
    relative to root.
    n2 should always be relative to root.
    '''
    if not n1:
        return localpath(n2)
    if os.path.isabs(n1):
        if os.path.splitdrive(root)[0] != os.path.splitdrive(n1)[0]:
            return os.path.join(root, localpath(n2))
        n2 = '/'.join((pconvert(root), n2))
    a, b = splitpath(n1), n2.split('/')
    a.reverse()
    b.reverse()
    while a and b and a[-1] == b[-1]:
        a.pop()
        b.pop()
    b.reverse()
    return pycompat.ossep.join((['..'] * len(a)) + b) or '.'

def mainfrozen():
    """return True if we are a frozen executable.

    The code supports py2exe (most common, Windows only) and tools/freeze
    (portable, not much used).
    """
    return (safehasattr(sys, "frozen") or # new py2exe
            safehasattr(sys, "importers") or # old py2exe
            imp.is_frozen(u"__main__")) # tools/freeze

# the location of data files matching the source code
if mainfrozen() and getattr(sys, 'frozen', None) != 'macosx_app':
    # executable version (py2exe) doesn't support __file__
    datapath = os.path.dirname(pycompat.sysexecutable)
else:
    datapath = os.path.dirname(__file__)

if not isinstance(datapath, bytes):
    datapath = pycompat.fsencode(datapath)

i18n.setdatapath(datapath)

_hgexecutable = None

def hgexecutable():
    """return location of the 'hg' executable.

    Defaults to $HG or 'hg' in the search path.
    """
    if _hgexecutable is None:
        hg = encoding.environ.get('HG')
        mainmod = sys.modules['__main__']
        if hg:
            _sethgexecutable(hg)
        elif mainfrozen():
            if getattr(sys, 'frozen', None) == 'macosx_app':
                # Env variable set by py2app
                _sethgexecutable(encoding.environ['EXECUTABLEPATH'])
            else:
                _sethgexecutable(pycompat.sysexecutable)
        elif os.path.basename(getattr(mainmod, '__file__', '')) == 'hg':
            _sethgexecutable(mainmod.__file__)
        else:
            exe = findexe('hg') or os.path.basename(sys.argv[0])
            _sethgexecutable(exe)
    return _hgexecutable

def _sethgexecutable(path):
    """set location of the 'hg' executable"""
    global _hgexecutable
    _hgexecutable = path

def _isstdout(f):
    fileno = getattr(f, 'fileno', None)
    return fileno and fileno() == sys.__stdout__.fileno()

def shellenviron(environ=None):
    """return environ with optional override, useful for shelling out"""
    def py2shell(val):
        'convert python object into string that is useful to shell'
        if val is None or val is False:
            return '0'
        if val is True:
            return '1'
        return str(val)
    env = dict(encoding.environ)
    if environ:
        env.update((k, py2shell(v)) for k, v in environ.iteritems())
    env['HG'] = hgexecutable()
    return env

def system(cmd, environ=None, cwd=None, onerr=None, errprefix=None, out=None):
    '''enhanced shell command execution.
    run with environment maybe modified, maybe in different dir.

    if command fails and onerr is None, return status, else raise onerr
    object as exception.

    if out is specified, it is assumed to be a file-like object that has a
    write() method. stdout and stderr will be redirected to out.'''
    try:
        stdout.flush()
    except Exception:
        pass
    origcmd = cmd
    cmd = quotecommand(cmd)
    if pycompat.sysplatform == 'plan9' and (sys.version_info[0] == 2
                                    and sys.version_info[1] < 7):
        # subprocess kludge to work around issues in half-baked Python
        # ports, notably bichued/python:
        if not cwd is None:
            os.chdir(cwd)
        rc = os.system(cmd)
    else:
        env = shellenviron(environ)
        if out is None or _isstdout(out):
            rc = subprocess.call(cmd, shell=True, close_fds=closefds,
                                 env=env, cwd=cwd)
        else:
            proc = subprocess.Popen(cmd, shell=True, close_fds=closefds,
                                    env=env, cwd=cwd, stdout=subprocess.PIPE,
                                    stderr=subprocess.STDOUT)
            for line in iter(proc.stdout.readline, ''):
                out.write(line)
            proc.wait()
            rc = proc.returncode
        if pycompat.sysplatform == 'OpenVMS' and rc & 1:
            rc = 0
    if rc and onerr:
        errmsg = '%s %s' % (os.path.basename(origcmd.split(None, 1)[0]),
                            explainexit(rc)[0])
        if errprefix:
            errmsg = '%s: %s' % (errprefix, errmsg)
        raise onerr(errmsg)
    return rc

def checksignature(func):
    '''wrap a function with code to check for calling errors'''
    def check(*args, **kwargs):
        try:
            return func(*args, **kwargs)
        except TypeError:
            if len(traceback.extract_tb(sys.exc_info()[2])) == 1:
                raise error.SignatureError
            raise

    return check

def copyfile(src, dest, hardlink=False, copystat=False, checkambig=False):
    '''copy a file, preserving mode and optionally other stat info like
    atime/mtime

    checkambig argument is used with filestat, and is useful only if
    destination file is guarded by any lock (e.g. repo.lock or
    repo.wlock).

    copystat and checkambig should be exclusive.
    '''
    assert not (copystat and checkambig)
    oldstat = None
    if os.path.lexists(dest):
        if checkambig:
            oldstat = checkambig and filestat(dest)
        unlink(dest)
    # hardlinks are problematic on CIFS, quietly ignore this flag
    # until we find a way to work around it cleanly (issue4546)
    if False and hardlink:
        try:
            oslink(src, dest)
            return
        except (IOError, OSError):
            pass # fall back to normal copy
    if os.path.islink(src):
        os.symlink(os.readlink(src), dest)
        # copytime is ignored for symlinks, but in general copytime isn't needed
        # for them anyway
    else:
        try:
            shutil.copyfile(src, dest)
            if copystat:
                # copystat also copies mode
                shutil.copystat(src, dest)
            else:
                shutil.copymode(src, dest)
                if oldstat and oldstat.stat:
                    newstat = filestat(dest)
                    if newstat.isambig(oldstat):
                        # stat of copied file is ambiguous to original one
                        advanced = (oldstat.stat.st_mtime + 1) & 0x7fffffff
                        os.utime(dest, (advanced, advanced))
        except shutil.Error as inst:
            raise Abort(str(inst))

def copyfiles(src, dst, hardlink=None, progress=lambda t, pos: None):
    """Copy a directory tree using hardlinks if possible."""
    num = 0

    if hardlink is None:
        hardlink = (os.stat(src).st_dev ==
                    os.stat(os.path.dirname(dst)).st_dev)
    if hardlink:
        topic = _('linking')
    else:
        topic = _('copying')

    if os.path.isdir(src):
        os.mkdir(dst)
        for name, kind in osutil.listdir(src):
            srcname = os.path.join(src, name)
            dstname = os.path.join(dst, name)
            def nprog(t, pos):
                if pos is not None:
                    return progress(t, pos + num)
            hardlink, n = copyfiles(srcname, dstname, hardlink, progress=nprog)
            num += n
    else:
        if hardlink:
            try:
                oslink(src, dst)
            except (IOError, OSError):
                hardlink = False
                shutil.copy(src, dst)
        else:
            shutil.copy(src, dst)
        num += 1
        progress(topic, num)
    progress(topic, None)

    return hardlink, num

_winreservednames = '''con prn aux nul
    com1 com2 com3 com4 com5 com6 com7 com8 com9
    lpt1 lpt2 lpt3 lpt4 lpt5 lpt6 lpt7 lpt8 lpt9'''.split()
_winreservedchars = ':*?"<>|'
def checkwinfilename(path):
    r'''Check that the base-relative path is a valid filename on Windows.
    Returns None if the path is ok, or a UI string describing the problem.

    >>> checkwinfilename("just/a/normal/path")
    >>> checkwinfilename("foo/bar/con.xml")
    "filename contains 'con', which is reserved on Windows"
    >>> checkwinfilename("foo/con.xml/bar")
    "filename contains 'con', which is reserved on Windows"
    >>> checkwinfilename("foo/bar/xml.con")
    >>> checkwinfilename("foo/bar/AUX/bla.txt")
    "filename contains 'AUX', which is reserved on Windows"
    >>> checkwinfilename("foo/bar/bla:.txt")
    "filename contains ':', which is reserved on Windows"
    >>> checkwinfilename("foo/bar/b\07la.txt")
    "filename contains '\\x07', which is invalid on Windows"
    >>> checkwinfilename("foo/bar/bla ")
    "filename ends with ' ', which is not allowed on Windows"
    >>> checkwinfilename("../bar")
    >>> checkwinfilename("foo\\")
    "filename ends with '\\', which is invalid on Windows"
    >>> checkwinfilename("foo\\/bar")
    "directory name ends with '\\', which is invalid on Windows"
    '''
    if path.endswith('\\'):
        return _("filename ends with '\\', which is invalid on Windows")
    if '\\/' in path:
        return _("directory name ends with '\\', which is invalid on Windows")
    for n in path.replace('\\', '/').split('/'):
        if not n:
            continue
        for c in n:
            if c in _winreservedchars:
                return _("filename contains '%s', which is reserved "
                         "on Windows") % c
            if ord(c) <= 31:
                return _("filename contains %r, which is invalid "
                         "on Windows") % c
        base = n.split('.')[0]
        if base and base.lower() in _winreservednames:
            return _("filename contains '%s', which is reserved "
                     "on Windows") % base
        t = n[-1]
        if t in '. ' and n not in '..':
            return _("filename ends with '%s', which is not allowed "
                     "on Windows") % t

if pycompat.osname == 'nt':
    checkosfilename = checkwinfilename
    timer = time.clock
else:
    checkosfilename = platform.checkosfilename
    timer = time.time

if safehasattr(time, "perf_counter"):
    timer = time.perf_counter

def makelock(info, pathname):
    try:
        return os.symlink(info, pathname)
    except OSError as why:
        if why.errno == errno.EEXIST:
            raise
    except AttributeError: # no symlink in os
        pass

    ld = os.open(pathname, os.O_CREAT | os.O_WRONLY | os.O_EXCL)
    os.write(ld, info)
    os.close(ld)

def readlock(pathname):
    try:
        return os.readlink(pathname)
    except OSError as why:
        if why.errno not in (errno.EINVAL, errno.ENOSYS):
            raise
    except AttributeError: # no symlink in os
        pass
    fp = posixfile(pathname)
    r = fp.read()
    fp.close()
    return r

def fstat(fp):
    '''stat file object that may not have fileno method.'''
    try:
        return os.fstat(fp.fileno())
    except AttributeError:
        return os.stat(fp.name)

# File system features

def fscasesensitive(path):
    """
    Return true if the given path is on a case-sensitive filesystem

    Requires a path (like /foo/.hg) ending with a foldable final
    directory component.
    """
    s1 = os.lstat(path)
    d, b = os.path.split(path)
    b2 = b.upper()
    if b == b2:
        b2 = b.lower()
        if b == b2:
            return True # no evidence against case sensitivity
    p2 = os.path.join(d, b2)
    try:
        s2 = os.lstat(p2)
        if s2 == s1:
            return False
        return True
    except OSError:
        return True

try:
    import re2
    _re2 = None
except ImportError:
    _re2 = False

class _re(object):
    def _checkre2(self):
        global _re2
        try:
            # check if match works, see issue3964
            _re2 = bool(re2.match(r'\[([^\[]+)\]', '[ui]'))
        except ImportError:
            _re2 = False

    def compile(self, pat, flags=0):
        '''Compile a regular expression, using re2 if possible

        For best performance, use only re2-compatible regexp features. The
        only flags from the re module that are re2-compatible are
        IGNORECASE and MULTILINE.'''
        if _re2 is None:
            self._checkre2()
        if _re2 and (flags & ~(remod.IGNORECASE | remod.MULTILINE)) == 0:
            if flags & remod.IGNORECASE:
                pat = '(?i)' + pat
            if flags & remod.MULTILINE:
                pat = '(?m)' + pat
            try:
                return re2.compile(pat)
            except re2.error:
                pass
        return remod.compile(pat, flags)

    @propertycache
    def escape(self):
        '''Return the version of escape corresponding to self.compile.

        This is imperfect because whether re2 or re is used for a particular
        function depends on the flags, etc, but it's the best we can do.
        '''
        global _re2
        if _re2 is None:
            self._checkre2()
        if _re2:
            return re2.escape
        else:
            return remod.escape

re = _re()

_fspathcache = {}
def fspath(name, root):
    '''Get name in the case stored in the filesystem

    The name should be relative to root, and be normcase-ed for efficiency.

    Note that this function is unnecessary, and should not be
    called, for case-sensitive filesystems (simply because it's expensive).

    The root should be normcase-ed, too.
    '''
    def _makefspathcacheentry(dir):
        return dict((normcase(n), n) for n in os.listdir(dir))

    seps = pycompat.ossep
    if pycompat.osaltsep:
        seps = seps + pycompat.osaltsep
    # Protect backslashes. This gets silly very quickly.
    seps.replace('\\','\\\\')
    pattern = remod.compile(r'([^%s]+)|([%s]+)' % (seps, seps))
    dir = os.path.normpath(root)
    result = []
    for part, sep in pattern.findall(name):
        if sep:
            result.append(sep)
            continue

        if dir not in _fspathcache:
            _fspathcache[dir] = _makefspathcacheentry(dir)
        contents = _fspathcache[dir]

        found = contents.get(part)
        if not found:
            # retry "once per directory" per "dirstate.walk" which
            # may take place for each patches of "hg qpush", for example
            _fspathcache[dir] = contents = _makefspathcacheentry(dir)
            found = contents.get(part)

        result.append(found or part)
        dir = os.path.join(dir, part)

    return ''.join(result)

def checknlink(testfile):
    '''check whether hardlink count reporting works properly'''

    # testfile may be open, so we need a separate file for checking to
    # work around issue2543 (or testfile may get lost on Samba shares)
    f1 = testfile + ".hgtmp1"
    if os.path.lexists(f1):
        return False
    try:
        posixfile(f1, 'w').close()
    except IOError:
        try:
            os.unlink(f1)
        except OSError:
            pass
        return False

    f2 = testfile + ".hgtmp2"
    fd = None
    try:
        oslink(f1, f2)
        # nlinks() may behave differently for files on Windows shares if
        # the file is open.
        fd = posixfile(f2)
        return nlinks(f2) > 1
    except OSError:
        return False
    finally:
        if fd is not None:
            fd.close()
        for f in (f1, f2):
            try:
                os.unlink(f)
            except OSError:
                pass

def endswithsep(path):
    '''Check path ends with os.sep or os.altsep.'''
    return (path.endswith(pycompat.ossep)
            or pycompat.osaltsep and path.endswith(pycompat.osaltsep))

def splitpath(path):
    '''Split path by os.sep.
    Note that this function does not use os.altsep because this is
    an alternative of simple "xxx.split(os.sep)".
    It is recommended to use os.path.normpath() before using this
    function if need.'''
    return path.split(pycompat.ossep)

def gui():
    '''Are we running in a GUI?'''
    if pycompat.sysplatform == 'darwin':
        if 'SSH_CONNECTION' in encoding.environ:
            # handle SSH access to a box where the user is logged in
            return False
        elif getattr(osutil, 'isgui', None):
            # check if a CoreGraphics session is available
            return osutil.isgui()
        else:
            # pure build; use a safe default
            return True
    else:
        return pycompat.osname == "nt" or encoding.environ.get("DISPLAY")

def mktempcopy(name, emptyok=False, createmode=None):
    """Create a temporary file with the same contents from name

    The permission bits are copied from the original file.

    If the temporary file is going to be truncated immediately, you
    can use emptyok=True as an optimization.

    Returns the name of the temporary file.
    """
    d, fn = os.path.split(name)
    fd, temp = tempfile.mkstemp(prefix='.%s-' % fn, dir=d)
    os.close(fd)
    # Temporary files are created with mode 0600, which is usually not
    # what we want.  If the original file already exists, just copy
    # its mode.  Otherwise, manually obey umask.
    copymode(name, temp, createmode)
    if emptyok:
        return temp
    try:
        try:
            ifp = posixfile(name, "rb")
        except IOError as inst:
            if inst.errno == errno.ENOENT:
                return temp
            if not getattr(inst, 'filename', None):
                inst.filename = name
            raise
        ofp = posixfile(temp, "wb")
        for chunk in filechunkiter(ifp):
            ofp.write(chunk)
        ifp.close()
        ofp.close()
    except: # re-raises
        try: os.unlink(temp)
        except OSError: pass
        raise
    return temp

class filestat(object):
    """help to exactly detect change of a file

    'stat' attribute is result of 'os.stat()' if specified 'path'
    exists. Otherwise, it is None. This can avoid preparative
    'exists()' examination on client side of this class.
    """
    def __init__(self, path):
        try:
            self.stat = os.stat(path)
        except OSError as err:
            if err.errno != errno.ENOENT:
                raise
            self.stat = None

    __hash__ = object.__hash__

    def __eq__(self, old):
        try:
            # if ambiguity between stat of new and old file is
            # avoided, comparison of size, ctime and mtime is enough
            # to exactly detect change of a file regardless of platform
            return (self.stat.st_size == old.stat.st_size and
                    self.stat.st_ctime == old.stat.st_ctime and
                    self.stat.st_mtime == old.stat.st_mtime)
        except AttributeError:
            return False

    def isambig(self, old):
        """Examine whether new (= self) stat is ambiguous against old one

        "S[N]" below means stat of a file at N-th change:

        - S[n-1].ctime  < S[n].ctime: can detect change of a file
        - S[n-1].ctime == S[n].ctime
          - S[n-1].ctime  < S[n].mtime: means natural advancing (*1)
          - S[n-1].ctime == S[n].mtime: is ambiguous (*2)
          - S[n-1].ctime  > S[n].mtime: never occurs naturally (don't care)
        - S[n-1].ctime  > S[n].ctime: never occurs naturally (don't care)

        Case (*2) above means that a file was changed twice or more at
        same time in sec (= S[n-1].ctime), and comparison of timestamp
        is ambiguous.

        Base idea to avoid such ambiguity is "advance mtime 1 sec, if
        timestamp is ambiguous".

        But advancing mtime only in case (*2) doesn't work as
        expected, because naturally advanced S[n].mtime in case (*1)
        might be equal to manually advanced S[n-1 or earlier].mtime.

        Therefore, all "S[n-1].ctime == S[n].ctime" cases should be
        treated as ambiguous regardless of mtime, to avoid overlooking
        by confliction between such mtime.

        Advancing mtime "if isambig(oldstat)" ensures "S[n-1].mtime !=
        S[n].mtime", even if size of a file isn't changed.
        """
        try:
            return (self.stat.st_ctime == old.stat.st_ctime)
        except AttributeError:
            return False

    def avoidambig(self, path, old):
        """Change file stat of specified path to avoid ambiguity

        'old' should be previous filestat of 'path'.

        This skips avoiding ambiguity, if a process doesn't have
        appropriate privileges for 'path'.
        """
        advanced = (old.stat.st_mtime + 1) & 0x7fffffff
        try:
            os.utime(path, (advanced, advanced))
        except OSError as inst:
            if inst.errno == errno.EPERM:
                # utime() on the file created by another user causes EPERM,
                # if a process doesn't have appropriate privileges
                return
            raise

    def __ne__(self, other):
        return not self == other

class atomictempfile(object):
    '''writable file object that atomically updates a file

    All writes will go to a temporary copy of the original file. Call
    close() when you are done writing, and atomictempfile will rename
    the temporary copy to the original name, making the changes
    visible. If the object is destroyed without being closed, all your
    writes are discarded.

    checkambig argument of constructor is used with filestat, and is
    useful only if target file is guarded by any lock (e.g. repo.lock
    or repo.wlock).
    '''
    def __init__(self, name, mode='w+b', createmode=None, checkambig=False):
        self.__name = name      # permanent name
        self._tempname = mktempcopy(name, emptyok=('w' in mode),
                                    createmode=createmode)
        self._fp = posixfile(self._tempname, mode)
        self._checkambig = checkambig

        # delegated methods
        self.read = self._fp.read
        self.write = self._fp.write
        self.seek = self._fp.seek
        self.tell = self._fp.tell
        self.fileno = self._fp.fileno

    def close(self):
        if not self._fp.closed:
            self._fp.close()
            filename = localpath(self.__name)
            oldstat = self._checkambig and filestat(filename)
            if oldstat and oldstat.stat:
                rename(self._tempname, filename)
                newstat = filestat(filename)
                if newstat.isambig(oldstat):
                    # stat of changed file is ambiguous to original one
                    advanced = (oldstat.stat.st_mtime + 1) & 0x7fffffff
                    os.utime(filename, (advanced, advanced))
            else:
                rename(self._tempname, filename)

    def discard(self):
        if not self._fp.closed:
            try:
                os.unlink(self._tempname)
            except OSError:
                pass
            self._fp.close()

    def __del__(self):
        if safehasattr(self, '_fp'): # constructor actually did something
            self.discard()

    def __enter__(self):
        return self

    def __exit__(self, exctype, excvalue, traceback):
        if exctype is not None:
            self.discard()
        else:
            self.close()

def makedirs(name, mode=None, notindexed=False):
    """recursive directory creation with parent mode inheritance

    Newly created directories are marked as "not to be indexed by
    the content indexing service", if ``notindexed`` is specified
    for "write" mode access.
    """
    try:
        makedir(name, not…