annotate mercurial/encoding.py @ 13425:0fe36c347c00

bookmarks: forbid \0 \r \n : in bookmark names (BC)

We restrict : to 1. make it easer to convert bookmarks to git branches,
2. use : later for a syntax to push a local bookmark to a remote bookmark
of a different name. \0, \n, \r are fobbidden they are used to separate
bookmarks in the bookmark file.

This change breaks backward compatbility as ':' was an allowed character in
previous versions.

# encoding.py - character transcoding support for Mercurial
#
#  Copyright 2005-2009 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.

import error
import unicodedata, locale, os

def _getpreferredencoding():
    '''
    On darwin, getpreferredencoding ignores the locale environment and
    always returns mac-roman. http://bugs.python.org/issue6202 fixes this
    for Python 2.7 and up. This is the same corrected code for earlier
    Python versions.

    However, we can't use a version check for this method, as some distributions
    patch Python to fix this. Instead, we use it as a 'fixer' for the mac-roman
    encoding, as it is unlikely that this encoding is the actually expected.
    '''
    try:
        locale.CODESET
    except AttributeError:
        # Fall back to parsing environment variables :-(
        return locale.getdefaultlocale()[1]

    oldloc = locale.setlocale(locale.LC_CTYPE)
    locale.setlocale(locale.LC_CTYPE, "")
    result = locale.nl_langinfo(locale.CODESET)
    locale.setlocale(locale.LC_CTYPE, oldloc)

    return result

_encodingfixers = {
    '646': lambda: 'ascii',
    'ANSI_X3.4-1968': lambda: 'ascii',
    'mac-roman': _getpreferredencoding
}

try:
    encoding = os.environ.get("HGENCODING")
    if not encoding:
        encoding = locale.getpreferredencoding() or 'ascii'
        encoding = _encodingfixers.get(encoding, lambda: encoding)()
except locale.Error:
    encoding = 'ascii'
encodingmode = os.environ.get("HGENCODINGMODE", "strict")
fallbackencoding = 'ISO-8859-1'

class localstr(str):
    '''This class allows strings that are unmodified to be
    round-tripped to the local encoding and back'''
    def __new__(cls, u, l):
        s = str.__new__(cls, l)
        s._utf8 = u
        return s
    def __hash__(self):
        return hash(self._utf8) # avoid collisions in local string space

def tolocal(s):
    """
    Convert a string from internal UTF-8 to local encoding

    All internal strings should be UTF-8 but some repos before the
    implementation of locale support may contain latin1 or possibly
    other character sets. We attempt to decode everything strictly
    using UTF-8, then Latin-1, and failing that, we use UTF-8 and
    replace unknown characters.

    The localstr class is used to cache the known UTF-8 encoding of
    strings next to their local representation to allow lossless
    round-trip conversion back to UTF-8.

    >>> u = 'foo: \\xc3\\xa4' # utf-8
    >>> l = tolocal(u)
    >>> l
    'foo: ?'
    >>> fromlocal(l)
    'foo: \\xc3\\xa4'
    >>> u2 = 'foo: \\xc3\\xa1'
    >>> d = { l: 1, tolocal(u2): 2 }
    >>> d # no collision
    {'foo: ?': 1, 'foo: ?': 2}
    >>> 'foo: ?' in d
    False
    >>> l1 = 'foo: \\xe4' # historical latin1 fallback
    >>> l = tolocal(l1)
    >>> l
    'foo: ?'
    >>> fromlocal(l) # magically in utf-8
    'foo: \\xc3\\xa4'
    """

    for e in ('UTF-8', fallbackencoding):
        try:
            u = s.decode(e) # attempt strict decoding
            if e == 'UTF-8':
                return localstr(s, u.encode(encoding, "replace"))
            else:
                return localstr(u.encode('UTF-8'),
                                u.encode(encoding, "replace"))
        except LookupError, k:
            raise error.Abort("%s, please check your locale settings" % k)
        except UnicodeDecodeError:
            pass
    u = s.decode("utf-8", "replace") # last ditch
    return u.encode(encoding, "replace") # can't round-trip

def fromlocal(s):
    """
    Convert a string from the local character encoding to UTF-8

    We attempt to decode strings using the encoding mode set by
    HGENCODINGMODE, which defaults to 'strict'. In this mode, unknown
    characters will cause an error message. Other modes include
    'replace', which replaces unknown characters with a special
    Unicode character, and 'ignore', which drops the character.
    """

    # can we do a lossless round-trip?
    if isinstance(s, localstr):
        return s._utf8

    try:
        return s.decode(encoding, encodingmode).encode("utf-8")
    except UnicodeDecodeError, inst:
        sub = s[max(0, inst.start - 10):inst.start + 10]
        raise error.Abort("decoding near '%s': %s!" % (sub, inst))
    except LookupError, k:
        raise error.Abort("%s, please check your locale settings" % k)

# How to treat ambiguous-width characters. Set to 'wide' to treat as wide.
ambiguous = os.environ.get("HGENCODINGAMBIGUOUS", "narrow")

def colwidth(s):
    "Find the column width of a UTF-8 string for display"
    d = s.decode(encoding, 'replace')
    if hasattr(unicodedata, 'east_asian_width'):
        wide = "WF"
        if ambiguous == "wide":
            wide = "WFA"
        w = unicodedata.east_asian_width
        return sum([w(c) in wide and 2 or 1 for c in d])
    return len(d)