+++ /dev/null
-# Copyright (C) 2001-2006 Python Software Foundation\r
-# Author: Ben Gertzfield, Barry Warsaw\r
-# Contact: email-sig@python.org\r
-\r
-__all__ = [\r
- 'Charset',\r
- 'add_alias',\r
- 'add_charset',\r
- 'add_codec',\r
- ]\r
-\r
-import codecs\r
-import email.base64mime\r
-import email.quoprimime\r
-\r
-from email import errors\r
-from email.encoders import encode_7or8bit\r
-\r
-\r
-\f\r
-# Flags for types of header encodings\r
-QP = 1 # Quoted-Printable\r
-BASE64 = 2 # Base64\r
-SHORTEST = 3 # the shorter of QP and base64, but only for headers\r
-\r
-# In "=?charset?q?hello_world?=", the =?, ?q?, and ?= add up to 7\r
-MISC_LEN = 7\r
-\r
-DEFAULT_CHARSET = 'us-ascii'\r
-\r
-\r
-\f\r
-# Defaults\r
-CHARSETS = {\r
- # input header enc body enc output conv\r
- 'iso-8859-1': (QP, QP, None),\r
- 'iso-8859-2': (QP, QP, None),\r
- 'iso-8859-3': (QP, QP, None),\r
- 'iso-8859-4': (QP, QP, None),\r
- # iso-8859-5 is Cyrillic, and not especially used\r
- # iso-8859-6 is Arabic, also not particularly used\r
- # iso-8859-7 is Greek, QP will not make it readable\r
- # iso-8859-8 is Hebrew, QP will not make it readable\r
- 'iso-8859-9': (QP, QP, None),\r
- 'iso-8859-10': (QP, QP, None),\r
- # iso-8859-11 is Thai, QP will not make it readable\r
- 'iso-8859-13': (QP, QP, None),\r
- 'iso-8859-14': (QP, QP, None),\r
- 'iso-8859-15': (QP, QP, None),\r
- 'iso-8859-16': (QP, QP, None),\r
- 'windows-1252':(QP, QP, None),\r
- 'viscii': (QP, QP, None),\r
- 'us-ascii': (None, None, None),\r
- 'big5': (BASE64, BASE64, None),\r
- 'gb2312': (BASE64, BASE64, None),\r
- 'euc-jp': (BASE64, None, 'iso-2022-jp'),\r
- 'shift_jis': (BASE64, None, 'iso-2022-jp'),\r
- 'iso-2022-jp': (BASE64, None, None),\r
- 'koi8-r': (BASE64, BASE64, None),\r
- 'utf-8': (SHORTEST, BASE64, 'utf-8'),\r
- # We're making this one up to represent raw unencoded 8-bit\r
- '8bit': (None, BASE64, 'utf-8'),\r
- }\r
-\r
-# Aliases for other commonly-used names for character sets. Map\r
-# them to the real ones used in email.\r
-ALIASES = {\r
- 'latin_1': 'iso-8859-1',\r
- 'latin-1': 'iso-8859-1',\r
- 'latin_2': 'iso-8859-2',\r
- 'latin-2': 'iso-8859-2',\r
- 'latin_3': 'iso-8859-3',\r
- 'latin-3': 'iso-8859-3',\r
- 'latin_4': 'iso-8859-4',\r
- 'latin-4': 'iso-8859-4',\r
- 'latin_5': 'iso-8859-9',\r
- 'latin-5': 'iso-8859-9',\r
- 'latin_6': 'iso-8859-10',\r
- 'latin-6': 'iso-8859-10',\r
- 'latin_7': 'iso-8859-13',\r
- 'latin-7': 'iso-8859-13',\r
- 'latin_8': 'iso-8859-14',\r
- 'latin-8': 'iso-8859-14',\r
- 'latin_9': 'iso-8859-15',\r
- 'latin-9': 'iso-8859-15',\r
- 'latin_10':'iso-8859-16',\r
- 'latin-10':'iso-8859-16',\r
- 'cp949': 'ks_c_5601-1987',\r
- 'euc_jp': 'euc-jp',\r
- 'euc_kr': 'euc-kr',\r
- 'ascii': 'us-ascii',\r
- }\r
-\r
-\r
-# Map charsets to their Unicode codec strings.\r
-CODEC_MAP = {\r
- 'gb2312': 'eucgb2312_cn',\r
- 'big5': 'big5_tw',\r
- # Hack: We don't want *any* conversion for stuff marked us-ascii, as all\r
- # sorts of garbage might be sent to us in the guise of 7-bit us-ascii.\r
- # Let that stuff pass through without conversion to/from Unicode.\r
- 'us-ascii': None,\r
- }\r
-\r
-\r
-\f\r
-# Convenience functions for extending the above mappings\r
-def add_charset(charset, header_enc=None, body_enc=None, output_charset=None):\r
- """Add character set properties to the global registry.\r
-\r
- charset is the input character set, and must be the canonical name of a\r
- character set.\r
-\r
- Optional header_enc and body_enc is either Charset.QP for\r
- quoted-printable, Charset.BASE64 for base64 encoding, Charset.SHORTEST for\r
- the shortest of qp or base64 encoding, or None for no encoding. SHORTEST\r
- is only valid for header_enc. It describes how message headers and\r
- message bodies in the input charset are to be encoded. Default is no\r
- encoding.\r
-\r
- Optional output_charset is the character set that the output should be\r
- in. Conversions will proceed from input charset, to Unicode, to the\r
- output charset when the method Charset.convert() is called. The default\r
- is to output in the same character set as the input.\r
-\r
- Both input_charset and output_charset must have Unicode codec entries in\r
- the module's charset-to-codec mapping; use add_codec(charset, codecname)\r
- to add codecs the module does not know about. See the codecs module's\r
- documentation for more information.\r
- """\r
- if body_enc == SHORTEST:\r
- raise ValueError('SHORTEST not allowed for body_enc')\r
- CHARSETS[charset] = (header_enc, body_enc, output_charset)\r
-\r
-\r
-def add_alias(alias, canonical):\r
- """Add a character set alias.\r
-\r
- alias is the alias name, e.g. latin-1\r
- canonical is the character set's canonical name, e.g. iso-8859-1\r
- """\r
- ALIASES[alias] = canonical\r
-\r
-\r
-def add_codec(charset, codecname):\r
- """Add a codec that map characters in the given charset to/from Unicode.\r
-\r
- charset is the canonical name of a character set. codecname is the name\r
- of a Python codec, as appropriate for the second argument to the unicode()\r
- built-in, or to the encode() method of a Unicode string.\r
- """\r
- CODEC_MAP[charset] = codecname\r
-\r
-\r
-\f\r
-class Charset:\r
- """Map character sets to their email properties.\r
-\r
- This class provides information about the requirements imposed on email\r
- for a specific character set. It also provides convenience routines for\r
- converting between character sets, given the availability of the\r
- applicable codecs. Given a character set, it will do its best to provide\r
- information on how to use that character set in an email in an\r
- RFC-compliant way.\r
-\r
- Certain character sets must be encoded with quoted-printable or base64\r
- when used in email headers or bodies. Certain character sets must be\r
- converted outright, and are not allowed in email. Instances of this\r
- module expose the following information about a character set:\r
-\r
- input_charset: The initial character set specified. Common aliases\r
- are converted to their `official' email names (e.g. latin_1\r
- is converted to iso-8859-1). Defaults to 7-bit us-ascii.\r
-\r
- header_encoding: If the character set must be encoded before it can be\r
- used in an email header, this attribute will be set to\r
- Charset.QP (for quoted-printable), Charset.BASE64 (for\r
- base64 encoding), or Charset.SHORTEST for the shortest of\r
- QP or BASE64 encoding. Otherwise, it will be None.\r
-\r
- body_encoding: Same as header_encoding, but describes the encoding for the\r
- mail message's body, which indeed may be different than the\r
- header encoding. Charset.SHORTEST is not allowed for\r
- body_encoding.\r
-\r
- output_charset: Some character sets must be converted before the can be\r
- used in email headers or bodies. If the input_charset is\r
- one of them, this attribute will contain the name of the\r
- charset output will be converted to. Otherwise, it will\r
- be None.\r
-\r
- input_codec: The name of the Python codec used to convert the\r
- input_charset to Unicode. If no conversion codec is\r
- necessary, this attribute will be None.\r
-\r
- output_codec: The name of the Python codec used to convert Unicode\r
- to the output_charset. If no conversion codec is necessary,\r
- this attribute will have the same value as the input_codec.\r
- """\r
- def __init__(self, input_charset=DEFAULT_CHARSET):\r
- # RFC 2046, $4.1.2 says charsets are not case sensitive. We coerce to\r
- # unicode because its .lower() is locale insensitive. If the argument\r
- # is already a unicode, we leave it at that, but ensure that the\r
- # charset is ASCII, as the standard (RFC XXX) requires.\r
- try:\r
- if isinstance(input_charset, unicode):\r
- input_charset.encode('ascii')\r
- else:\r
- input_charset = unicode(input_charset, 'ascii')\r
- except UnicodeError:\r
- raise errors.CharsetError(input_charset)\r
- input_charset = input_charset.lower().encode('ascii')\r
- # Set the input charset after filtering through the aliases and/or codecs\r
- if not (input_charset in ALIASES or input_charset in CHARSETS):\r
- try:\r
- input_charset = codecs.lookup(input_charset).name\r
- except LookupError:\r
- pass\r
- self.input_charset = ALIASES.get(input_charset, input_charset)\r
- # We can try to guess which encoding and conversion to use by the\r
- # charset_map dictionary. Try that first, but let the user override\r
- # it.\r
- henc, benc, conv = CHARSETS.get(self.input_charset,\r
- (SHORTEST, BASE64, None))\r
- if not conv:\r
- conv = self.input_charset\r
- # Set the attributes, allowing the arguments to override the default.\r
- self.header_encoding = henc\r
- self.body_encoding = benc\r
- self.output_charset = ALIASES.get(conv, conv)\r
- # Now set the codecs. If one isn't defined for input_charset,\r
- # guess and try a Unicode codec with the same name as input_codec.\r
- self.input_codec = CODEC_MAP.get(self.input_charset,\r
- self.input_charset)\r
- self.output_codec = CODEC_MAP.get(self.output_charset,\r
- self.output_charset)\r
-\r
- def __str__(self):\r
- return self.input_charset.lower()\r
-\r
- __repr__ = __str__\r
-\r
- def __eq__(self, other):\r
- return str(self) == str(other).lower()\r
-\r
- def __ne__(self, other):\r
- return not self.__eq__(other)\r
-\r
- def get_body_encoding(self):\r
- """Return the content-transfer-encoding used for body encoding.\r
-\r
- This is either the string `quoted-printable' or `base64' depending on\r
- the encoding used, or it is a function in which case you should call\r
- the function with a single argument, the Message object being\r
- encoded. The function should then set the Content-Transfer-Encoding\r
- header itself to whatever is appropriate.\r
-\r
- Returns "quoted-printable" if self.body_encoding is QP.\r
- Returns "base64" if self.body_encoding is BASE64.\r
- Returns "7bit" otherwise.\r
- """\r
- assert self.body_encoding != SHORTEST\r
- if self.body_encoding == QP:\r
- return 'quoted-printable'\r
- elif self.body_encoding == BASE64:\r
- return 'base64'\r
- else:\r
- return encode_7or8bit\r
-\r
- def convert(self, s):\r
- """Convert a string from the input_codec to the output_codec."""\r
- if self.input_codec != self.output_codec:\r
- return unicode(s, self.input_codec).encode(self.output_codec)\r
- else:\r
- return s\r
-\r
- def to_splittable(self, s):\r
- """Convert a possibly multibyte string to a safely splittable format.\r
-\r
- Uses the input_codec to try and convert the string to Unicode, so it\r
- can be safely split on character boundaries (even for multibyte\r
- characters).\r
-\r
- Returns the string as-is if it isn't known how to convert it to\r
- Unicode with the input_charset.\r
-\r
- Characters that could not be converted to Unicode will be replaced\r
- with the Unicode replacement character U+FFFD.\r
- """\r
- if isinstance(s, unicode) or self.input_codec is None:\r
- return s\r
- try:\r
- return unicode(s, self.input_codec, 'replace')\r
- except LookupError:\r
- # Input codec not installed on system, so return the original\r
- # string unchanged.\r
- return s\r
-\r
- def from_splittable(self, ustr, to_output=True):\r
- """Convert a splittable string back into an encoded string.\r
-\r
- Uses the proper codec to try and convert the string from Unicode back\r
- into an encoded format. Return the string as-is if it is not Unicode,\r
- or if it could not be converted from Unicode.\r
-\r
- Characters that could not be converted from Unicode will be replaced\r
- with an appropriate character (usually '?').\r
-\r
- If to_output is True (the default), uses output_codec to convert to an\r
- encoded format. If to_output is False, uses input_codec.\r
- """\r
- if to_output:\r
- codec = self.output_codec\r
- else:\r
- codec = self.input_codec\r
- if not isinstance(ustr, unicode) or codec is None:\r
- return ustr\r
- try:\r
- return ustr.encode(codec, 'replace')\r
- except LookupError:\r
- # Output codec not installed\r
- return ustr\r
-\r
- def get_output_charset(self):\r
- """Return the output character set.\r
-\r
- This is self.output_charset if that is not None, otherwise it is\r
- self.input_charset.\r
- """\r
- return self.output_charset or self.input_charset\r
-\r
- def encoded_header_len(self, s):\r
- """Return the length of the encoded header string."""\r
- cset = self.get_output_charset()\r
- # The len(s) of a 7bit encoding is len(s)\r
- if self.header_encoding == BASE64:\r
- return email.base64mime.base64_len(s) + len(cset) + MISC_LEN\r
- elif self.header_encoding == QP:\r
- return email.quoprimime.header_quopri_len(s) + len(cset) + MISC_LEN\r
- elif self.header_encoding == SHORTEST:\r
- lenb64 = email.base64mime.base64_len(s)\r
- lenqp = email.quoprimime.header_quopri_len(s)\r
- return min(lenb64, lenqp) + len(cset) + MISC_LEN\r
- else:\r
- return len(s)\r
-\r
- def header_encode(self, s, convert=False):\r
- """Header-encode a string, optionally converting it to output_charset.\r
-\r
- If convert is True, the string will be converted from the input\r
- charset to the output charset automatically. This is not useful for\r
- multibyte character sets, which have line length issues (multibyte\r
- characters must be split on a character, not a byte boundary); use the\r
- high-level Header class to deal with these issues. convert defaults\r
- to False.\r
-\r
- The type of encoding (base64 or quoted-printable) will be based on\r
- self.header_encoding.\r
- """\r
- cset = self.get_output_charset()\r
- if convert:\r
- s = self.convert(s)\r
- # 7bit/8bit encodings return the string unchanged (modulo conversions)\r
- if self.header_encoding == BASE64:\r
- return email.base64mime.header_encode(s, cset)\r
- elif self.header_encoding == QP:\r
- return email.quoprimime.header_encode(s, cset, maxlinelen=None)\r
- elif self.header_encoding == SHORTEST:\r
- lenb64 = email.base64mime.base64_len(s)\r
- lenqp = email.quoprimime.header_quopri_len(s)\r
- if lenb64 < lenqp:\r
- return email.base64mime.header_encode(s, cset)\r
- else:\r
- return email.quoprimime.header_encode(s, cset, maxlinelen=None)\r
- else:\r
- return s\r
-\r
- def body_encode(self, s, convert=True):\r
- """Body-encode a string and convert it to output_charset.\r
-\r
- If convert is True (the default), the string will be converted from\r
- the input charset to output charset automatically. Unlike\r
- header_encode(), there are no issues with byte boundaries and\r
- multibyte charsets in email bodies, so this is usually pretty safe.\r
-\r
- The type of encoding (base64 or quoted-printable) will be based on\r
- self.body_encoding.\r
- """\r
- if convert:\r
- s = self.convert(s)\r
- # 7bit/8bit encodings return the string unchanged (module conversions)\r
- if self.body_encoding is BASE64:\r
- return email.base64mime.body_encode(s)\r
- elif self.body_encoding is QP:\r
- return email.quoprimime.body_encode(s)\r
- else:\r
- return s\r