charset.py

来自「mallet是自然语言处理、机器学习领域的一个开源项目。」· Python 代码 · 共 394 行 · 第 1/2 页

# Copyright (C) 2001,2002 Python Software Foundation
# Author: [e-mail] (Ben Gertzfield), [e-mail] (Barry Warsaw)

from types import UnicodeType
from email.Encoders import encode_7or8bit
import email.base64MIME
import email.quopriMIME

def _isunicode(s):
    return isinstance(s, UnicodeType)

# Python 2.2.1 and beyond has these symbols
try:
    True, False
except NameError:
    True = 1
    False = 0



# Flags for types of header encodings
QP     = 1   # Quoted-Printable
BASE64 = 2   # Base64
SHORTEST = 3 # the shorter of QP and base64, but only for headers

# In "=?charset?q?hello_world?=", the =?, ?q?, and ?= add up to 7
MISC_LEN = 7

DEFAULT_CHARSET = 'us-ascii'



# Defaults
CHARSETS = {
    # input        header enc  body enc output conv
    'iso-8859-1':  (QP,        QP,      None),
    'iso-8859-2':  (QP,        QP,      None),
    'iso-8859-3':  (QP,        QP,      None),
    'iso-8859-4':  (QP,        QP,      None),
    # iso-8859-5 is Cyrillic, and not especially used
    # iso-8859-6 is Arabic, also not particularly used
    # iso-8859-7 is Greek, QP will not make it readable
    # iso-8859-8 is Hebrew, QP will not make it readable
    'iso-8859-9':  (QP,        QP,      None),
    'iso-8859-10': (QP,        QP,      None),
    # iso-8859-11 is Thai, QP will not make it readable
    'iso-8859-13': (QP,        QP,      None),
    'iso-8859-14': (QP,        QP,      None),
    'iso-8859-15': (QP,        QP,      None),
    'windows-1252':(QP,        QP,      None),
    'viscii':      (QP,        QP,      None),
    'us-ascii':    (None,      None,    None),
    'big5':        (BASE64,    BASE64,  None),
    'gb2312':      (BASE64,    BASE64,  None),
    'euc-jp':      (BASE64,    None,    'iso-2022-jp'),
    'shift_jis':   (BASE64,    None,    'iso-2022-jp'),
    'iso-2022-jp': (BASE64,    None,    None),
    'koi8-r':      (BASE64,    BASE64,  None),
    'utf-8':       (SHORTEST,  BASE64, 'utf-8'),
    # We're making this one up to represent raw unencoded 8-bit
    '8bit':        (None,      BASE64, 'utf-8'),
    }

# Aliases for other commonly-used names for character sets.  Map
# them to the real ones used in email.
ALIASES = {
    'latin_1': 'iso-8859-1',
    'latin-1': 'iso-8859-1',
    'latin_2': 'iso-8859-2',
    'latin-2': 'iso-8859-2',
    'latin_3': 'iso-8859-3',
    'latin-3': 'iso-8859-3',
    'latin_4': 'iso-8859-4',
    'latin-4': 'iso-8859-4',
    'latin_5': 'iso-8859-9',
    'latin-5': 'iso-8859-9',
    'latin_6': 'iso-8859-10',
    'latin-6': 'iso-8859-10',
    'latin_7': 'iso-8859-13',
    'latin-7': 'iso-8859-13',
    'latin_8': 'iso-8859-14',
    'latin-8': 'iso-8859-14',
    'latin_9': 'iso-8859-15',
    'latin-9': 'iso-8859-15',
    'cp949':   'ks_c_5601-1987',
    'euc_jp':  'euc-jp',
    'euc_kr':  'euc-kr',
    'ascii':   'us-ascii',
    }

# Map charsets to their Unicode codec strings.  Note that Python doesn't come
# with any Asian codecs by default.  Here's where to get them:
#
# Japanese -- http://www.asahi-net.or.jp/~rd6t-kjym/python
# Korean   -- http://sf.net/projects/koco
# Chinese  -- http://sf.net/projects/python-codecs
#
# Note that these codecs have their own lifecycle and may be in varying states
# of stability and useability.

CODEC_MAP = {
    'euc-jp':      'japanese.euc-jp',
    'iso-2022-jp': 'japanese.iso-2022-jp',
    'shift_jis':   'japanese.shift_jis',
    'euc-kr':      'korean.euc-kr',
    'ks_c_5601-1987': 'korean.cp949',
    'iso-2022-kr': 'korean.iso-2022-kr',
    'johab':       'korean.johab',
    'gb2132':      'eucgb2312_cn',
    'big5':        'big5_tw',
    'utf-8':       'utf-8',
    # Hack: We don't want *any* conversion for stuff marked us-ascii, as all
    # sorts of garbage might be sent to us in the guise of 7-bit us-ascii.
    # Let that stuff pass through without conversion to/from Unicode.
    'us-ascii':    None,
    }



# Convenience functions for extending the above mappings
def add_charset(charset, header_enc=None, body_enc=None, output_charset=None):
    """Add character set properties to the global registry.

    charset is the input character set, and must be the canonical name of a
    character set.

    Optional header_enc and body_enc is either Charset.QP for
    quoted-printable, Charset.BASE64 for base64 encoding, Charset.SHORTEST for
    the shortest of qp or base64 encoding, or None for no encoding.  SHORTEST
    is only valid for header_enc.  It describes how message headers and
    message bodies in the input charset are to be encoded.  Default is no
    encoding.

    Optional output_charset is the character set that the output should be
    in.  Conversions will proceed from input charset, to Unicode, to the
    output charset when the method Charset.convert() is called.  The default
    is to output in the same character set as the input.

    Both input_charset and output_charset must have Unicode codec entries in
    the module's charset-to-codec mapping; use add_codec(charset, codecname)
    to add codecs the module does not know about.  See the codecs module's
    documentation for more information.
    """
    if body_enc == SHORTEST:
        raise ValueError, 'SHORTEST not allowed for body_enc'
    CHARSETS[charset] = (header_enc, body_enc, output_charset)


def add_alias(alias, canonical):
    """Add a character set alias.

    alias is the alias name, e.g. latin-1
    canonical is the character set's canonical name, e.g. iso-8859-1
    """
    ALIASES[alias] = canonical


def add_codec(charset, codecname):
    """Add a codec that map characters in the given charset to/from Unicode.

    charset is the canonical name of a character set.  codecname is the name
    of a Python codec, as appropriate for the second argument to the unicode()
    built-in, or to the encode() method of a Unicode string.
    """
    CODEC_MAP[charset] = codecname



class Charset:
    """Map character sets to their email properties.

    This class provides information about the requirements imposed on email
    for a specific character set.  It also provides convenience routines for
    converting between character sets, given the availability of the
    applicable codecs.  Given a character set, it will do its best to provide
    information on how to use that character set in an email in an
    RFC-compliant way.

    Certain character sets must be encoded with quoted-printable or base64
    when used in email headers or bodies.  Certain character sets must be
    converted outright, and are not allowed in email.  Instances of this
    module expose the following information about a character set:

    input_charset: The initial character set specified.  Common aliases
                   are converted to their `official' email names (e.g. latin_1
                   is converted to iso-8859-1).  Defaults to 7-bit us-ascii.

    header_encoding: If the character set must be encoded before it can be
                     used in an email header, this attribute will be set to
                     Charset.QP (for quoted-printable), Charset.BASE64 (for
                     base64 encoding), or Charset.SHORTEST for the shortest of
                     QP or BASE64 encoding.  Otherwise, it will be None.

    body_encoding: Same as header_encoding, but describes the encoding for the
                   mail message's body, which indeed may be different than the
                   header encoding.  Charset.SHORTEST is not allowed for
                   body_encoding.