1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
|
"""
Functions for handling encoding of web pages
"""
from __future__ import annotations
import codecs
import encodings
import re
from re import Match
from typing import Callable, cast
import w3lib.util
from w3lib._types import AnyUnicodeError, StrOrBytes
_HEADER_ENCODING_RE = re.compile(r"charset=([\w-]+)", re.I)
def http_content_type_encoding(content_type: str | None) -> str | None:
"""Extract the encoding in the content-type header
>>> import w3lib.encoding
>>> w3lib.encoding.http_content_type_encoding("Content-Type: text/html; charset=ISO-8859-4")
'iso8859-4'
"""
if content_type:
match = _HEADER_ENCODING_RE.search(content_type)
if match:
return resolve_encoding(match.group(1))
return None
# regexp for parsing HTTP meta tags
_TEMPLATE = r"""%s\s*=\s*["']?\s*%s\s*["']?"""
_SKIP_ATTRS = """(?:\\s+
[^=<>/\\s"'\x00-\x1f\x7f]+ # Attribute name
(?:\\s*=\\s*
(?: # ' and " are entity encoded (', "), so no need for \', \"
'[^']*' # attr in '
|
"[^"]*" # attr in "
|
[^'"\\s]+ # attr having no ' nor "
))?
)*?""" # must be used with re.VERBOSE flag
_HTTPEQUIV_RE = _TEMPLATE % ("http-equiv", "Content-Type")
_CONTENT_RE = _TEMPLATE % ("content", r"(?P<mime>[^;]+);\s*charset=(?P<charset>[\w-]+)")
_CONTENT2_RE = _TEMPLATE % ("charset", r"(?P<charset2>[\w-]+)")
_XML_ENCODING_RE = _TEMPLATE % ("encoding", r"(?P<xmlcharset>[\w-]+)")
# check for meta tags, or xml decl. and stop search if a body tag is encountered
_BODY_ENCODING_PATTERN = (
r"<\s*(?:meta%s(?:(?:\s+%s|\s+%s){2}|\s+%s)|\?xml\s[^>]+%s|body)"
% (_SKIP_ATTRS, _HTTPEQUIV_RE, _CONTENT_RE, _CONTENT2_RE, _XML_ENCODING_RE)
)
_BODY_ENCODING_STR_RE = re.compile(_BODY_ENCODING_PATTERN, re.I | re.VERBOSE)
_BODY_ENCODING_BYTES_RE = re.compile(
_BODY_ENCODING_PATTERN.encode("ascii"), re.I | re.VERBOSE
)
def html_body_declared_encoding(html_body_str: StrOrBytes) -> str | None:
'''Return the encoding specified in meta tags in the html body,
or ``None`` if no suitable encoding was found
>>> import w3lib.encoding
>>> w3lib.encoding.html_body_declared_encoding(
... """<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
... "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
... <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
... <head>
... <title>Some title</title>
... <meta http-equiv="content-type" content="text/html;charset=utf-8" />
... </head>
... <body>
... ...
... </body>
... </html>""")
'utf-8'
>>>
'''
# html5 suggests the first 1024 bytes are sufficient, we allow for more
chunk = html_body_str[:4096]
match: Match[bytes] | Match[str] | None
if isinstance(chunk, bytes):
match = _BODY_ENCODING_BYTES_RE.search(chunk)
else:
match = _BODY_ENCODING_STR_RE.search(chunk)
if match:
encoding = (
match.group("charset")
or match.group("charset2")
or match.group("xmlcharset")
)
if encoding:
return resolve_encoding(w3lib.util.to_unicode(encoding))
return None
# Default encoding translation
# this maps cannonicalized encodings to target encodings
# see http://www.whatwg.org/specs/web-apps/current-work/multipage/parsing.html#character-encodings-0
# in addition, gb18030 supercedes gb2312 & gbk
# the keys are converted using _c18n_encoding and in sorted order
DEFAULT_ENCODING_TRANSLATION = {
"ascii": "cp1252",
"big5": "big5hkscs",
"euc_kr": "cp949",
"gb2312": "gb18030",
"gb_2312_80": "gb18030",
"gbk": "gb18030",
"iso8859_11": "cp874",
"iso8859_9": "cp1254",
"latin_1": "cp1252",
"macintosh": "mac_roman",
"shift_jis": "cp932",
"tis_620": "cp874",
"win_1251": "cp1251",
"windows_31j": "cp932",
"win_31j": "cp932",
"windows_874": "cp874",
"win_874": "cp874",
"x_sjis": "cp932",
"zh_cn": "gb18030",
}
def _c18n_encoding(encoding: str) -> str:
"""Canonicalize an encoding name
This performs normalization and translates aliases using python's
encoding aliases
"""
normed = encodings.normalize_encoding(encoding).lower()
return cast(str, encodings.aliases.aliases.get(normed, normed))
def resolve_encoding(encoding_alias: str) -> str | None:
"""Return the encoding that `encoding_alias` maps to, or ``None``
if the encoding cannot be interpreted
>>> import w3lib.encoding
>>> w3lib.encoding.resolve_encoding('latin1')
'cp1252'
>>> w3lib.encoding.resolve_encoding('gb_2312-80')
'gb18030'
>>>
"""
c18n_encoding = _c18n_encoding(encoding_alias)
translated = DEFAULT_ENCODING_TRANSLATION.get(c18n_encoding, c18n_encoding)
try:
return codecs.lookup(translated).name
except LookupError:
return None
_BOM_TABLE = [
(codecs.BOM_UTF32_BE, "utf-32-be"),
(codecs.BOM_UTF32_LE, "utf-32-le"),
(codecs.BOM_UTF16_BE, "utf-16-be"),
(codecs.BOM_UTF16_LE, "utf-16-le"),
(codecs.BOM_UTF8, "utf-8"),
]
_FIRST_CHARS = {c[0] for (c, _) in _BOM_TABLE}
def read_bom(data: bytes) -> tuple[None, None] | tuple[str, bytes]:
r"""Read the byte order mark in the text, if present, and
return the encoding represented by the BOM and the BOM.
If no BOM can be detected, ``(None, None)`` is returned.
>>> import w3lib.encoding
>>> w3lib.encoding.read_bom(b'\xfe\xff\x6c\x34')
('utf-16-be', '\xfe\xff')
>>> w3lib.encoding.read_bom(b'\xff\xfe\x34\x6c')
('utf-16-le', '\xff\xfe')
>>> w3lib.encoding.read_bom(b'\x00\x00\xfe\xff\x00\x00\x6c\x34')
('utf-32-be', '\x00\x00\xfe\xff')
>>> w3lib.encoding.read_bom(b'\xff\xfe\x00\x00\x34\x6c\x00\x00')
('utf-32-le', '\xff\xfe\x00\x00')
>>> w3lib.encoding.read_bom(b'\x01\x02\x03\x04')
(None, None)
>>>
"""
# common case is no BOM, so this is fast
if data and data[0] in _FIRST_CHARS:
for bom, encoding in _BOM_TABLE:
if data.startswith(bom):
return encoding, bom
return None, None
# Python decoder doesn't follow unicode standard when handling
# bad utf-8 encoded strings. see http://bugs.python.org/issue8271
codecs.register_error(
"w3lib_replace", lambda exc: ("\ufffd", cast(AnyUnicodeError, exc).end)
)
def to_unicode(data_str: bytes, encoding: str) -> str:
"""Convert a str object to unicode using the encoding given
Characters that cannot be converted will be converted to ``\\ufffd`` (the
unicode replacement character).
"""
return data_str.decode(encoding, "replace")
def html_to_unicode(
content_type_header: str | None,
html_body_str: bytes,
default_encoding: str = "utf8",
auto_detect_fun: Callable[[bytes], str | None] | None = None,
) -> tuple[str, str]:
r'''Convert raw html bytes to unicode
This attempts to make a reasonable guess at the content encoding of the
html body, following a similar process to a web browser.
It will try in order:
* BOM (byte-order mark)
* http content type header
* meta or xml tag declarations
* auto-detection, if the `auto_detect_fun` keyword argument is not ``None``
* default encoding in keyword arg (which defaults to utf8)
If an encoding other than the auto-detected or default encoding is used,
overrides will be applied, converting some character encodings to more
suitable alternatives.
If a BOM is found matching the encoding, it will be stripped.
The `auto_detect_fun` argument can be used to pass a function that will
sniff the encoding of the text. This function must take the raw text as an
argument and return the name of an encoding that python can process, or
None. To use chardet, for example, you can define the function as::
auto_detect_fun=lambda x: chardet.detect(x).get('encoding')
or to use UnicodeDammit (shipped with the BeautifulSoup library)::
auto_detect_fun=lambda x: UnicodeDammit(x).originalEncoding
If the locale of the website or user language preference is known, then a
better default encoding can be supplied.
If `content_type_header` is not present, ``None`` can be passed signifying
that the header was not present.
This method will not fail, if characters cannot be converted to unicode,
``\\ufffd`` (the unicode replacement character) will be inserted instead.
Returns a tuple of ``(<encoding used>, <unicode_string>)``
Examples:
>>> import w3lib.encoding
>>> w3lib.encoding.html_to_unicode(None,
... b"""<!DOCTYPE html>
... <head>
... <meta charset="UTF-8" />
... <meta name="viewport" content="width=device-width" />
... <title>Creative Commons France</title>
... <link rel='canonical' href='http://creativecommons.fr/' />
... <body>
... <p>Creative Commons est une organisation \xc3\xa0 but non lucratif
... qui a pour dessein de faciliter la diffusion et le partage des oeuvres
... tout en accompagnant les nouvelles pratiques de cr\xc3\xa9ation \xc3\xa0 l\xe2\x80\x99\xc3\xa8re numerique.</p>
... </body>
... </html>""")
('utf-8', '<!DOCTYPE html>\n<head>\n<meta charset="UTF-8" />\n<meta name="viewport" content="width=device-width" />\n<title>Creative Commons France</title>\n<link rel=\'canonical\' href=\'http://creativecommons.fr/\' />\n<body>\n<p>Creative Commons est une organisation \xe0 but non lucratif\nqui a pour dessein de faciliter la diffusion et le partage des oeuvres\ntout en accompagnant les nouvelles pratiques de cr\xe9ation \xe0 l\u2019\xe8re numerique.</p>\n</body>\n</html>')
>>>
'''
bom_enc, bom = read_bom(html_body_str)
if bom_enc is not None:
bom = cast(bytes, bom)
return bom_enc, to_unicode(html_body_str[len(bom) :], bom_enc)
enc = http_content_type_encoding(content_type_header)
if enc is not None:
if enc in {"utf-16", "utf-32"}:
enc += "-be"
return enc, to_unicode(html_body_str, enc)
enc = html_body_declared_encoding(html_body_str)
if enc is None and (auto_detect_fun is not None):
enc = auto_detect_fun(html_body_str)
if enc is None:
enc = default_encoding
return enc, to_unicode(html_body_str, enc)
|
