]>
Commit | Line | Data |
---|---|---|
1 | #!/usr/bin/env python3 | |
2 | # coding: utf-8 | |
3 | ||
4 | from __future__ import unicode_literals | |
5 | ||
6 | import base64 | |
7 | import binascii | |
8 | import calendar | |
9 | import codecs | |
10 | import collections | |
11 | import contextlib | |
12 | import ctypes | |
13 | import datetime | |
14 | import email.utils | |
15 | import email.header | |
16 | import errno | |
17 | import functools | |
18 | import gzip | |
19 | import hashlib | |
20 | import hmac | |
21 | import importlib.util | |
22 | import io | |
23 | import itertools | |
24 | import json | |
25 | import locale | |
26 | import math | |
27 | import operator | |
28 | import os | |
29 | import platform | |
30 | import random | |
31 | import re | |
32 | import socket | |
33 | import ssl | |
34 | import subprocess | |
35 | import sys | |
36 | import tempfile | |
37 | import time | |
38 | import traceback | |
39 | import xml.etree.ElementTree | |
40 | import zlib | |
41 | import mimetypes | |
42 | ||
43 | from .compat import ( | |
44 | compat_HTMLParseError, | |
45 | compat_HTMLParser, | |
46 | compat_HTTPError, | |
47 | compat_basestring, | |
48 | compat_chr, | |
49 | compat_cookiejar, | |
50 | compat_ctypes_WINFUNCTYPE, | |
51 | compat_etree_fromstring, | |
52 | compat_expanduser, | |
53 | compat_html_entities, | |
54 | compat_html_entities_html5, | |
55 | compat_http_client, | |
56 | compat_integer_types, | |
57 | compat_numeric_types, | |
58 | compat_kwargs, | |
59 | compat_os_name, | |
60 | compat_parse_qs, | |
61 | compat_shlex_split, | |
62 | compat_shlex_quote, | |
63 | compat_str, | |
64 | compat_struct_pack, | |
65 | compat_struct_unpack, | |
66 | compat_urllib_error, | |
67 | compat_urllib_parse, | |
68 | compat_urllib_parse_urlencode, | |
69 | compat_urllib_parse_urlparse, | |
70 | compat_urllib_parse_urlunparse, | |
71 | compat_urllib_parse_quote, | |
72 | compat_urllib_parse_quote_plus, | |
73 | compat_urllib_parse_unquote_plus, | |
74 | compat_urllib_request, | |
75 | compat_urlparse, | |
76 | compat_xpath, | |
77 | ) | |
78 | ||
79 | from .socks import ( | |
80 | ProxyType, | |
81 | sockssocket, | |
82 | ) | |
83 | ||
84 | ||
85 | def register_socks_protocols(): | |
86 | # "Register" SOCKS protocols | |
87 | # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904 | |
88 | # URLs with protocols not in urlparse.uses_netloc are not handled correctly | |
89 | for scheme in ('socks', 'socks4', 'socks4a', 'socks5'): | |
90 | if scheme not in compat_urlparse.uses_netloc: | |
91 | compat_urlparse.uses_netloc.append(scheme) | |
92 | ||
93 | ||
94 | # This is not clearly defined otherwise | |
95 | compiled_regex_type = type(re.compile('')) | |
96 | ||
97 | ||
98 | def random_user_agent(): | |
99 | _USER_AGENT_TPL = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36' | |
100 | _CHROME_VERSIONS = ( | |
101 | '90.0.4430.212', | |
102 | '90.0.4430.24', | |
103 | '90.0.4430.70', | |
104 | '90.0.4430.72', | |
105 | '90.0.4430.85', | |
106 | '90.0.4430.93', | |
107 | '91.0.4472.101', | |
108 | '91.0.4472.106', | |
109 | '91.0.4472.114', | |
110 | '91.0.4472.124', | |
111 | '91.0.4472.164', | |
112 | '91.0.4472.19', | |
113 | '91.0.4472.77', | |
114 | '92.0.4515.107', | |
115 | '92.0.4515.115', | |
116 | '92.0.4515.131', | |
117 | '92.0.4515.159', | |
118 | '92.0.4515.43', | |
119 | '93.0.4556.0', | |
120 | '93.0.4577.15', | |
121 | '93.0.4577.63', | |
122 | '93.0.4577.82', | |
123 | '94.0.4606.41', | |
124 | '94.0.4606.54', | |
125 | '94.0.4606.61', | |
126 | '94.0.4606.71', | |
127 | '94.0.4606.81', | |
128 | '94.0.4606.85', | |
129 | '95.0.4638.17', | |
130 | '95.0.4638.50', | |
131 | '95.0.4638.54', | |
132 | '95.0.4638.69', | |
133 | '95.0.4638.74', | |
134 | '96.0.4664.18', | |
135 | '96.0.4664.45', | |
136 | '96.0.4664.55', | |
137 | '96.0.4664.93', | |
138 | '97.0.4692.20', | |
139 | ) | |
140 | return _USER_AGENT_TPL % random.choice(_CHROME_VERSIONS) | |
141 | ||
142 | ||
143 | std_headers = { | |
144 | 'User-Agent': random_user_agent(), | |
145 | 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8', | |
146 | 'Accept-Encoding': 'gzip, deflate', | |
147 | 'Accept-Language': 'en-us,en;q=0.5', | |
148 | 'Sec-Fetch-Mode': 'navigate', | |
149 | } | |
150 | ||
151 | ||
152 | USER_AGENTS = { | |
153 | 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27', | |
154 | } | |
155 | ||
156 | ||
157 | NO_DEFAULT = object() | |
158 | ||
159 | ENGLISH_MONTH_NAMES = [ | |
160 | 'January', 'February', 'March', 'April', 'May', 'June', | |
161 | 'July', 'August', 'September', 'October', 'November', 'December'] | |
162 | ||
163 | MONTH_NAMES = { | |
164 | 'en': ENGLISH_MONTH_NAMES, | |
165 | 'fr': [ | |
166 | 'janvier', 'février', 'mars', 'avril', 'mai', 'juin', | |
167 | 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'], | |
168 | } | |
169 | ||
170 | KNOWN_EXTENSIONS = ( | |
171 | 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac', | |
172 | 'flv', 'f4v', 'f4a', 'f4b', | |
173 | 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus', | |
174 | 'mkv', 'mka', 'mk3d', | |
175 | 'avi', 'divx', | |
176 | 'mov', | |
177 | 'asf', 'wmv', 'wma', | |
178 | '3gp', '3g2', | |
179 | 'mp3', | |
180 | 'flac', | |
181 | 'ape', | |
182 | 'wav', | |
183 | 'f4f', 'f4m', 'm3u8', 'smil') | |
184 | ||
185 | # needed for sanitizing filenames in restricted mode | |
186 | ACCENT_CHARS = dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ', | |
187 | itertools.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'], | |
188 | 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y'))) | |
189 | ||
190 | DATE_FORMATS = ( | |
191 | '%d %B %Y', | |
192 | '%d %b %Y', | |
193 | '%B %d %Y', | |
194 | '%B %dst %Y', | |
195 | '%B %dnd %Y', | |
196 | '%B %drd %Y', | |
197 | '%B %dth %Y', | |
198 | '%b %d %Y', | |
199 | '%b %dst %Y', | |
200 | '%b %dnd %Y', | |
201 | '%b %drd %Y', | |
202 | '%b %dth %Y', | |
203 | '%b %dst %Y %I:%M', | |
204 | '%b %dnd %Y %I:%M', | |
205 | '%b %drd %Y %I:%M', | |
206 | '%b %dth %Y %I:%M', | |
207 | '%Y %m %d', | |
208 | '%Y-%m-%d', | |
209 | '%Y.%m.%d.', | |
210 | '%Y/%m/%d', | |
211 | '%Y/%m/%d %H:%M', | |
212 | '%Y/%m/%d %H:%M:%S', | |
213 | '%Y%m%d%H%M', | |
214 | '%Y%m%d%H%M%S', | |
215 | '%Y%m%d', | |
216 | '%Y-%m-%d %H:%M', | |
217 | '%Y-%m-%d %H:%M:%S', | |
218 | '%Y-%m-%d %H:%M:%S.%f', | |
219 | '%Y-%m-%d %H:%M:%S:%f', | |
220 | '%d.%m.%Y %H:%M', | |
221 | '%d.%m.%Y %H.%M', | |
222 | '%Y-%m-%dT%H:%M:%SZ', | |
223 | '%Y-%m-%dT%H:%M:%S.%fZ', | |
224 | '%Y-%m-%dT%H:%M:%S.%f0Z', | |
225 | '%Y-%m-%dT%H:%M:%S', | |
226 | '%Y-%m-%dT%H:%M:%S.%f', | |
227 | '%Y-%m-%dT%H:%M', | |
228 | '%b %d %Y at %H:%M', | |
229 | '%b %d %Y at %H:%M:%S', | |
230 | '%B %d %Y at %H:%M', | |
231 | '%B %d %Y at %H:%M:%S', | |
232 | '%H:%M %d-%b-%Y', | |
233 | ) | |
234 | ||
235 | DATE_FORMATS_DAY_FIRST = list(DATE_FORMATS) | |
236 | DATE_FORMATS_DAY_FIRST.extend([ | |
237 | '%d-%m-%Y', | |
238 | '%d.%m.%Y', | |
239 | '%d.%m.%y', | |
240 | '%d/%m/%Y', | |
241 | '%d/%m/%y', | |
242 | '%d/%m/%Y %H:%M:%S', | |
243 | ]) | |
244 | ||
245 | DATE_FORMATS_MONTH_FIRST = list(DATE_FORMATS) | |
246 | DATE_FORMATS_MONTH_FIRST.extend([ | |
247 | '%m-%d-%Y', | |
248 | '%m.%d.%Y', | |
249 | '%m/%d/%Y', | |
250 | '%m/%d/%y', | |
251 | '%m/%d/%Y %H:%M:%S', | |
252 | ]) | |
253 | ||
254 | PACKED_CODES_RE = r"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)" | |
255 | JSON_LD_RE = r'(?is)<script[^>]+type=(["\']?)application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>' | |
256 | ||
257 | ||
258 | def preferredencoding(): | |
259 | """Get preferred encoding. | |
260 | ||
261 | Returns the best encoding scheme for the system, based on | |
262 | locale.getpreferredencoding() and some further tweaks. | |
263 | """ | |
264 | try: | |
265 | pref = locale.getpreferredencoding() | |
266 | 'TEST'.encode(pref) | |
267 | except Exception: | |
268 | pref = 'UTF-8' | |
269 | ||
270 | return pref | |
271 | ||
272 | ||
273 | def write_json_file(obj, fn): | |
274 | """ Encode obj as JSON and write it to fn, atomically if possible """ | |
275 | ||
276 | fn = encodeFilename(fn) | |
277 | if sys.version_info < (3, 0) and sys.platform != 'win32': | |
278 | encoding = get_filesystem_encoding() | |
279 | # os.path.basename returns a bytes object, but NamedTemporaryFile | |
280 | # will fail if the filename contains non ascii characters unless we | |
281 | # use a unicode object | |
282 | path_basename = lambda f: os.path.basename(fn).decode(encoding) | |
283 | # the same for os.path.dirname | |
284 | path_dirname = lambda f: os.path.dirname(fn).decode(encoding) | |
285 | else: | |
286 | path_basename = os.path.basename | |
287 | path_dirname = os.path.dirname | |
288 | ||
289 | args = { | |
290 | 'suffix': '.tmp', | |
291 | 'prefix': path_basename(fn) + '.', | |
292 | 'dir': path_dirname(fn), | |
293 | 'delete': False, | |
294 | } | |
295 | ||
296 | # In Python 2.x, json.dump expects a bytestream. | |
297 | # In Python 3.x, it writes to a character stream | |
298 | if sys.version_info < (3, 0): | |
299 | args['mode'] = 'wb' | |
300 | else: | |
301 | args.update({ | |
302 | 'mode': 'w', | |
303 | 'encoding': 'utf-8', | |
304 | }) | |
305 | ||
306 | tf = tempfile.NamedTemporaryFile(**compat_kwargs(args)) | |
307 | ||
308 | try: | |
309 | with tf: | |
310 | json.dump(obj, tf, ensure_ascii=False) | |
311 | if sys.platform == 'win32': | |
312 | # Need to remove existing file on Windows, else os.rename raises | |
313 | # WindowsError or FileExistsError. | |
314 | try: | |
315 | os.unlink(fn) | |
316 | except OSError: | |
317 | pass | |
318 | try: | |
319 | mask = os.umask(0) | |
320 | os.umask(mask) | |
321 | os.chmod(tf.name, 0o666 & ~mask) | |
322 | except OSError: | |
323 | pass | |
324 | os.rename(tf.name, fn) | |
325 | except Exception: | |
326 | try: | |
327 | os.remove(tf.name) | |
328 | except OSError: | |
329 | pass | |
330 | raise | |
331 | ||
332 | ||
333 | if sys.version_info >= (2, 7): | |
334 | def find_xpath_attr(node, xpath, key, val=None): | |
335 | """ Find the xpath xpath[@key=val] """ | |
336 | assert re.match(r'^[a-zA-Z_-]+$', key) | |
337 | expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val)) | |
338 | return node.find(expr) | |
339 | else: | |
340 | def find_xpath_attr(node, xpath, key, val=None): | |
341 | for f in node.findall(compat_xpath(xpath)): | |
342 | if key not in f.attrib: | |
343 | continue | |
344 | if val is None or f.attrib.get(key) == val: | |
345 | return f | |
346 | return None | |
347 | ||
348 | # On python2.6 the xml.etree.ElementTree.Element methods don't support | |
349 | # the namespace parameter | |
350 | ||
351 | ||
352 | def xpath_with_ns(path, ns_map): | |
353 | components = [c.split(':') for c in path.split('/')] | |
354 | replaced = [] | |
355 | for c in components: | |
356 | if len(c) == 1: | |
357 | replaced.append(c[0]) | |
358 | else: | |
359 | ns, tag = c | |
360 | replaced.append('{%s}%s' % (ns_map[ns], tag)) | |
361 | return '/'.join(replaced) | |
362 | ||
363 | ||
364 | def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT): | |
365 | def _find_xpath(xpath): | |
366 | return node.find(compat_xpath(xpath)) | |
367 | ||
368 | if isinstance(xpath, (str, compat_str)): | |
369 | n = _find_xpath(xpath) | |
370 | else: | |
371 | for xp in xpath: | |
372 | n = _find_xpath(xp) | |
373 | if n is not None: | |
374 | break | |
375 | ||
376 | if n is None: | |
377 | if default is not NO_DEFAULT: | |
378 | return default | |
379 | elif fatal: | |
380 | name = xpath if name is None else name | |
381 | raise ExtractorError('Could not find XML element %s' % name) | |
382 | else: | |
383 | return None | |
384 | return n | |
385 | ||
386 | ||
387 | def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT): | |
388 | n = xpath_element(node, xpath, name, fatal=fatal, default=default) | |
389 | if n is None or n == default: | |
390 | return n | |
391 | if n.text is None: | |
392 | if default is not NO_DEFAULT: | |
393 | return default | |
394 | elif fatal: | |
395 | name = xpath if name is None else name | |
396 | raise ExtractorError('Could not find XML element\'s text %s' % name) | |
397 | else: | |
398 | return None | |
399 | return n.text | |
400 | ||
401 | ||
402 | def xpath_attr(node, xpath, key, name=None, fatal=False, default=NO_DEFAULT): | |
403 | n = find_xpath_attr(node, xpath, key) | |
404 | if n is None: | |
405 | if default is not NO_DEFAULT: | |
406 | return default | |
407 | elif fatal: | |
408 | name = '%s[@%s]' % (xpath, key) if name is None else name | |
409 | raise ExtractorError('Could not find XML attribute %s' % name) | |
410 | else: | |
411 | return None | |
412 | return n.attrib[key] | |
413 | ||
414 | ||
415 | def get_element_by_id(id, html): | |
416 | """Return the content of the tag with the specified ID in the passed HTML document""" | |
417 | return get_element_by_attribute('id', id, html) | |
418 | ||
419 | ||
420 | def get_element_html_by_id(id, html): | |
421 | """Return the html of the tag with the specified ID in the passed HTML document""" | |
422 | return get_element_html_by_attribute('id', id, html) | |
423 | ||
424 | ||
425 | def get_element_by_class(class_name, html): | |
426 | """Return the content of the first tag with the specified class in the passed HTML document""" | |
427 | retval = get_elements_by_class(class_name, html) | |
428 | return retval[0] if retval else None | |
429 | ||
430 | ||
431 | def get_element_html_by_class(class_name, html): | |
432 | """Return the html of the first tag with the specified class in the passed HTML document""" | |
433 | retval = get_elements_html_by_class(class_name, html) | |
434 | return retval[0] if retval else None | |
435 | ||
436 | ||
437 | def get_element_by_attribute(attribute, value, html, escape_value=True): | |
438 | retval = get_elements_by_attribute(attribute, value, html, escape_value) | |
439 | return retval[0] if retval else None | |
440 | ||
441 | ||
442 | def get_element_html_by_attribute(attribute, value, html, escape_value=True): | |
443 | retval = get_elements_html_by_attribute(attribute, value, html, escape_value) | |
444 | return retval[0] if retval else None | |
445 | ||
446 | ||
447 | def get_elements_by_class(class_name, html): | |
448 | """Return the content of all tags with the specified class in the passed HTML document as a list""" | |
449 | return get_elements_by_attribute( | |
450 | 'class', r'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name), | |
451 | html, escape_value=False) | |
452 | ||
453 | ||
454 | def get_elements_html_by_class(class_name, html): | |
455 | """Return the html of all tags with the specified class in the passed HTML document as a list""" | |
456 | return get_elements_html_by_attribute( | |
457 | 'class', r'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name), | |
458 | html, escape_value=False) | |
459 | ||
460 | ||
461 | def get_elements_by_attribute(*args, **kwargs): | |
462 | """Return the content of the tag with the specified attribute in the passed HTML document""" | |
463 | return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)] | |
464 | ||
465 | ||
466 | def get_elements_html_by_attribute(*args, **kwargs): | |
467 | """Return the html of the tag with the specified attribute in the passed HTML document""" | |
468 | return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)] | |
469 | ||
470 | ||
471 | def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True): | |
472 | """ | |
473 | Return the text (content) and the html (whole) of the tag with the specified | |
474 | attribute in the passed HTML document | |
475 | """ | |
476 | ||
477 | value_quote_optional = '' if re.match(r'''[\s"'`=<>]''', value) else '?' | |
478 | ||
479 | value = re.escape(value) if escape_value else value | |
480 | ||
481 | partial_element_re = r'''(?x) | |
482 | <(?P<tag>[a-zA-Z0-9:._-]+) | |
483 | (?:\s(?:[^>"']|"[^"]*"|'[^']*')*)? | |
484 | \s%(attribute)s\s*=\s*(?P<_q>['"]%(vqo)s)(?-x:%(value)s)(?P=_q) | |
485 | ''' % {'attribute': re.escape(attribute), 'value': value, 'vqo': value_quote_optional} | |
486 | ||
487 | for m in re.finditer(partial_element_re, html): | |
488 | content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():]) | |
489 | ||
490 | yield ( | |
491 | unescapeHTML(re.sub(r'^(?P<q>["\'])(?P<content>.*)(?P=q)$', r'\g<content>', content, flags=re.DOTALL)), | |
492 | whole | |
493 | ) | |
494 | ||
495 | ||
496 | class HTMLBreakOnClosingTagParser(compat_HTMLParser): | |
497 | """ | |
498 | HTML parser which raises HTMLBreakOnClosingTagException upon reaching the | |
499 | closing tag for the first opening tag it has encountered, and can be used | |
500 | as a context manager | |
501 | """ | |
502 | ||
503 | class HTMLBreakOnClosingTagException(Exception): | |
504 | pass | |
505 | ||
506 | def __init__(self): | |
507 | self.tagstack = collections.deque() | |
508 | compat_HTMLParser.__init__(self) | |
509 | ||
510 | def __enter__(self): | |
511 | return self | |
512 | ||
513 | def __exit__(self, *_): | |
514 | self.close() | |
515 | ||
516 | def close(self): | |
517 | # handle_endtag does not return upon raising HTMLBreakOnClosingTagException, | |
518 | # so data remains buffered; we no longer have any interest in it, thus | |
519 | # override this method to discard it | |
520 | pass | |
521 | ||
522 | def handle_starttag(self, tag, _): | |
523 | self.tagstack.append(tag) | |
524 | ||
525 | def handle_endtag(self, tag): | |
526 | if not self.tagstack: | |
527 | raise compat_HTMLParseError('no tags in the stack') | |
528 | while self.tagstack: | |
529 | inner_tag = self.tagstack.pop() | |
530 | if inner_tag == tag: | |
531 | break | |
532 | else: | |
533 | raise compat_HTMLParseError(f'matching opening tag for closing {tag} tag not found') | |
534 | if not self.tagstack: | |
535 | raise self.HTMLBreakOnClosingTagException() | |
536 | ||
537 | ||
538 | def get_element_text_and_html_by_tag(tag, html): | |
539 | """ | |
540 | For the first element with the specified tag in the passed HTML document | |
541 | return its' content (text) and the whole element (html) | |
542 | """ | |
543 | def find_or_raise(haystack, needle, exc): | |
544 | try: | |
545 | return haystack.index(needle) | |
546 | except ValueError: | |
547 | raise exc | |
548 | closing_tag = f'</{tag}>' | |
549 | whole_start = find_or_raise( | |
550 | html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found')) | |
551 | content_start = find_or_raise( | |
552 | html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag')) | |
553 | content_start += whole_start + 1 | |
554 | with HTMLBreakOnClosingTagParser() as parser: | |
555 | parser.feed(html[whole_start:content_start]) | |
556 | if not parser.tagstack or parser.tagstack[0] != tag: | |
557 | raise compat_HTMLParseError(f'parser did not match opening {tag} tag') | |
558 | offset = content_start | |
559 | while offset < len(html): | |
560 | next_closing_tag_start = find_or_raise( | |
561 | html[offset:], closing_tag, | |
562 | compat_HTMLParseError(f'closing {tag} tag not found')) | |
563 | next_closing_tag_end = next_closing_tag_start + len(closing_tag) | |
564 | try: | |
565 | parser.feed(html[offset:offset + next_closing_tag_end]) | |
566 | offset += next_closing_tag_end | |
567 | except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException: | |
568 | return html[content_start:offset + next_closing_tag_start], \ | |
569 | html[whole_start:offset + next_closing_tag_end] | |
570 | raise compat_HTMLParseError('unexpected end of html') | |
571 | ||
572 | ||
573 | class HTMLAttributeParser(compat_HTMLParser): | |
574 | """Trivial HTML parser to gather the attributes for a single element""" | |
575 | ||
576 | def __init__(self): | |
577 | self.attrs = {} | |
578 | compat_HTMLParser.__init__(self) | |
579 | ||
580 | def handle_starttag(self, tag, attrs): | |
581 | self.attrs = dict(attrs) | |
582 | ||
583 | ||
584 | class HTMLListAttrsParser(compat_HTMLParser): | |
585 | """HTML parser to gather the attributes for the elements of a list""" | |
586 | ||
587 | def __init__(self): | |
588 | compat_HTMLParser.__init__(self) | |
589 | self.items = [] | |
590 | self._level = 0 | |
591 | ||
592 | def handle_starttag(self, tag, attrs): | |
593 | if tag == 'li' and self._level == 0: | |
594 | self.items.append(dict(attrs)) | |
595 | self._level += 1 | |
596 | ||
597 | def handle_endtag(self, tag): | |
598 | self._level -= 1 | |
599 | ||
600 | ||
601 | def extract_attributes(html_element): | |
602 | """Given a string for an HTML element such as | |
603 | <el | |
604 | a="foo" B="bar" c="&98;az" d=boz | |
605 | empty= noval entity="&" | |
606 | sq='"' dq="'" | |
607 | > | |
608 | Decode and return a dictionary of attributes. | |
609 | { | |
610 | 'a': 'foo', 'b': 'bar', c: 'baz', d: 'boz', | |
611 | 'empty': '', 'noval': None, 'entity': '&', | |
612 | 'sq': '"', 'dq': '\'' | |
613 | }. | |
614 | NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions, | |
615 | but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5. | |
616 | """ | |
617 | parser = HTMLAttributeParser() | |
618 | try: | |
619 | parser.feed(html_element) | |
620 | parser.close() | |
621 | # Older Python may throw HTMLParseError in case of malformed HTML | |
622 | except compat_HTMLParseError: | |
623 | pass | |
624 | return parser.attrs | |
625 | ||
626 | ||
627 | def parse_list(webpage): | |
628 | """Given a string for an series of HTML <li> elements, | |
629 | return a dictionary of their attributes""" | |
630 | parser = HTMLListAttrsParser() | |
631 | parser.feed(webpage) | |
632 | parser.close() | |
633 | return parser.items | |
634 | ||
635 | ||
636 | def clean_html(html): | |
637 | """Clean an HTML snippet into a readable string""" | |
638 | ||
639 | if html is None: # Convenience for sanitizing descriptions etc. | |
640 | return html | |
641 | ||
642 | html = re.sub(r'\s+', ' ', html) | |
643 | html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html) | |
644 | html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html) | |
645 | # Strip html tags | |
646 | html = re.sub('<.*?>', '', html) | |
647 | # Replace html entities | |
648 | html = unescapeHTML(html) | |
649 | return html.strip() | |
650 | ||
651 | ||
652 | def sanitize_open(filename, open_mode): | |
653 | """Try to open the given filename, and slightly tweak it if this fails. | |
654 | ||
655 | Attempts to open the given filename. If this fails, it tries to change | |
656 | the filename slightly, step by step, until it's either able to open it | |
657 | or it fails and raises a final exception, like the standard open() | |
658 | function. | |
659 | ||
660 | It returns the tuple (stream, definitive_file_name). | |
661 | """ | |
662 | try: | |
663 | if filename == '-': | |
664 | if sys.platform == 'win32': | |
665 | import msvcrt | |
666 | msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY) | |
667 | return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename) | |
668 | stream = locked_file(filename, open_mode, block=False).open() | |
669 | return (stream, filename) | |
670 | except (IOError, OSError) as err: | |
671 | if err.errno in (errno.EACCES,): | |
672 | raise | |
673 | ||
674 | # In case of error, try to remove win32 forbidden chars | |
675 | alt_filename = sanitize_path(filename) | |
676 | if alt_filename == filename: | |
677 | raise | |
678 | else: | |
679 | # An exception here should be caught in the caller | |
680 | stream = locked_file(filename, open_mode, block=False).open() | |
681 | return (stream, alt_filename) | |
682 | ||
683 | ||
684 | def timeconvert(timestr): | |
685 | """Convert RFC 2822 defined time string into system timestamp""" | |
686 | timestamp = None | |
687 | timetuple = email.utils.parsedate_tz(timestr) | |
688 | if timetuple is not None: | |
689 | timestamp = email.utils.mktime_tz(timetuple) | |
690 | return timestamp | |
691 | ||
692 | ||
693 | def sanitize_filename(s, restricted=False, is_id=False): | |
694 | """Sanitizes a string so it could be used as part of a filename. | |
695 | If restricted is set, use a stricter subset of allowed characters. | |
696 | Set is_id if this is not an arbitrary string, but an ID that should be kept | |
697 | if possible. | |
698 | """ | |
699 | def replace_insane(char): | |
700 | if restricted and char in ACCENT_CHARS: | |
701 | return ACCENT_CHARS[char] | |
702 | elif not restricted and char == '\n': | |
703 | return ' ' | |
704 | elif char == '?' or ord(char) < 32 or ord(char) == 127: | |
705 | return '' | |
706 | elif char == '"': | |
707 | return '' if restricted else '\'' | |
708 | elif char == ':': | |
709 | return '_-' if restricted else ' -' | |
710 | elif char in '\\/|*<>': | |
711 | return '_' | |
712 | if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()): | |
713 | return '_' | |
714 | if restricted and ord(char) > 127: | |
715 | return '_' | |
716 | return char | |
717 | ||
718 | if s == '': | |
719 | return '' | |
720 | # Handle timestamps | |
721 | s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s) | |
722 | result = ''.join(map(replace_insane, s)) | |
723 | if not is_id: | |
724 | while '__' in result: | |
725 | result = result.replace('__', '_') | |
726 | result = result.strip('_') | |
727 | # Common case of "Foreign band name - English song title" | |
728 | if restricted and result.startswith('-_'): | |
729 | result = result[2:] | |
730 | if result.startswith('-'): | |
731 | result = '_' + result[len('-'):] | |
732 | result = result.lstrip('.') | |
733 | if not result: | |
734 | result = '_' | |
735 | return result | |
736 | ||
737 | ||
738 | def sanitize_path(s, force=False): | |
739 | """Sanitizes and normalizes path on Windows""" | |
740 | if sys.platform == 'win32': | |
741 | force = False | |
742 | drive_or_unc, _ = os.path.splitdrive(s) | |
743 | if sys.version_info < (2, 7) and not drive_or_unc: | |
744 | drive_or_unc, _ = os.path.splitunc(s) | |
745 | elif force: | |
746 | drive_or_unc = '' | |
747 | else: | |
748 | return s | |
749 | ||
750 | norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep) | |
751 | if drive_or_unc: | |
752 | norm_path.pop(0) | |
753 | sanitized_path = [ | |
754 | path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part) | |
755 | for path_part in norm_path] | |
756 | if drive_or_unc: | |
757 | sanitized_path.insert(0, drive_or_unc + os.path.sep) | |
758 | elif force and s[0] == os.path.sep: | |
759 | sanitized_path.insert(0, os.path.sep) | |
760 | return os.path.join(*sanitized_path) | |
761 | ||
762 | ||
763 | def sanitize_url(url): | |
764 | # Prepend protocol-less URLs with `http:` scheme in order to mitigate | |
765 | # the number of unwanted failures due to missing protocol | |
766 | if url.startswith('//'): | |
767 | return 'http:%s' % url | |
768 | # Fix some common typos seen so far | |
769 | COMMON_TYPOS = ( | |
770 | # https://github.com/ytdl-org/youtube-dl/issues/15649 | |
771 | (r'^httpss://', r'https://'), | |
772 | # https://bx1.be/lives/direct-tv/ | |
773 | (r'^rmtp([es]?)://', r'rtmp\1://'), | |
774 | ) | |
775 | for mistake, fixup in COMMON_TYPOS: | |
776 | if re.match(mistake, url): | |
777 | return re.sub(mistake, fixup, url) | |
778 | return url | |
779 | ||
780 | ||
781 | def extract_basic_auth(url): | |
782 | parts = compat_urlparse.urlsplit(url) | |
783 | if parts.username is None: | |
784 | return url, None | |
785 | url = compat_urlparse.urlunsplit(parts._replace(netloc=( | |
786 | parts.hostname if parts.port is None | |
787 | else '%s:%d' % (parts.hostname, parts.port)))) | |
788 | auth_payload = base64.b64encode( | |
789 | ('%s:%s' % (parts.username, parts.password or '')).encode('utf-8')) | |
790 | return url, 'Basic ' + auth_payload.decode('utf-8') | |
791 | ||
792 | ||
793 | def sanitized_Request(url, *args, **kwargs): | |
794 | url, auth_header = extract_basic_auth(escape_url(sanitize_url(url))) | |
795 | if auth_header is not None: | |
796 | headers = args[1] if len(args) >= 2 else kwargs.setdefault('headers', {}) | |
797 | headers['Authorization'] = auth_header | |
798 | return compat_urllib_request.Request(url, *args, **kwargs) | |
799 | ||
800 | ||
801 | def expand_path(s): | |
802 | """Expand shell variables and ~""" | |
803 | return os.path.expandvars(compat_expanduser(s)) | |
804 | ||
805 | ||
806 | def orderedSet(iterable): | |
807 | """ Remove all duplicates from the input iterable """ | |
808 | res = [] | |
809 | for el in iterable: | |
810 | if el not in res: | |
811 | res.append(el) | |
812 | return res | |
813 | ||
814 | ||
815 | def _htmlentity_transform(entity_with_semicolon): | |
816 | """Transforms an HTML entity to a character.""" | |
817 | entity = entity_with_semicolon[:-1] | |
818 | ||
819 | # Known non-numeric HTML entity | |
820 | if entity in compat_html_entities.name2codepoint: | |
821 | return compat_chr(compat_html_entities.name2codepoint[entity]) | |
822 | ||
823 | # TODO: HTML5 allows entities without a semicolon. For example, | |
824 | # 'Éric' should be decoded as 'Éric'. | |
825 | if entity_with_semicolon in compat_html_entities_html5: | |
826 | return compat_html_entities_html5[entity_with_semicolon] | |
827 | ||
828 | mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity) | |
829 | if mobj is not None: | |
830 | numstr = mobj.group(1) | |
831 | if numstr.startswith('x'): | |
832 | base = 16 | |
833 | numstr = '0%s' % numstr | |
834 | else: | |
835 | base = 10 | |
836 | # See https://github.com/ytdl-org/youtube-dl/issues/7518 | |
837 | try: | |
838 | return compat_chr(int(numstr, base)) | |
839 | except ValueError: | |
840 | pass | |
841 | ||
842 | # Unknown entity in name, return its literal representation | |
843 | return '&%s;' % entity | |
844 | ||
845 | ||
846 | def unescapeHTML(s): | |
847 | if s is None: | |
848 | return None | |
849 | assert type(s) == compat_str | |
850 | ||
851 | return re.sub( | |
852 | r'&([^&;]+;)', lambda m: _htmlentity_transform(m.group(1)), s) | |
853 | ||
854 | ||
855 | def escapeHTML(text): | |
856 | return ( | |
857 | text | |
858 | .replace('&', '&') | |
859 | .replace('<', '<') | |
860 | .replace('>', '>') | |
861 | .replace('"', '"') | |
862 | .replace("'", ''') | |
863 | ) | |
864 | ||
865 | ||
866 | def process_communicate_or_kill(p, *args, **kwargs): | |
867 | try: | |
868 | return p.communicate(*args, **kwargs) | |
869 | except BaseException: # Including KeyboardInterrupt | |
870 | p.kill() | |
871 | p.wait() | |
872 | raise | |
873 | ||
874 | ||
875 | class Popen(subprocess.Popen): | |
876 | if sys.platform == 'win32': | |
877 | _startupinfo = subprocess.STARTUPINFO() | |
878 | _startupinfo.dwFlags |= subprocess.STARTF_USESHOWWINDOW | |
879 | else: | |
880 | _startupinfo = None | |
881 | ||
882 | def __init__(self, *args, **kwargs): | |
883 | super(Popen, self).__init__(*args, **kwargs, startupinfo=self._startupinfo) | |
884 | ||
885 | def communicate_or_kill(self, *args, **kwargs): | |
886 | return process_communicate_or_kill(self, *args, **kwargs) | |
887 | ||
888 | ||
889 | def get_subprocess_encoding(): | |
890 | if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5: | |
891 | # For subprocess calls, encode with locale encoding | |
892 | # Refer to http://stackoverflow.com/a/9951851/35070 | |
893 | encoding = preferredencoding() | |
894 | else: | |
895 | encoding = sys.getfilesystemencoding() | |
896 | if encoding is None: | |
897 | encoding = 'utf-8' | |
898 | return encoding | |
899 | ||
900 | ||
901 | def encodeFilename(s, for_subprocess=False): | |
902 | """ | |
903 | @param s The name of the file | |
904 | """ | |
905 | ||
906 | assert type(s) == compat_str | |
907 | ||
908 | # Python 3 has a Unicode API | |
909 | if sys.version_info >= (3, 0): | |
910 | return s | |
911 | ||
912 | # Pass '' directly to use Unicode APIs on Windows 2000 and up | |
913 | # (Detecting Windows NT 4 is tricky because 'major >= 4' would | |
914 | # match Windows 9x series as well. Besides, NT 4 is obsolete.) | |
915 | if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5: | |
916 | return s | |
917 | ||
918 | # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible | |
919 | if sys.platform.startswith('java'): | |
920 | return s | |
921 | ||
922 | return s.encode(get_subprocess_encoding(), 'ignore') | |
923 | ||
924 | ||
925 | def decodeFilename(b, for_subprocess=False): | |
926 | ||
927 | if sys.version_info >= (3, 0): | |
928 | return b | |
929 | ||
930 | if not isinstance(b, bytes): | |
931 | return b | |
932 | ||
933 | return b.decode(get_subprocess_encoding(), 'ignore') | |
934 | ||
935 | ||
936 | def encodeArgument(s): | |
937 | if not isinstance(s, compat_str): | |
938 | # Legacy code that uses byte strings | |
939 | # Uncomment the following line after fixing all post processors | |
940 | # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s)) | |
941 | s = s.decode('ascii') | |
942 | return encodeFilename(s, True) | |
943 | ||
944 | ||
945 | def decodeArgument(b): | |
946 | return decodeFilename(b, True) | |
947 | ||
948 | ||
949 | def decodeOption(optval): | |
950 | if optval is None: | |
951 | return optval | |
952 | if isinstance(optval, bytes): | |
953 | optval = optval.decode(preferredencoding()) | |
954 | ||
955 | assert isinstance(optval, compat_str) | |
956 | return optval | |
957 | ||
958 | ||
959 | _timetuple = collections.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds')) | |
960 | ||
961 | ||
962 | def timetuple_from_msec(msec): | |
963 | secs, msec = divmod(msec, 1000) | |
964 | mins, secs = divmod(secs, 60) | |
965 | hrs, mins = divmod(mins, 60) | |
966 | return _timetuple(hrs, mins, secs, msec) | |
967 | ||
968 | ||
969 | def formatSeconds(secs, delim=':', msec=False): | |
970 | time = timetuple_from_msec(secs * 1000) | |
971 | if time.hours: | |
972 | ret = '%d%s%02d%s%02d' % (time.hours, delim, time.minutes, delim, time.seconds) | |
973 | elif time.minutes: | |
974 | ret = '%d%s%02d' % (time.minutes, delim, time.seconds) | |
975 | else: | |
976 | ret = '%d' % time.seconds | |
977 | return '%s.%03d' % (ret, time.milliseconds) if msec else ret | |
978 | ||
979 | ||
980 | def _ssl_load_windows_store_certs(ssl_context, storename): | |
981 | # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py | |
982 | try: | |
983 | certs = [cert for cert, encoding, trust in ssl.enum_certificates(storename) | |
984 | if encoding == 'x509_asn' and ( | |
985 | trust is True or ssl.Purpose.SERVER_AUTH.oid in trust)] | |
986 | except PermissionError: | |
987 | return | |
988 | for cert in certs: | |
989 | try: | |
990 | ssl_context.load_verify_locations(cadata=cert) | |
991 | except ssl.SSLError: | |
992 | pass | |
993 | ||
994 | ||
995 | def make_HTTPS_handler(params, **kwargs): | |
996 | opts_check_certificate = not params.get('nocheckcertificate') | |
997 | context = ssl.SSLContext(ssl.PROTOCOL_TLS_CLIENT) | |
998 | context.check_hostname = opts_check_certificate | |
999 | if params.get('legacyserverconnect'): | |
1000 | context.options |= 4 # SSL_OP_LEGACY_SERVER_CONNECT | |
1001 | context.verify_mode = ssl.CERT_REQUIRED if opts_check_certificate else ssl.CERT_NONE | |
1002 | if opts_check_certificate: | |
1003 | try: | |
1004 | context.load_default_certs() | |
1005 | # Work around the issue in load_default_certs when there are bad certificates. See: | |
1006 | # https://github.com/yt-dlp/yt-dlp/issues/1060, | |
1007 | # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312 | |
1008 | except ssl.SSLError: | |
1009 | # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151 | |
1010 | if sys.platform == 'win32' and hasattr(ssl, 'enum_certificates'): | |
1011 | # Create a new context to discard any certificates that were already loaded | |
1012 | context = ssl.SSLContext(ssl.PROTOCOL_TLS_CLIENT) | |
1013 | context.check_hostname, context.verify_mode = True, ssl.CERT_REQUIRED | |
1014 | for storename in ('CA', 'ROOT'): | |
1015 | _ssl_load_windows_store_certs(context, storename) | |
1016 | context.set_default_verify_paths() | |
1017 | return YoutubeDLHTTPSHandler(params, context=context, **kwargs) | |
1018 | ||
1019 | ||
1020 | def bug_reports_message(before=';'): | |
1021 | msg = ('please report this issue on https://github.com/yt-dlp/yt-dlp , ' | |
1022 | 'filling out the "Broken site" issue template properly. ' | |
1023 | 'Confirm you are on the latest version using -U') | |
1024 | ||
1025 | before = before.rstrip() | |
1026 | if not before or before.endswith(('.', '!', '?')): | |
1027 | msg = msg[0].title() + msg[1:] | |
1028 | ||
1029 | return (before + ' ' if before else '') + msg | |
1030 | ||
1031 | ||
1032 | class YoutubeDLError(Exception): | |
1033 | """Base exception for YoutubeDL errors.""" | |
1034 | msg = None | |
1035 | ||
1036 | def __init__(self, msg=None): | |
1037 | if msg is not None: | |
1038 | self.msg = msg | |
1039 | elif self.msg is None: | |
1040 | self.msg = type(self).__name__ | |
1041 | super().__init__(self.msg) | |
1042 | ||
1043 | ||
1044 | network_exceptions = [compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error] | |
1045 | if hasattr(ssl, 'CertificateError'): | |
1046 | network_exceptions.append(ssl.CertificateError) | |
1047 | network_exceptions = tuple(network_exceptions) | |
1048 | ||
1049 | ||
1050 | class ExtractorError(YoutubeDLError): | |
1051 | """Error during info extraction.""" | |
1052 | ||
1053 | def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None, ie=None): | |
1054 | """ tb, if given, is the original traceback (so that it can be printed out). | |
1055 | If expected is set, this is a normal error message and most likely not a bug in yt-dlp. | |
1056 | """ | |
1057 | if sys.exc_info()[0] in network_exceptions: | |
1058 | expected = True | |
1059 | ||
1060 | self.msg = str(msg) | |
1061 | self.traceback = tb | |
1062 | self.expected = expected | |
1063 | self.cause = cause | |
1064 | self.video_id = video_id | |
1065 | self.ie = ie | |
1066 | self.exc_info = sys.exc_info() # preserve original exception | |
1067 | ||
1068 | super(ExtractorError, self).__init__(''.join(( | |
1069 | format_field(ie, template='[%s] '), | |
1070 | format_field(video_id, template='%s: '), | |
1071 | self.msg, | |
1072 | format_field(cause, template=' (caused by %r)'), | |
1073 | '' if expected else bug_reports_message()))) | |
1074 | ||
1075 | def format_traceback(self): | |
1076 | if self.traceback is None: | |
1077 | return None | |
1078 | return ''.join(traceback.format_tb(self.traceback)) | |
1079 | ||
1080 | ||
1081 | class UnsupportedError(ExtractorError): | |
1082 | def __init__(self, url): | |
1083 | super(UnsupportedError, self).__init__( | |
1084 | 'Unsupported URL: %s' % url, expected=True) | |
1085 | self.url = url | |
1086 | ||
1087 | ||
1088 | class RegexNotFoundError(ExtractorError): | |
1089 | """Error when a regex didn't match""" | |
1090 | pass | |
1091 | ||
1092 | ||
1093 | class GeoRestrictedError(ExtractorError): | |
1094 | """Geographic restriction Error exception. | |
1095 | ||
1096 | This exception may be thrown when a video is not available from your | |
1097 | geographic location due to geographic restrictions imposed by a website. | |
1098 | """ | |
1099 | ||
1100 | def __init__(self, msg, countries=None, **kwargs): | |
1101 | kwargs['expected'] = True | |
1102 | super(GeoRestrictedError, self).__init__(msg, **kwargs) | |
1103 | self.countries = countries | |
1104 | ||
1105 | ||
1106 | class DownloadError(YoutubeDLError): | |
1107 | """Download Error exception. | |
1108 | ||
1109 | This exception may be thrown by FileDownloader objects if they are not | |
1110 | configured to continue on errors. They will contain the appropriate | |
1111 | error message. | |
1112 | """ | |
1113 | ||
1114 | def __init__(self, msg, exc_info=None): | |
1115 | """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """ | |
1116 | super(DownloadError, self).__init__(msg) | |
1117 | self.exc_info = exc_info | |
1118 | ||
1119 | ||
1120 | class EntryNotInPlaylist(YoutubeDLError): | |
1121 | """Entry not in playlist exception. | |
1122 | ||
1123 | This exception will be thrown by YoutubeDL when a requested entry | |
1124 | is not found in the playlist info_dict | |
1125 | """ | |
1126 | msg = 'Entry not found in info' | |
1127 | ||
1128 | ||
1129 | class SameFileError(YoutubeDLError): | |
1130 | """Same File exception. | |
1131 | ||
1132 | This exception will be thrown by FileDownloader objects if they detect | |
1133 | multiple files would have to be downloaded to the same file on disk. | |
1134 | """ | |
1135 | msg = 'Fixed output name but more than one file to download' | |
1136 | ||
1137 | def __init__(self, filename=None): | |
1138 | if filename is not None: | |
1139 | self.msg += f': {filename}' | |
1140 | super().__init__(self.msg) | |
1141 | ||
1142 | ||
1143 | class PostProcessingError(YoutubeDLError): | |
1144 | """Post Processing exception. | |
1145 | ||
1146 | This exception may be raised by PostProcessor's .run() method to | |
1147 | indicate an error in the postprocessing task. | |
1148 | """ | |
1149 | ||
1150 | ||
1151 | class DownloadCancelled(YoutubeDLError): | |
1152 | """ Exception raised when the download queue should be interrupted """ | |
1153 | msg = 'The download was cancelled' | |
1154 | ||
1155 | ||
1156 | class ExistingVideoReached(DownloadCancelled): | |
1157 | """ --break-on-existing triggered """ | |
1158 | msg = 'Encountered a video that is already in the archive, stopping due to --break-on-existing' | |
1159 | ||
1160 | ||
1161 | class RejectedVideoReached(DownloadCancelled): | |
1162 | """ --break-on-reject triggered """ | |
1163 | msg = 'Encountered a video that did not match filter, stopping due to --break-on-reject' | |
1164 | ||
1165 | ||
1166 | class MaxDownloadsReached(DownloadCancelled): | |
1167 | """ --max-downloads limit has been reached. """ | |
1168 | msg = 'Maximum number of downloads reached, stopping due to --max-downloads' | |
1169 | ||
1170 | ||
1171 | class ReExtractInfo(YoutubeDLError): | |
1172 | """ Video info needs to be re-extracted. """ | |
1173 | ||
1174 | def __init__(self, msg, expected=False): | |
1175 | super().__init__(msg) | |
1176 | self.expected = expected | |
1177 | ||
1178 | ||
1179 | class ThrottledDownload(ReExtractInfo): | |
1180 | """ Download speed below --throttled-rate. """ | |
1181 | msg = 'The download speed is below throttle limit' | |
1182 | ||
1183 | def __init__(self): | |
1184 | super().__init__(self.msg, expected=False) | |
1185 | ||
1186 | ||
1187 | class UnavailableVideoError(YoutubeDLError): | |
1188 | """Unavailable Format exception. | |
1189 | ||
1190 | This exception will be thrown when a video is requested | |
1191 | in a format that is not available for that video. | |
1192 | """ | |
1193 | msg = 'Unable to download video' | |
1194 | ||
1195 | def __init__(self, err=None): | |
1196 | if err is not None: | |
1197 | self.msg += f': {err}' | |
1198 | super().__init__(self.msg) | |
1199 | ||
1200 | ||
1201 | class ContentTooShortError(YoutubeDLError): | |
1202 | """Content Too Short exception. | |
1203 | ||
1204 | This exception may be raised by FileDownloader objects when a file they | |
1205 | download is too small for what the server announced first, indicating | |
1206 | the connection was probably interrupted. | |
1207 | """ | |
1208 | ||
1209 | def __init__(self, downloaded, expected): | |
1210 | super(ContentTooShortError, self).__init__( | |
1211 | 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded, expected) | |
1212 | ) | |
1213 | # Both in bytes | |
1214 | self.downloaded = downloaded | |
1215 | self.expected = expected | |
1216 | ||
1217 | ||
1218 | class XAttrMetadataError(YoutubeDLError): | |
1219 | def __init__(self, code=None, msg='Unknown error'): | |
1220 | super(XAttrMetadataError, self).__init__(msg) | |
1221 | self.code = code | |
1222 | self.msg = msg | |
1223 | ||
1224 | # Parsing code and msg | |
1225 | if (self.code in (errno.ENOSPC, errno.EDQUOT) | |
1226 | or 'No space left' in self.msg or 'Disk quota exceeded' in self.msg): | |
1227 | self.reason = 'NO_SPACE' | |
1228 | elif self.code == errno.E2BIG or 'Argument list too long' in self.msg: | |
1229 | self.reason = 'VALUE_TOO_LONG' | |
1230 | else: | |
1231 | self.reason = 'NOT_SUPPORTED' | |
1232 | ||
1233 | ||
1234 | class XAttrUnavailableError(YoutubeDLError): | |
1235 | pass | |
1236 | ||
1237 | ||
1238 | def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs): | |
1239 | # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting | |
1240 | # expected HTTP responses to meet HTTP/1.0 or later (see also | |
1241 | # https://github.com/ytdl-org/youtube-dl/issues/6727) | |
1242 | if sys.version_info < (3, 0): | |
1243 | kwargs['strict'] = True | |
1244 | hc = http_class(*args, **compat_kwargs(kwargs)) | |
1245 | source_address = ydl_handler._params.get('source_address') | |
1246 | ||
1247 | if source_address is not None: | |
1248 | # This is to workaround _create_connection() from socket where it will try all | |
1249 | # address data from getaddrinfo() including IPv6. This filters the result from | |
1250 | # getaddrinfo() based on the source_address value. | |
1251 | # This is based on the cpython socket.create_connection() function. | |
1252 | # https://github.com/python/cpython/blob/master/Lib/socket.py#L691 | |
1253 | def _create_connection(address, timeout=socket._GLOBAL_DEFAULT_TIMEOUT, source_address=None): | |
1254 | host, port = address | |
1255 | err = None | |
1256 | addrs = socket.getaddrinfo(host, port, 0, socket.SOCK_STREAM) | |
1257 | af = socket.AF_INET if '.' in source_address[0] else socket.AF_INET6 | |
1258 | ip_addrs = [addr for addr in addrs if addr[0] == af] | |
1259 | if addrs and not ip_addrs: | |
1260 | ip_version = 'v4' if af == socket.AF_INET else 'v6' | |
1261 | raise socket.error( | |
1262 | "No remote IP%s addresses available for connect, can't use '%s' as source address" | |
1263 | % (ip_version, source_address[0])) | |
1264 | for res in ip_addrs: | |
1265 | af, socktype, proto, canonname, sa = res | |
1266 | sock = None | |
1267 | try: | |
1268 | sock = socket.socket(af, socktype, proto) | |
1269 | if timeout is not socket._GLOBAL_DEFAULT_TIMEOUT: | |
1270 | sock.settimeout(timeout) | |
1271 | sock.bind(source_address) | |
1272 | sock.connect(sa) | |
1273 | err = None # Explicitly break reference cycle | |
1274 | return sock | |
1275 | except socket.error as _: | |
1276 | err = _ | |
1277 | if sock is not None: | |
1278 | sock.close() | |
1279 | if err is not None: | |
1280 | raise err | |
1281 | else: | |
1282 | raise socket.error('getaddrinfo returns an empty list') | |
1283 | if hasattr(hc, '_create_connection'): | |
1284 | hc._create_connection = _create_connection | |
1285 | sa = (source_address, 0) | |
1286 | if hasattr(hc, 'source_address'): # Python 2.7+ | |
1287 | hc.source_address = sa | |
1288 | else: # Python 2.6 | |
1289 | def _hc_connect(self, *args, **kwargs): | |
1290 | sock = _create_connection( | |
1291 | (self.host, self.port), self.timeout, sa) | |
1292 | if is_https: | |
1293 | self.sock = ssl.wrap_socket( | |
1294 | sock, self.key_file, self.cert_file, | |
1295 | ssl_version=ssl.PROTOCOL_TLSv1) | |
1296 | else: | |
1297 | self.sock = sock | |
1298 | hc.connect = functools.partial(_hc_connect, hc) | |
1299 | ||
1300 | return hc | |
1301 | ||
1302 | ||
1303 | def handle_youtubedl_headers(headers): | |
1304 | filtered_headers = headers | |
1305 | ||
1306 | if 'Youtubedl-no-compression' in filtered_headers: | |
1307 | filtered_headers = dict((k, v) for k, v in filtered_headers.items() if k.lower() != 'accept-encoding') | |
1308 | del filtered_headers['Youtubedl-no-compression'] | |
1309 | ||
1310 | return filtered_headers | |
1311 | ||
1312 | ||
1313 | class YoutubeDLHandler(compat_urllib_request.HTTPHandler): | |
1314 | """Handler for HTTP requests and responses. | |
1315 | ||
1316 | This class, when installed with an OpenerDirector, automatically adds | |
1317 | the standard headers to every HTTP request and handles gzipped and | |
1318 | deflated responses from web servers. If compression is to be avoided in | |
1319 | a particular request, the original request in the program code only has | |
1320 | to include the HTTP header "Youtubedl-no-compression", which will be | |
1321 | removed before making the real request. | |
1322 | ||
1323 | Part of this code was copied from: | |
1324 | ||
1325 | http://techknack.net/python-urllib2-handlers/ | |
1326 | ||
1327 | Andrew Rowls, the author of that code, agreed to release it to the | |
1328 | public domain. | |
1329 | """ | |
1330 | ||
1331 | def __init__(self, params, *args, **kwargs): | |
1332 | compat_urllib_request.HTTPHandler.__init__(self, *args, **kwargs) | |
1333 | self._params = params | |
1334 | ||
1335 | def http_open(self, req): | |
1336 | conn_class = compat_http_client.HTTPConnection | |
1337 | ||
1338 | socks_proxy = req.headers.get('Ytdl-socks-proxy') | |
1339 | if socks_proxy: | |
1340 | conn_class = make_socks_conn_class(conn_class, socks_proxy) | |
1341 | del req.headers['Ytdl-socks-proxy'] | |
1342 | ||
1343 | return self.do_open(functools.partial( | |
1344 | _create_http_connection, self, conn_class, False), | |
1345 | req) | |
1346 | ||
1347 | @staticmethod | |
1348 | def deflate(data): | |
1349 | if not data: | |
1350 | return data | |
1351 | try: | |
1352 | return zlib.decompress(data, -zlib.MAX_WBITS) | |
1353 | except zlib.error: | |
1354 | return zlib.decompress(data) | |
1355 | ||
1356 | def http_request(self, req): | |
1357 | # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not | |
1358 | # always respected by websites, some tend to give out URLs with non percent-encoded | |
1359 | # non-ASCII characters (see telemb.py, ard.py [#3412]) | |
1360 | # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991) | |
1361 | # To work around aforementioned issue we will replace request's original URL with | |
1362 | # percent-encoded one | |
1363 | # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09) | |
1364 | # the code of this workaround has been moved here from YoutubeDL.urlopen() | |
1365 | url = req.get_full_url() | |
1366 | url_escaped = escape_url(url) | |
1367 | ||
1368 | # Substitute URL if any change after escaping | |
1369 | if url != url_escaped: | |
1370 | req = update_Request(req, url=url_escaped) | |
1371 | ||
1372 | for h, v in std_headers.items(): | |
1373 | # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275 | |
1374 | # The dict keys are capitalized because of this bug by urllib | |
1375 | if h.capitalize() not in req.headers: | |
1376 | req.add_header(h, v) | |
1377 | ||
1378 | req.headers = handle_youtubedl_headers(req.headers) | |
1379 | ||
1380 | if sys.version_info < (2, 7) and '#' in req.get_full_url(): | |
1381 | # Python 2.6 is brain-dead when it comes to fragments | |
1382 | req._Request__original = req._Request__original.partition('#')[0] | |
1383 | req._Request__r_type = req._Request__r_type.partition('#')[0] | |
1384 | ||
1385 | return req | |
1386 | ||
1387 | def http_response(self, req, resp): | |
1388 | old_resp = resp | |
1389 | # gzip | |
1390 | if resp.headers.get('Content-encoding', '') == 'gzip': | |
1391 | content = resp.read() | |
1392 | gz = gzip.GzipFile(fileobj=io.BytesIO(content), mode='rb') | |
1393 | try: | |
1394 | uncompressed = io.BytesIO(gz.read()) | |
1395 | except IOError as original_ioerror: | |
1396 | # There may be junk add the end of the file | |
1397 | # See http://stackoverflow.com/q/4928560/35070 for details | |
1398 | for i in range(1, 1024): | |
1399 | try: | |
1400 | gz = gzip.GzipFile(fileobj=io.BytesIO(content[:-i]), mode='rb') | |
1401 | uncompressed = io.BytesIO(gz.read()) | |
1402 | except IOError: | |
1403 | continue | |
1404 | break | |
1405 | else: | |
1406 | raise original_ioerror | |
1407 | resp = compat_urllib_request.addinfourl(uncompressed, old_resp.headers, old_resp.url, old_resp.code) | |
1408 | resp.msg = old_resp.msg | |
1409 | del resp.headers['Content-encoding'] | |
1410 | # deflate | |
1411 | if resp.headers.get('Content-encoding', '') == 'deflate': | |
1412 | gz = io.BytesIO(self.deflate(resp.read())) | |
1413 | resp = compat_urllib_request.addinfourl(gz, old_resp.headers, old_resp.url, old_resp.code) | |
1414 | resp.msg = old_resp.msg | |
1415 | del resp.headers['Content-encoding'] | |
1416 | # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see | |
1417 | # https://github.com/ytdl-org/youtube-dl/issues/6457). | |
1418 | if 300 <= resp.code < 400: | |
1419 | location = resp.headers.get('Location') | |
1420 | if location: | |
1421 | # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3 | |
1422 | if sys.version_info >= (3, 0): | |
1423 | location = location.encode('iso-8859-1').decode('utf-8') | |
1424 | else: | |
1425 | location = location.decode('utf-8') | |
1426 | location_escaped = escape_url(location) | |
1427 | if location != location_escaped: | |
1428 | del resp.headers['Location'] | |
1429 | if sys.version_info < (3, 0): | |
1430 | location_escaped = location_escaped.encode('utf-8') | |
1431 | resp.headers['Location'] = location_escaped | |
1432 | return resp | |
1433 | ||
1434 | https_request = http_request | |
1435 | https_response = http_response | |
1436 | ||
1437 | ||
1438 | def make_socks_conn_class(base_class, socks_proxy): | |
1439 | assert issubclass(base_class, ( | |
1440 | compat_http_client.HTTPConnection, compat_http_client.HTTPSConnection)) | |
1441 | ||
1442 | url_components = compat_urlparse.urlparse(socks_proxy) | |
1443 | if url_components.scheme.lower() == 'socks5': | |
1444 | socks_type = ProxyType.SOCKS5 | |
1445 | elif url_components.scheme.lower() in ('socks', 'socks4'): | |
1446 | socks_type = ProxyType.SOCKS4 | |
1447 | elif url_components.scheme.lower() == 'socks4a': | |
1448 | socks_type = ProxyType.SOCKS4A | |
1449 | ||
1450 | def unquote_if_non_empty(s): | |
1451 | if not s: | |
1452 | return s | |
1453 | return compat_urllib_parse_unquote_plus(s) | |
1454 | ||
1455 | proxy_args = ( | |
1456 | socks_type, | |
1457 | url_components.hostname, url_components.port or 1080, | |
1458 | True, # Remote DNS | |
1459 | unquote_if_non_empty(url_components.username), | |
1460 | unquote_if_non_empty(url_components.password), | |
1461 | ) | |
1462 | ||
1463 | class SocksConnection(base_class): | |
1464 | def connect(self): | |
1465 | self.sock = sockssocket() | |
1466 | self.sock.setproxy(*proxy_args) | |
1467 | if type(self.timeout) in (int, float): | |
1468 | self.sock.settimeout(self.timeout) | |
1469 | self.sock.connect((self.host, self.port)) | |
1470 | ||
1471 | if isinstance(self, compat_http_client.HTTPSConnection): | |
1472 | if hasattr(self, '_context'): # Python > 2.6 | |
1473 | self.sock = self._context.wrap_socket( | |
1474 | self.sock, server_hostname=self.host) | |
1475 | else: | |
1476 | self.sock = ssl.wrap_socket(self.sock) | |
1477 | ||
1478 | return SocksConnection | |
1479 | ||
1480 | ||
1481 | class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler): | |
1482 | def __init__(self, params, https_conn_class=None, *args, **kwargs): | |
1483 | compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs) | |
1484 | self._https_conn_class = https_conn_class or compat_http_client.HTTPSConnection | |
1485 | self._params = params | |
1486 | ||
1487 | def https_open(self, req): | |
1488 | kwargs = {} | |
1489 | conn_class = self._https_conn_class | |
1490 | ||
1491 | if hasattr(self, '_context'): # python > 2.6 | |
1492 | kwargs['context'] = self._context | |
1493 | if hasattr(self, '_check_hostname'): # python 3.x | |
1494 | kwargs['check_hostname'] = self._check_hostname | |
1495 | ||
1496 | socks_proxy = req.headers.get('Ytdl-socks-proxy') | |
1497 | if socks_proxy: | |
1498 | conn_class = make_socks_conn_class(conn_class, socks_proxy) | |
1499 | del req.headers['Ytdl-socks-proxy'] | |
1500 | ||
1501 | return self.do_open(functools.partial( | |
1502 | _create_http_connection, self, conn_class, True), | |
1503 | req, **kwargs) | |
1504 | ||
1505 | ||
1506 | class YoutubeDLCookieJar(compat_cookiejar.MozillaCookieJar): | |
1507 | """ | |
1508 | See [1] for cookie file format. | |
1509 | ||
1510 | 1. https://curl.haxx.se/docs/http-cookies.html | |
1511 | """ | |
1512 | _HTTPONLY_PREFIX = '#HttpOnly_' | |
1513 | _ENTRY_LEN = 7 | |
1514 | _HEADER = '''# Netscape HTTP Cookie File | |
1515 | # This file is generated by yt-dlp. Do not edit. | |
1516 | ||
1517 | ''' | |
1518 | _CookieFileEntry = collections.namedtuple( | |
1519 | 'CookieFileEntry', | |
1520 | ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value')) | |
1521 | ||
1522 | def save(self, filename=None, ignore_discard=False, ignore_expires=False): | |
1523 | """ | |
1524 | Save cookies to a file. | |
1525 | ||
1526 | Most of the code is taken from CPython 3.8 and slightly adapted | |
1527 | to support cookie files with UTF-8 in both python 2 and 3. | |
1528 | """ | |
1529 | if filename is None: | |
1530 | if self.filename is not None: | |
1531 | filename = self.filename | |
1532 | else: | |
1533 | raise ValueError(compat_cookiejar.MISSING_FILENAME_TEXT) | |
1534 | ||
1535 | # Store session cookies with `expires` set to 0 instead of an empty | |
1536 | # string | |
1537 | for cookie in self: | |
1538 | if cookie.expires is None: | |
1539 | cookie.expires = 0 | |
1540 | ||
1541 | with io.open(filename, 'w', encoding='utf-8') as f: | |
1542 | f.write(self._HEADER) | |
1543 | now = time.time() | |
1544 | for cookie in self: | |
1545 | if not ignore_discard and cookie.discard: | |
1546 | continue | |
1547 | if not ignore_expires and cookie.is_expired(now): | |
1548 | continue | |
1549 | if cookie.secure: | |
1550 | secure = 'TRUE' | |
1551 | else: | |
1552 | secure = 'FALSE' | |
1553 | if cookie.domain.startswith('.'): | |
1554 | initial_dot = 'TRUE' | |
1555 | else: | |
1556 | initial_dot = 'FALSE' | |
1557 | if cookie.expires is not None: | |
1558 | expires = compat_str(cookie.expires) | |
1559 | else: | |
1560 | expires = '' | |
1561 | if cookie.value is None: | |
1562 | # cookies.txt regards 'Set-Cookie: foo' as a cookie | |
1563 | # with no name, whereas http.cookiejar regards it as a | |
1564 | # cookie with no value. | |
1565 | name = '' | |
1566 | value = cookie.name | |
1567 | else: | |
1568 | name = cookie.name | |
1569 | value = cookie.value | |
1570 | f.write( | |
1571 | '\t'.join([cookie.domain, initial_dot, cookie.path, | |
1572 | secure, expires, name, value]) + '\n') | |
1573 | ||
1574 | def load(self, filename=None, ignore_discard=False, ignore_expires=False): | |
1575 | """Load cookies from a file.""" | |
1576 | if filename is None: | |
1577 | if self.filename is not None: | |
1578 | filename = self.filename | |
1579 | else: | |
1580 | raise ValueError(compat_cookiejar.MISSING_FILENAME_TEXT) | |
1581 | ||
1582 | def prepare_line(line): | |
1583 | if line.startswith(self._HTTPONLY_PREFIX): | |
1584 | line = line[len(self._HTTPONLY_PREFIX):] | |
1585 | # comments and empty lines are fine | |
1586 | if line.startswith('#') or not line.strip(): | |
1587 | return line | |
1588 | cookie_list = line.split('\t') | |
1589 | if len(cookie_list) != self._ENTRY_LEN: | |
1590 | raise compat_cookiejar.LoadError('invalid length %d' % len(cookie_list)) | |
1591 | cookie = self._CookieFileEntry(*cookie_list) | |
1592 | if cookie.expires_at and not cookie.expires_at.isdigit(): | |
1593 | raise compat_cookiejar.LoadError('invalid expires at %s' % cookie.expires_at) | |
1594 | return line | |
1595 | ||
1596 | cf = io.StringIO() | |
1597 | with io.open(filename, encoding='utf-8') as f: | |
1598 | for line in f: | |
1599 | try: | |
1600 | cf.write(prepare_line(line)) | |
1601 | except compat_cookiejar.LoadError as e: | |
1602 | write_string( | |
1603 | 'WARNING: skipping cookie file entry due to %s: %r\n' | |
1604 | % (e, line), sys.stderr) | |
1605 | continue | |
1606 | cf.seek(0) | |
1607 | self._really_load(cf, filename, ignore_discard, ignore_expires) | |
1608 | # Session cookies are denoted by either `expires` field set to | |
1609 | # an empty string or 0. MozillaCookieJar only recognizes the former | |
1610 | # (see [1]). So we need force the latter to be recognized as session | |
1611 | # cookies on our own. | |
1612 | # Session cookies may be important for cookies-based authentication, | |
1613 | # e.g. usually, when user does not check 'Remember me' check box while | |
1614 | # logging in on a site, some important cookies are stored as session | |
1615 | # cookies so that not recognizing them will result in failed login. | |
1616 | # 1. https://bugs.python.org/issue17164 | |
1617 | for cookie in self: | |
1618 | # Treat `expires=0` cookies as session cookies | |
1619 | if cookie.expires == 0: | |
1620 | cookie.expires = None | |
1621 | cookie.discard = True | |
1622 | ||
1623 | ||
1624 | class YoutubeDLCookieProcessor(compat_urllib_request.HTTPCookieProcessor): | |
1625 | def __init__(self, cookiejar=None): | |
1626 | compat_urllib_request.HTTPCookieProcessor.__init__(self, cookiejar) | |
1627 | ||
1628 | def http_response(self, request, response): | |
1629 | # Python 2 will choke on next HTTP request in row if there are non-ASCII | |
1630 | # characters in Set-Cookie HTTP header of last response (see | |
1631 | # https://github.com/ytdl-org/youtube-dl/issues/6769). | |
1632 | # In order to at least prevent crashing we will percent encode Set-Cookie | |
1633 | # header before HTTPCookieProcessor starts processing it. | |
1634 | # if sys.version_info < (3, 0) and response.headers: | |
1635 | # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'): | |
1636 | # set_cookie = response.headers.get(set_cookie_header) | |
1637 | # if set_cookie: | |
1638 | # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ") | |
1639 | # if set_cookie != set_cookie_escaped: | |
1640 | # del response.headers[set_cookie_header] | |
1641 | # response.headers[set_cookie_header] = set_cookie_escaped | |
1642 | return compat_urllib_request.HTTPCookieProcessor.http_response(self, request, response) | |
1643 | ||
1644 | https_request = compat_urllib_request.HTTPCookieProcessor.http_request | |
1645 | https_response = http_response | |
1646 | ||
1647 | ||
1648 | class YoutubeDLRedirectHandler(compat_urllib_request.HTTPRedirectHandler): | |
1649 | """YoutubeDL redirect handler | |
1650 | ||
1651 | The code is based on HTTPRedirectHandler implementation from CPython [1]. | |
1652 | ||
1653 | This redirect handler solves two issues: | |
1654 | - ensures redirect URL is always unicode under python 2 | |
1655 | - introduces support for experimental HTTP response status code | |
1656 | 308 Permanent Redirect [2] used by some sites [3] | |
1657 | ||
1658 | 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py | |
1659 | 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308 | |
1660 | 3. https://github.com/ytdl-org/youtube-dl/issues/28768 | |
1661 | """ | |
1662 | ||
1663 | http_error_301 = http_error_303 = http_error_307 = http_error_308 = compat_urllib_request.HTTPRedirectHandler.http_error_302 | |
1664 | ||
1665 | def redirect_request(self, req, fp, code, msg, headers, newurl): | |
1666 | """Return a Request or None in response to a redirect. | |
1667 | ||
1668 | This is called by the http_error_30x methods when a | |
1669 | redirection response is received. If a redirection should | |
1670 | take place, return a new Request to allow http_error_30x to | |
1671 | perform the redirect. Otherwise, raise HTTPError if no-one | |
1672 | else should try to handle this url. Return None if you can't | |
1673 | but another Handler might. | |
1674 | """ | |
1675 | m = req.get_method() | |
1676 | if (not (code in (301, 302, 303, 307, 308) and m in ("GET", "HEAD") | |
1677 | or code in (301, 302, 303) and m == "POST")): | |
1678 | raise compat_HTTPError(req.full_url, code, msg, headers, fp) | |
1679 | # Strictly (according to RFC 2616), 301 or 302 in response to | |
1680 | # a POST MUST NOT cause a redirection without confirmation | |
1681 | # from the user (of urllib.request, in this case). In practice, | |
1682 | # essentially all clients do redirect in this case, so we do | |
1683 | # the same. | |
1684 | ||
1685 | # On python 2 urlh.geturl() may sometimes return redirect URL | |
1686 | # as byte string instead of unicode. This workaround allows | |
1687 | # to force it always return unicode. | |
1688 | if sys.version_info[0] < 3: | |
1689 | newurl = compat_str(newurl) | |
1690 | ||
1691 | # Be conciliant with URIs containing a space. This is mainly | |
1692 | # redundant with the more complete encoding done in http_error_302(), | |
1693 | # but it is kept for compatibility with other callers. | |
1694 | newurl = newurl.replace(' ', '%20') | |
1695 | ||
1696 | CONTENT_HEADERS = ("content-length", "content-type") | |
1697 | # NB: don't use dict comprehension for python 2.6 compatibility | |
1698 | newheaders = dict((k, v) for k, v in req.headers.items() | |
1699 | if k.lower() not in CONTENT_HEADERS) | |
1700 | return compat_urllib_request.Request( | |
1701 | newurl, headers=newheaders, origin_req_host=req.origin_req_host, | |
1702 | unverifiable=True) | |
1703 | ||
1704 | ||
1705 | def extract_timezone(date_str): | |
1706 | m = re.search( | |
1707 | r'''(?x) | |
1708 | ^.{8,}? # >=8 char non-TZ prefix, if present | |
1709 | (?P<tz>Z| # just the UTC Z, or | |
1710 | (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or | |
1711 | (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits | |
1712 | [ ]? # optional space | |
1713 | (?P<sign>\+|-) # +/- | |
1714 | (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm | |
1715 | $) | |
1716 | ''', date_str) | |
1717 | if not m: | |
1718 | timezone = datetime.timedelta() | |
1719 | else: | |
1720 | date_str = date_str[:-len(m.group('tz'))] | |
1721 | if not m.group('sign'): | |
1722 | timezone = datetime.timedelta() | |
1723 | else: | |
1724 | sign = 1 if m.group('sign') == '+' else -1 | |
1725 | timezone = datetime.timedelta( | |
1726 | hours=sign * int(m.group('hours')), | |
1727 | minutes=sign * int(m.group('minutes'))) | |
1728 | return timezone, date_str | |
1729 | ||
1730 | ||
1731 | def parse_iso8601(date_str, delimiter='T', timezone=None): | |
1732 | """ Return a UNIX timestamp from the given date """ | |
1733 | ||
1734 | if date_str is None: | |
1735 | return None | |
1736 | ||
1737 | date_str = re.sub(r'\.[0-9]+', '', date_str) | |
1738 | ||
1739 | if timezone is None: | |
1740 | timezone, date_str = extract_timezone(date_str) | |
1741 | ||
1742 | try: | |
1743 | date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter) | |
1744 | dt = datetime.datetime.strptime(date_str, date_format) - timezone | |
1745 | return calendar.timegm(dt.timetuple()) | |
1746 | except ValueError: | |
1747 | pass | |
1748 | ||
1749 | ||
1750 | def date_formats(day_first=True): | |
1751 | return DATE_FORMATS_DAY_FIRST if day_first else DATE_FORMATS_MONTH_FIRST | |
1752 | ||
1753 | ||
1754 | def unified_strdate(date_str, day_first=True): | |
1755 | """Return a string with the date in the format YYYYMMDD""" | |
1756 | ||
1757 | if date_str is None: | |
1758 | return None | |
1759 | upload_date = None | |
1760 | # Replace commas | |
1761 | date_str = date_str.replace(',', ' ') | |
1762 | # Remove AM/PM + timezone | |
1763 | date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str) | |
1764 | _, date_str = extract_timezone(date_str) | |
1765 | ||
1766 | for expression in date_formats(day_first): | |
1767 | try: | |
1768 | upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d') | |
1769 | except ValueError: | |
1770 | pass | |
1771 | if upload_date is None: | |
1772 | timetuple = email.utils.parsedate_tz(date_str) | |
1773 | if timetuple: | |
1774 | try: | |
1775 | upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d') | |
1776 | except ValueError: | |
1777 | pass | |
1778 | if upload_date is not None: | |
1779 | return compat_str(upload_date) | |
1780 | ||
1781 | ||
1782 | def unified_timestamp(date_str, day_first=True): | |
1783 | if date_str is None: | |
1784 | return None | |
1785 | ||
1786 | date_str = re.sub(r'[,|]', '', date_str) | |
1787 | ||
1788 | pm_delta = 12 if re.search(r'(?i)PM', date_str) else 0 | |
1789 | timezone, date_str = extract_timezone(date_str) | |
1790 | ||
1791 | # Remove AM/PM + timezone | |
1792 | date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str) | |
1793 | ||
1794 | # Remove unrecognized timezones from ISO 8601 alike timestamps | |
1795 | m = re.search(r'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str) | |
1796 | if m: | |
1797 | date_str = date_str[:-len(m.group('tz'))] | |
1798 | ||
1799 | # Python only supports microseconds, so remove nanoseconds | |
1800 | m = re.search(r'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str) | |
1801 | if m: | |
1802 | date_str = m.group(1) | |
1803 | ||
1804 | for expression in date_formats(day_first): | |
1805 | try: | |
1806 | dt = datetime.datetime.strptime(date_str, expression) - timezone + datetime.timedelta(hours=pm_delta) | |
1807 | return calendar.timegm(dt.timetuple()) | |
1808 | except ValueError: | |
1809 | pass | |
1810 | timetuple = email.utils.parsedate_tz(date_str) | |
1811 | if timetuple: | |
1812 | return calendar.timegm(timetuple) + pm_delta * 3600 | |
1813 | ||
1814 | ||
1815 | def determine_ext(url, default_ext='unknown_video'): | |
1816 | if url is None or '.' not in url: | |
1817 | return default_ext | |
1818 | guess = url.partition('?')[0].rpartition('.')[2] | |
1819 | if re.match(r'^[A-Za-z0-9]+$', guess): | |
1820 | return guess | |
1821 | # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download | |
1822 | elif guess.rstrip('/') in KNOWN_EXTENSIONS: | |
1823 | return guess.rstrip('/') | |
1824 | else: | |
1825 | return default_ext | |
1826 | ||
1827 | ||
1828 | def subtitles_filename(filename, sub_lang, sub_format, expected_real_ext=None): | |
1829 | return replace_extension(filename, sub_lang + '.' + sub_format, expected_real_ext) | |
1830 | ||
1831 | ||
1832 | def datetime_from_str(date_str, precision='auto', format='%Y%m%d'): | |
1833 | """ | |
1834 | Return a datetime object from a string in the format YYYYMMDD or | |
1835 | (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)? | |
1836 | ||
1837 | format: string date format used to return datetime object from | |
1838 | precision: round the time portion of a datetime object. | |
1839 | auto|microsecond|second|minute|hour|day. | |
1840 | auto: round to the unit provided in date_str (if applicable). | |
1841 | """ | |
1842 | auto_precision = False | |
1843 | if precision == 'auto': | |
1844 | auto_precision = True | |
1845 | precision = 'microsecond' | |
1846 | today = datetime_round(datetime.datetime.utcnow(), precision) | |
1847 | if date_str in ('now', 'today'): | |
1848 | return today | |
1849 | if date_str == 'yesterday': | |
1850 | return today - datetime.timedelta(days=1) | |
1851 | match = re.match( | |
1852 | r'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?', | |
1853 | date_str) | |
1854 | if match is not None: | |
1855 | start_time = datetime_from_str(match.group('start'), precision, format) | |
1856 | time = int(match.group('time')) * (-1 if match.group('sign') == '-' else 1) | |
1857 | unit = match.group('unit') | |
1858 | if unit == 'month' or unit == 'year': | |
1859 | new_date = datetime_add_months(start_time, time * 12 if unit == 'year' else time) | |
1860 | unit = 'day' | |
1861 | else: | |
1862 | if unit == 'week': | |
1863 | unit = 'day' | |
1864 | time *= 7 | |
1865 | delta = datetime.timedelta(**{unit + 's': time}) | |
1866 | new_date = start_time + delta | |
1867 | if auto_precision: | |
1868 | return datetime_round(new_date, unit) | |
1869 | return new_date | |
1870 | ||
1871 | return datetime_round(datetime.datetime.strptime(date_str, format), precision) | |
1872 | ||
1873 | ||
1874 | def date_from_str(date_str, format='%Y%m%d'): | |
1875 | """ | |
1876 | Return a datetime object from a string in the format YYYYMMDD or | |
1877 | (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)? | |
1878 | ||
1879 | format: string date format used to return datetime object from | |
1880 | """ | |
1881 | return datetime_from_str(date_str, precision='microsecond', format=format).date() | |
1882 | ||
1883 | ||
1884 | def datetime_add_months(dt, months): | |
1885 | """Increment/Decrement a datetime object by months.""" | |
1886 | month = dt.month + months - 1 | |
1887 | year = dt.year + month // 12 | |
1888 | month = month % 12 + 1 | |
1889 | day = min(dt.day, calendar.monthrange(year, month)[1]) | |
1890 | return dt.replace(year, month, day) | |
1891 | ||
1892 | ||
1893 | def datetime_round(dt, precision='day'): | |
1894 | """ | |
1895 | Round a datetime object's time to a specific precision | |
1896 | """ | |
1897 | if precision == 'microsecond': | |
1898 | return dt | |
1899 | ||
1900 | unit_seconds = { | |
1901 | 'day': 86400, | |
1902 | 'hour': 3600, | |
1903 | 'minute': 60, | |
1904 | 'second': 1, | |
1905 | } | |
1906 | roundto = lambda x, n: ((x + n / 2) // n) * n | |
1907 | timestamp = calendar.timegm(dt.timetuple()) | |
1908 | return datetime.datetime.utcfromtimestamp(roundto(timestamp, unit_seconds[precision])) | |
1909 | ||
1910 | ||
1911 | def hyphenate_date(date_str): | |
1912 | """ | |
1913 | Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format""" | |
1914 | match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str) | |
1915 | if match is not None: | |
1916 | return '-'.join(match.groups()) | |
1917 | else: | |
1918 | return date_str | |
1919 | ||
1920 | ||
1921 | class DateRange(object): | |
1922 | """Represents a time interval between two dates""" | |
1923 | ||
1924 | def __init__(self, start=None, end=None): | |
1925 | """start and end must be strings in the format accepted by date""" | |
1926 | if start is not None: | |
1927 | self.start = date_from_str(start) | |
1928 | else: | |
1929 | self.start = datetime.datetime.min.date() | |
1930 | if end is not None: | |
1931 | self.end = date_from_str(end) | |
1932 | else: | |
1933 | self.end = datetime.datetime.max.date() | |
1934 | if self.start > self.end: | |
1935 | raise ValueError('Date range: "%s" , the start date must be before the end date' % self) | |
1936 | ||
1937 | @classmethod | |
1938 | def day(cls, day): | |
1939 | """Returns a range that only contains the given day""" | |
1940 | return cls(day, day) | |
1941 | ||
1942 | def __contains__(self, date): | |
1943 | """Check if the date is in the range""" | |
1944 | if not isinstance(date, datetime.date): | |
1945 | date = date_from_str(date) | |
1946 | return self.start <= date <= self.end | |
1947 | ||
1948 | def __str__(self): | |
1949 | return '%s - %s' % (self.start.isoformat(), self.end.isoformat()) | |
1950 | ||
1951 | ||
1952 | def platform_name(): | |
1953 | """ Returns the platform name as a compat_str """ | |
1954 | res = platform.platform() | |
1955 | if isinstance(res, bytes): | |
1956 | res = res.decode(preferredencoding()) | |
1957 | ||
1958 | assert isinstance(res, compat_str) | |
1959 | return res | |
1960 | ||
1961 | ||
1962 | def get_windows_version(): | |
1963 | ''' Get Windows version. None if it's not running on Windows ''' | |
1964 | if compat_os_name == 'nt': | |
1965 | return version_tuple(platform.win32_ver()[1]) | |
1966 | else: | |
1967 | return None | |
1968 | ||
1969 | ||
1970 | def _windows_write_string(s, out): | |
1971 | """ Returns True if the string was written using special methods, | |
1972 | False if it has yet to be written out.""" | |
1973 | # Adapted from http://stackoverflow.com/a/3259271/35070 | |
1974 | ||
1975 | import ctypes.wintypes | |
1976 | ||
1977 | WIN_OUTPUT_IDS = { | |
1978 | 1: -11, | |
1979 | 2: -12, | |
1980 | } | |
1981 | ||
1982 | try: | |
1983 | fileno = out.fileno() | |
1984 | except AttributeError: | |
1985 | # If the output stream doesn't have a fileno, it's virtual | |
1986 | return False | |
1987 | except io.UnsupportedOperation: | |
1988 | # Some strange Windows pseudo files? | |
1989 | return False | |
1990 | if fileno not in WIN_OUTPUT_IDS: | |
1991 | return False | |
1992 | ||
1993 | GetStdHandle = compat_ctypes_WINFUNCTYPE( | |
1994 | ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)( | |
1995 | ('GetStdHandle', ctypes.windll.kernel32)) | |
1996 | h = GetStdHandle(WIN_OUTPUT_IDS[fileno]) | |
1997 | ||
1998 | WriteConsoleW = compat_ctypes_WINFUNCTYPE( | |
1999 | ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR, | |
2000 | ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD), | |
2001 | ctypes.wintypes.LPVOID)(('WriteConsoleW', ctypes.windll.kernel32)) | |
2002 | written = ctypes.wintypes.DWORD(0) | |
2003 | ||
2004 | GetFileType = compat_ctypes_WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)(('GetFileType', ctypes.windll.kernel32)) | |
2005 | FILE_TYPE_CHAR = 0x0002 | |
2006 | FILE_TYPE_REMOTE = 0x8000 | |
2007 | GetConsoleMode = compat_ctypes_WINFUNCTYPE( | |
2008 | ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, | |
2009 | ctypes.POINTER(ctypes.wintypes.DWORD))( | |
2010 | ('GetConsoleMode', ctypes.windll.kernel32)) | |
2011 | INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value | |
2012 | ||
2013 | def not_a_console(handle): | |
2014 | if handle == INVALID_HANDLE_VALUE or handle is None: | |
2015 | return True | |
2016 | return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR | |
2017 | or GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0) | |
2018 | ||
2019 | if not_a_console(h): | |
2020 | return False | |
2021 | ||
2022 | def next_nonbmp_pos(s): | |
2023 | try: | |
2024 | return next(i for i, c in enumerate(s) if ord(c) > 0xffff) | |
2025 | except StopIteration: | |
2026 | return len(s) | |
2027 | ||
2028 | while s: | |
2029 | count = min(next_nonbmp_pos(s), 1024) | |
2030 | ||
2031 | ret = WriteConsoleW( | |
2032 | h, s, count if count else 2, ctypes.byref(written), None) | |
2033 | if ret == 0: | |
2034 | raise OSError('Failed to write string') | |
2035 | if not count: # We just wrote a non-BMP character | |
2036 | assert written.value == 2 | |
2037 | s = s[1:] | |
2038 | else: | |
2039 | assert written.value > 0 | |
2040 | s = s[written.value:] | |
2041 | return True | |
2042 | ||
2043 | ||
2044 | def write_string(s, out=None, encoding=None): | |
2045 | if out is None: | |
2046 | out = sys.stderr | |
2047 | assert type(s) == compat_str | |
2048 | ||
2049 | if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'): | |
2050 | if _windows_write_string(s, out): | |
2051 | return | |
2052 | ||
2053 | if ('b' in getattr(out, 'mode', '') | |
2054 | or sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr | |
2055 | byt = s.encode(encoding or preferredencoding(), 'ignore') | |
2056 | out.write(byt) | |
2057 | elif hasattr(out, 'buffer'): | |
2058 | enc = encoding or getattr(out, 'encoding', None) or preferredencoding() | |
2059 | byt = s.encode(enc, 'ignore') | |
2060 | out.buffer.write(byt) | |
2061 | else: | |
2062 | out.write(s) | |
2063 | out.flush() | |
2064 | ||
2065 | ||
2066 | def bytes_to_intlist(bs): | |
2067 | if not bs: | |
2068 | return [] | |
2069 | if isinstance(bs[0], int): # Python 3 | |
2070 | return list(bs) | |
2071 | else: | |
2072 | return [ord(c) for c in bs] | |
2073 | ||
2074 | ||
2075 | def intlist_to_bytes(xs): | |
2076 | if not xs: | |
2077 | return b'' | |
2078 | return compat_struct_pack('%dB' % len(xs), *xs) | |
2079 | ||
2080 | ||
2081 | # Cross-platform file locking | |
2082 | if sys.platform == 'win32': | |
2083 | import ctypes.wintypes | |
2084 | import msvcrt | |
2085 | ||
2086 | class OVERLAPPED(ctypes.Structure): | |
2087 | _fields_ = [ | |
2088 | ('Internal', ctypes.wintypes.LPVOID), | |
2089 | ('InternalHigh', ctypes.wintypes.LPVOID), | |
2090 | ('Offset', ctypes.wintypes.DWORD), | |
2091 | ('OffsetHigh', ctypes.wintypes.DWORD), | |
2092 | ('hEvent', ctypes.wintypes.HANDLE), | |
2093 | ] | |
2094 | ||
2095 | kernel32 = ctypes.windll.kernel32 | |
2096 | LockFileEx = kernel32.LockFileEx | |
2097 | LockFileEx.argtypes = [ | |
2098 | ctypes.wintypes.HANDLE, # hFile | |
2099 | ctypes.wintypes.DWORD, # dwFlags | |
2100 | ctypes.wintypes.DWORD, # dwReserved | |
2101 | ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow | |
2102 | ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh | |
2103 | ctypes.POINTER(OVERLAPPED) # Overlapped | |
2104 | ] | |
2105 | LockFileEx.restype = ctypes.wintypes.BOOL | |
2106 | UnlockFileEx = kernel32.UnlockFileEx | |
2107 | UnlockFileEx.argtypes = [ | |
2108 | ctypes.wintypes.HANDLE, # hFile | |
2109 | ctypes.wintypes.DWORD, # dwReserved | |
2110 | ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow | |
2111 | ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh | |
2112 | ctypes.POINTER(OVERLAPPED) # Overlapped | |
2113 | ] | |
2114 | UnlockFileEx.restype = ctypes.wintypes.BOOL | |
2115 | whole_low = 0xffffffff | |
2116 | whole_high = 0x7fffffff | |
2117 | ||
2118 | def _lock_file(f, exclusive, block): # todo: block unused on win32 | |
2119 | overlapped = OVERLAPPED() | |
2120 | overlapped.Offset = 0 | |
2121 | overlapped.OffsetHigh = 0 | |
2122 | overlapped.hEvent = 0 | |
2123 | f._lock_file_overlapped_p = ctypes.pointer(overlapped) | |
2124 | handle = msvcrt.get_osfhandle(f.fileno()) | |
2125 | if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0, | |
2126 | whole_low, whole_high, f._lock_file_overlapped_p): | |
2127 | raise OSError('Locking file failed: %r' % ctypes.FormatError()) | |
2128 | ||
2129 | def _unlock_file(f): | |
2130 | assert f._lock_file_overlapped_p | |
2131 | handle = msvcrt.get_osfhandle(f.fileno()) | |
2132 | if not UnlockFileEx(handle, 0, | |
2133 | whole_low, whole_high, f._lock_file_overlapped_p): | |
2134 | raise OSError('Unlocking file failed: %r' % ctypes.FormatError()) | |
2135 | ||
2136 | else: | |
2137 | # Some platforms, such as Jython, is missing fcntl | |
2138 | try: | |
2139 | import fcntl | |
2140 | ||
2141 | def _lock_file(f, exclusive, block): | |
2142 | fcntl.flock(f, | |
2143 | fcntl.LOCK_SH if not exclusive | |
2144 | else fcntl.LOCK_EX if block | |
2145 | else fcntl.LOCK_EX | fcntl.LOCK_NB) | |
2146 | ||
2147 | def _unlock_file(f): | |
2148 | fcntl.flock(f, fcntl.LOCK_UN) | |
2149 | ||
2150 | except ImportError: | |
2151 | UNSUPPORTED_MSG = 'file locking is not supported on this platform' | |
2152 | ||
2153 | def _lock_file(f, exclusive, block): | |
2154 | raise IOError(UNSUPPORTED_MSG) | |
2155 | ||
2156 | def _unlock_file(f): | |
2157 | raise IOError(UNSUPPORTED_MSG) | |
2158 | ||
2159 | ||
2160 | class locked_file(object): | |
2161 | def __init__(self, filename, mode, block=True, encoding=None): | |
2162 | assert mode in ['r', 'rb', 'a', 'ab', 'w', 'wb'] | |
2163 | self.f = io.open(filename, mode, encoding=encoding) | |
2164 | self.mode = mode | |
2165 | self.block = block | |
2166 | ||
2167 | def __enter__(self): | |
2168 | exclusive = 'r' not in self.mode | |
2169 | try: | |
2170 | _lock_file(self.f, exclusive, self.block) | |
2171 | except IOError: | |
2172 | self.f.close() | |
2173 | raise | |
2174 | return self | |
2175 | ||
2176 | def __exit__(self, etype, value, traceback): | |
2177 | try: | |
2178 | _unlock_file(self.f) | |
2179 | finally: | |
2180 | self.f.close() | |
2181 | ||
2182 | def __iter__(self): | |
2183 | return iter(self.f) | |
2184 | ||
2185 | def write(self, *args): | |
2186 | return self.f.write(*args) | |
2187 | ||
2188 | def read(self, *args): | |
2189 | return self.f.read(*args) | |
2190 | ||
2191 | def flush(self): | |
2192 | self.f.flush() | |
2193 | ||
2194 | def open(self): | |
2195 | return self.__enter__() | |
2196 | ||
2197 | def close(self, *args): | |
2198 | self.__exit__(self, *args, value=False, traceback=False) | |
2199 | ||
2200 | ||
2201 | def get_filesystem_encoding(): | |
2202 | encoding = sys.getfilesystemencoding() | |
2203 | return encoding if encoding is not None else 'utf-8' | |
2204 | ||
2205 | ||
2206 | def shell_quote(args): | |
2207 | quoted_args = [] | |
2208 | encoding = get_filesystem_encoding() | |
2209 | for a in args: | |
2210 | if isinstance(a, bytes): | |
2211 | # We may get a filename encoded with 'encodeFilename' | |
2212 | a = a.decode(encoding) | |
2213 | quoted_args.append(compat_shlex_quote(a)) | |
2214 | return ' '.join(quoted_args) | |
2215 | ||
2216 | ||
2217 | def smuggle_url(url, data): | |
2218 | """ Pass additional data in a URL for internal use. """ | |
2219 | ||
2220 | url, idata = unsmuggle_url(url, {}) | |
2221 | data.update(idata) | |
2222 | sdata = compat_urllib_parse_urlencode( | |
2223 | {'__youtubedl_smuggle': json.dumps(data)}) | |
2224 | return url + '#' + sdata | |
2225 | ||
2226 | ||
2227 | def unsmuggle_url(smug_url, default=None): | |
2228 | if '#__youtubedl_smuggle' not in smug_url: | |
2229 | return smug_url, default | |
2230 | url, _, sdata = smug_url.rpartition('#') | |
2231 | jsond = compat_parse_qs(sdata)['__youtubedl_smuggle'][0] | |
2232 | data = json.loads(jsond) | |
2233 | return url, data | |
2234 | ||
2235 | ||
2236 | def format_decimal_suffix(num, fmt='%d%s', *, factor=1000): | |
2237 | """ Formats numbers with decimal sufixes like K, M, etc """ | |
2238 | num, factor = float_or_none(num), float(factor) | |
2239 | if num is None: | |
2240 | return None | |
2241 | exponent = 0 if num == 0 else int(math.log(num, factor)) | |
2242 | suffix = ['', *'kMGTPEZY'][exponent] | |
2243 | if factor == 1024: | |
2244 | suffix = {'k': 'Ki', '': ''}.get(suffix, f'{suffix}i') | |
2245 | converted = num / (factor ** exponent) | |
2246 | return fmt % (converted, suffix) | |
2247 | ||
2248 | ||
2249 | def format_bytes(bytes): | |
2250 | return format_decimal_suffix(bytes, '%.2f%sB', factor=1024) or 'N/A' | |
2251 | ||
2252 | ||
2253 | def lookup_unit_table(unit_table, s): | |
2254 | units_re = '|'.join(re.escape(u) for u in unit_table) | |
2255 | m = re.match( | |
2256 | r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re, s) | |
2257 | if not m: | |
2258 | return None | |
2259 | num_str = m.group('num').replace(',', '.') | |
2260 | mult = unit_table[m.group('unit')] | |
2261 | return int(float(num_str) * mult) | |
2262 | ||
2263 | ||
2264 | def parse_filesize(s): | |
2265 | if s is None: | |
2266 | return None | |
2267 | ||
2268 | # The lower-case forms are of course incorrect and unofficial, | |
2269 | # but we support those too | |
2270 | _UNIT_TABLE = { | |
2271 | 'B': 1, | |
2272 | 'b': 1, | |
2273 | 'bytes': 1, | |
2274 | 'KiB': 1024, | |
2275 | 'KB': 1000, | |
2276 | 'kB': 1024, | |
2277 | 'Kb': 1000, | |
2278 | 'kb': 1000, | |
2279 | 'kilobytes': 1000, | |
2280 | 'kibibytes': 1024, | |
2281 | 'MiB': 1024 ** 2, | |
2282 | 'MB': 1000 ** 2, | |
2283 | 'mB': 1024 ** 2, | |
2284 | 'Mb': 1000 ** 2, | |
2285 | 'mb': 1000 ** 2, | |
2286 | 'megabytes': 1000 ** 2, | |
2287 | 'mebibytes': 1024 ** 2, | |
2288 | 'GiB': 1024 ** 3, | |
2289 | 'GB': 1000 ** 3, | |
2290 | 'gB': 1024 ** 3, | |
2291 | 'Gb': 1000 ** 3, | |
2292 | 'gb': 1000 ** 3, | |
2293 | 'gigabytes': 1000 ** 3, | |
2294 | 'gibibytes': 1024 ** 3, | |
2295 | 'TiB': 1024 ** 4, | |
2296 | 'TB': 1000 ** 4, | |
2297 | 'tB': 1024 ** 4, | |
2298 | 'Tb': 1000 ** 4, | |
2299 | 'tb': 1000 ** 4, | |
2300 | 'terabytes': 1000 ** 4, | |
2301 | 'tebibytes': 1024 ** 4, | |
2302 | 'PiB': 1024 ** 5, | |
2303 | 'PB': 1000 ** 5, | |
2304 | 'pB': 1024 ** 5, | |
2305 | 'Pb': 1000 ** 5, | |
2306 | 'pb': 1000 ** 5, | |
2307 | 'petabytes': 1000 ** 5, | |
2308 | 'pebibytes': 1024 ** 5, | |
2309 | 'EiB': 1024 ** 6, | |
2310 | 'EB': 1000 ** 6, | |
2311 | 'eB': 1024 ** 6, | |
2312 | 'Eb': 1000 ** 6, | |
2313 | 'eb': 1000 ** 6, | |
2314 | 'exabytes': 1000 ** 6, | |
2315 | 'exbibytes': 1024 ** 6, | |
2316 | 'ZiB': 1024 ** 7, | |
2317 | 'ZB': 1000 ** 7, | |
2318 | 'zB': 1024 ** 7, | |
2319 | 'Zb': 1000 ** 7, | |
2320 | 'zb': 1000 ** 7, | |
2321 | 'zettabytes': 1000 ** 7, | |
2322 | 'zebibytes': 1024 ** 7, | |
2323 | 'YiB': 1024 ** 8, | |
2324 | 'YB': 1000 ** 8, | |
2325 | 'yB': 1024 ** 8, | |
2326 | 'Yb': 1000 ** 8, | |
2327 | 'yb': 1000 ** 8, | |
2328 | 'yottabytes': 1000 ** 8, | |
2329 | 'yobibytes': 1024 ** 8, | |
2330 | } | |
2331 | ||
2332 | return lookup_unit_table(_UNIT_TABLE, s) | |
2333 | ||
2334 | ||
2335 | def parse_count(s): | |
2336 | if s is None: | |
2337 | return None | |
2338 | ||
2339 | s = re.sub(r'^[^\d]+\s', '', s).strip() | |
2340 | ||
2341 | if re.match(r'^[\d,.]+$', s): | |
2342 | return str_to_int(s) | |
2343 | ||
2344 | _UNIT_TABLE = { | |
2345 | 'k': 1000, | |
2346 | 'K': 1000, | |
2347 | 'm': 1000 ** 2, | |
2348 | 'M': 1000 ** 2, | |
2349 | 'kk': 1000 ** 2, | |
2350 | 'KK': 1000 ** 2, | |
2351 | 'b': 1000 ** 3, | |
2352 | 'B': 1000 ** 3, | |
2353 | } | |
2354 | ||
2355 | ret = lookup_unit_table(_UNIT_TABLE, s) | |
2356 | if ret is not None: | |
2357 | return ret | |
2358 | ||
2359 | mobj = re.match(r'([\d,.]+)(?:$|\s)', s) | |
2360 | if mobj: | |
2361 | return str_to_int(mobj.group(1)) | |
2362 | ||
2363 | ||
2364 | def parse_resolution(s): | |
2365 | if s is None: | |
2366 | return {} | |
2367 | ||
2368 | mobj = re.search(r'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s) | |
2369 | if mobj: | |
2370 | return { | |
2371 | 'width': int(mobj.group('w')), | |
2372 | 'height': int(mobj.group('h')), | |
2373 | } | |
2374 | ||
2375 | mobj = re.search(r'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s) | |
2376 | if mobj: | |
2377 | return {'height': int(mobj.group(1))} | |
2378 | ||
2379 | mobj = re.search(r'\b([48])[kK]\b', s) | |
2380 | if mobj: | |
2381 | return {'height': int(mobj.group(1)) * 540} | |
2382 | ||
2383 | return {} | |
2384 | ||
2385 | ||
2386 | def parse_bitrate(s): | |
2387 | if not isinstance(s, compat_str): | |
2388 | return | |
2389 | mobj = re.search(r'\b(\d+)\s*kbps', s) | |
2390 | if mobj: | |
2391 | return int(mobj.group(1)) | |
2392 | ||
2393 | ||
2394 | def month_by_name(name, lang='en'): | |
2395 | """ Return the number of a month by (locale-independently) English name """ | |
2396 | ||
2397 | month_names = MONTH_NAMES.get(lang, MONTH_NAMES['en']) | |
2398 | ||
2399 | try: | |
2400 | return month_names.index(name) + 1 | |
2401 | except ValueError: | |
2402 | return None | |
2403 | ||
2404 | ||
2405 | def month_by_abbreviation(abbrev): | |
2406 | """ Return the number of a month by (locale-independently) English | |
2407 | abbreviations """ | |
2408 | ||
2409 | try: | |
2410 | return [s[:3] for s in ENGLISH_MONTH_NAMES].index(abbrev) + 1 | |
2411 | except ValueError: | |
2412 | return None | |
2413 | ||
2414 | ||
2415 | def fix_xml_ampersands(xml_str): | |
2416 | """Replace all the '&' by '&' in XML""" | |
2417 | return re.sub( | |
2418 | r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)', | |
2419 | '&', | |
2420 | xml_str) | |
2421 | ||
2422 | ||
2423 | def setproctitle(title): | |
2424 | assert isinstance(title, compat_str) | |
2425 | ||
2426 | # ctypes in Jython is not complete | |
2427 | # http://bugs.jython.org/issue2148 | |
2428 | if sys.platform.startswith('java'): | |
2429 | return | |
2430 | ||
2431 | try: | |
2432 | libc = ctypes.cdll.LoadLibrary('libc.so.6') | |
2433 | except OSError: | |
2434 | return | |
2435 | except TypeError: | |
2436 | # LoadLibrary in Windows Python 2.7.13 only expects | |
2437 | # a bytestring, but since unicode_literals turns | |
2438 | # every string into a unicode string, it fails. | |
2439 | return | |
2440 | title_bytes = title.encode('utf-8') | |
2441 | buf = ctypes.create_string_buffer(len(title_bytes)) | |
2442 | buf.value = title_bytes | |
2443 | try: | |
2444 | libc.prctl(15, buf, 0, 0, 0) | |
2445 | except AttributeError: | |
2446 | return # Strange libc, just skip this | |
2447 | ||
2448 | ||
2449 | def remove_start(s, start): | |
2450 | return s[len(start):] if s is not None and s.startswith(start) else s | |
2451 | ||
2452 | ||
2453 | def remove_end(s, end): | |
2454 | return s[:-len(end)] if s is not None and s.endswith(end) else s | |
2455 | ||
2456 | ||
2457 | def remove_quotes(s): | |
2458 | if s is None or len(s) < 2: | |
2459 | return s | |
2460 | for quote in ('"', "'", ): | |
2461 | if s[0] == quote and s[-1] == quote: | |
2462 | return s[1:-1] | |
2463 | return s | |
2464 | ||
2465 | ||
2466 | def get_domain(url): | |
2467 | domain = re.match(r'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url) | |
2468 | return domain.group('domain') if domain else None | |
2469 | ||
2470 | ||
2471 | def url_basename(url): | |
2472 | path = compat_urlparse.urlparse(url).path | |
2473 | return path.strip('/').split('/')[-1] | |
2474 | ||
2475 | ||
2476 | def base_url(url): | |
2477 | return re.match(r'https?://[^?#&]+/', url).group() | |
2478 | ||
2479 | ||
2480 | def urljoin(base, path): | |
2481 | if isinstance(path, bytes): | |
2482 | path = path.decode('utf-8') | |
2483 | if not isinstance(path, compat_str) or not path: | |
2484 | return None | |
2485 | if re.match(r'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path): | |
2486 | return path | |
2487 | if isinstance(base, bytes): | |
2488 | base = base.decode('utf-8') | |
2489 | if not isinstance(base, compat_str) or not re.match( | |
2490 | r'^(?:https?:)?//', base): | |
2491 | return None | |
2492 | return compat_urlparse.urljoin(base, path) | |
2493 | ||
2494 | ||
2495 | class HEADRequest(compat_urllib_request.Request): | |
2496 | def get_method(self): | |
2497 | return 'HEAD' | |
2498 | ||
2499 | ||
2500 | class PUTRequest(compat_urllib_request.Request): | |
2501 | def get_method(self): | |
2502 | return 'PUT' | |
2503 | ||
2504 | ||
2505 | def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1): | |
2506 | if get_attr and v is not None: | |
2507 | v = getattr(v, get_attr, None) | |
2508 | try: | |
2509 | return int(v) * invscale // scale | |
2510 | except (ValueError, TypeError, OverflowError): | |
2511 | return default | |
2512 | ||
2513 | ||
2514 | def str_or_none(v, default=None): | |
2515 | return default if v is None else compat_str(v) | |
2516 | ||
2517 | ||
2518 | def str_to_int(int_str): | |
2519 | """ A more relaxed version of int_or_none """ | |
2520 | if isinstance(int_str, compat_integer_types): | |
2521 | return int_str | |
2522 | elif isinstance(int_str, compat_str): | |
2523 | int_str = re.sub(r'[,\.\+]', '', int_str) | |
2524 | return int_or_none(int_str) | |
2525 | ||
2526 | ||
2527 | def float_or_none(v, scale=1, invscale=1, default=None): | |
2528 | if v is None: | |
2529 | return default | |
2530 | try: | |
2531 | return float(v) * invscale / scale | |
2532 | except (ValueError, TypeError): | |
2533 | return default | |
2534 | ||
2535 | ||
2536 | def bool_or_none(v, default=None): | |
2537 | return v if isinstance(v, bool) else default | |
2538 | ||
2539 | ||
2540 | def strip_or_none(v, default=None): | |
2541 | return v.strip() if isinstance(v, compat_str) else default | |
2542 | ||
2543 | ||
2544 | def url_or_none(url): | |
2545 | if not url or not isinstance(url, compat_str): | |
2546 | return None | |
2547 | url = url.strip() | |
2548 | return url if re.match(r'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url) else None | |
2549 | ||
2550 | ||
2551 | def strftime_or_none(timestamp, date_format, default=None): | |
2552 | datetime_object = None | |
2553 | try: | |
2554 | if isinstance(timestamp, compat_numeric_types): # unix timestamp | |
2555 | datetime_object = datetime.datetime.utcfromtimestamp(timestamp) | |
2556 | elif isinstance(timestamp, compat_str): # assume YYYYMMDD | |
2557 | datetime_object = datetime.datetime.strptime(timestamp, '%Y%m%d') | |
2558 | return datetime_object.strftime(date_format) | |
2559 | except (ValueError, TypeError, AttributeError): | |
2560 | return default | |
2561 | ||
2562 | ||
2563 | def parse_duration(s): | |
2564 | if not isinstance(s, compat_basestring): | |
2565 | return None | |
2566 | s = s.strip() | |
2567 | if not s: | |
2568 | return None | |
2569 | ||
2570 | days, hours, mins, secs, ms = [None] * 5 | |
2571 | m = re.match(r'''(?x) | |
2572 | (?P<before_secs> | |
2573 | (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)? | |
2574 | (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+)) | |
2575 | (?P<ms>[.:][0-9]+)?Z?$ | |
2576 | ''', s) | |
2577 | if m: | |
2578 | days, hours, mins, secs, ms = m.group('days', 'hours', 'mins', 'secs', 'ms') | |
2579 | else: | |
2580 | m = re.match( | |
2581 | r'''(?ix)(?:P? | |
2582 | (?: | |
2583 | [0-9]+\s*y(?:ears?)?\s* | |
2584 | )? | |
2585 | (?: | |
2586 | [0-9]+\s*m(?:onths?)?\s* | |
2587 | )? | |
2588 | (?: | |
2589 | [0-9]+\s*w(?:eeks?)?\s* | |
2590 | )? | |
2591 | (?: | |
2592 | (?P<days>[0-9]+)\s*d(?:ays?)?\s* | |
2593 | )? | |
2594 | T)? | |
2595 | (?: | |
2596 | (?P<hours>[0-9]+)\s*h(?:ours?)?\s* | |
2597 | )? | |
2598 | (?: | |
2599 | (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s* | |
2600 | )? | |
2601 | (?: | |
2602 | (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s* | |
2603 | )?Z?$''', s) | |
2604 | if m: | |
2605 | days, hours, mins, secs, ms = m.groups() | |
2606 | else: | |
2607 | m = re.match(r'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s) | |
2608 | if m: | |
2609 | hours, mins = m.groups() | |
2610 | else: | |
2611 | return None | |
2612 | ||
2613 | duration = 0 | |
2614 | if secs: | |
2615 | duration += float(secs) | |
2616 | if mins: | |
2617 | duration += float(mins) * 60 | |
2618 | if hours: | |
2619 | duration += float(hours) * 60 * 60 | |
2620 | if days: | |
2621 | duration += float(days) * 24 * 60 * 60 | |
2622 | if ms: | |
2623 | duration += float(ms.replace(':', '.')) | |
2624 | return duration | |
2625 | ||
2626 | ||
2627 | def prepend_extension(filename, ext, expected_real_ext=None): | |
2628 | name, real_ext = os.path.splitext(filename) | |
2629 | return ( | |
2630 | '{0}.{1}{2}'.format(name, ext, real_ext) | |
2631 | if not expected_real_ext or real_ext[1:] == expected_real_ext | |
2632 | else '{0}.{1}'.format(filename, ext)) | |
2633 | ||
2634 | ||
2635 | def replace_extension(filename, ext, expected_real_ext=None): | |
2636 | name, real_ext = os.path.splitext(filename) | |
2637 | return '{0}.{1}'.format( | |
2638 | name if not expected_real_ext or real_ext[1:] == expected_real_ext else filename, | |
2639 | ext) | |
2640 | ||
2641 | ||
2642 | def check_executable(exe, args=[]): | |
2643 | """ Checks if the given binary is installed somewhere in PATH, and returns its name. | |
2644 | args can be a list of arguments for a short output (like -version) """ | |
2645 | try: | |
2646 | Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate_or_kill() | |
2647 | except OSError: | |
2648 | return False | |
2649 | return exe | |
2650 | ||
2651 | ||
2652 | def _get_exe_version_output(exe, args): | |
2653 | try: | |
2654 | # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers | |
2655 | # SIGTTOU if yt-dlp is run in the background. | |
2656 | # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656 | |
2657 | out, _ = Popen( | |
2658 | [encodeArgument(exe)] + args, stdin=subprocess.PIPE, | |
2659 | stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate_or_kill() | |
2660 | except OSError: | |
2661 | return False | |
2662 | if isinstance(out, bytes): # Python 2.x | |
2663 | out = out.decode('ascii', 'ignore') | |
2664 | return out | |
2665 | ||
2666 | ||
2667 | def detect_exe_version(output, version_re=None, unrecognized='present'): | |
2668 | assert isinstance(output, compat_str) | |
2669 | if version_re is None: | |
2670 | version_re = r'version\s+([-0-9._a-zA-Z]+)' | |
2671 | m = re.search(version_re, output) | |
2672 | if m: | |
2673 | return m.group(1) | |
2674 | else: | |
2675 | return unrecognized | |
2676 | ||
2677 | ||
2678 | def get_exe_version(exe, args=['--version'], | |
2679 | version_re=None, unrecognized='present'): | |
2680 | """ Returns the version of the specified executable, | |
2681 | or False if the executable is not present """ | |
2682 | out = _get_exe_version_output(exe, args) | |
2683 | return detect_exe_version(out, version_re, unrecognized) if out else False | |
2684 | ||
2685 | ||
2686 | class LazyList(collections.abc.Sequence): | |
2687 | ''' Lazy immutable list from an iterable | |
2688 | Note that slices of a LazyList are lists and not LazyList''' | |
2689 | ||
2690 | class IndexError(IndexError): | |
2691 | pass | |
2692 | ||
2693 | def __init__(self, iterable, *, reverse=False, _cache=None): | |
2694 | self.__iterable = iter(iterable) | |
2695 | self.__cache = [] if _cache is None else _cache | |
2696 | self.__reversed = reverse | |
2697 | ||
2698 | def __iter__(self): | |
2699 | if self.__reversed: | |
2700 | # We need to consume the entire iterable to iterate in reverse | |
2701 | yield from self.exhaust() | |
2702 | return | |
2703 | yield from self.__cache | |
2704 | for item in self.__iterable: | |
2705 | self.__cache.append(item) | |
2706 | yield item | |
2707 | ||
2708 | def __exhaust(self): | |
2709 | self.__cache.extend(self.__iterable) | |
2710 | # Discard the emptied iterable to make it pickle-able | |
2711 | self.__iterable = [] | |
2712 | return self.__cache | |
2713 | ||
2714 | def exhaust(self): | |
2715 | ''' Evaluate the entire iterable ''' | |
2716 | return self.__exhaust()[::-1 if self.__reversed else 1] | |
2717 | ||
2718 | @staticmethod | |
2719 | def __reverse_index(x): | |
2720 | return None if x is None else -(x + 1) | |
2721 | ||
2722 | def __getitem__(self, idx): | |
2723 | if isinstance(idx, slice): | |
2724 | if self.__reversed: | |
2725 | idx = slice(self.__reverse_index(idx.start), self.__reverse_index(idx.stop), -(idx.step or 1)) | |
2726 | start, stop, step = idx.start, idx.stop, idx.step or 1 | |
2727 | elif isinstance(idx, int): | |
2728 | if self.__reversed: | |
2729 | idx = self.__reverse_index(idx) | |
2730 | start, stop, step = idx, idx, 0 | |
2731 | else: | |
2732 | raise TypeError('indices must be integers or slices') | |
2733 | if ((start or 0) < 0 or (stop or 0) < 0 | |
2734 | or (start is None and step < 0) | |
2735 | or (stop is None and step > 0)): | |
2736 | # We need to consume the entire iterable to be able to slice from the end | |
2737 | # Obviously, never use this with infinite iterables | |
2738 | self.__exhaust() | |
2739 | try: | |
2740 | return self.__cache[idx] | |
2741 | except IndexError as e: | |
2742 | raise self.IndexError(e) from e | |
2743 | n = max(start or 0, stop or 0) - len(self.__cache) + 1 | |
2744 | if n > 0: | |
2745 | self.__cache.extend(itertools.islice(self.__iterable, n)) | |
2746 | try: | |
2747 | return self.__cache[idx] | |
2748 | except IndexError as e: | |
2749 | raise self.IndexError(e) from e | |
2750 | ||
2751 | def __bool__(self): | |
2752 | try: | |
2753 | self[-1] if self.__reversed else self[0] | |
2754 | except self.IndexError: | |
2755 | return False | |
2756 | return True | |
2757 | ||
2758 | def __len__(self): | |
2759 | self.__exhaust() | |
2760 | return len(self.__cache) | |
2761 | ||
2762 | def __reversed__(self): | |
2763 | return type(self)(self.__iterable, reverse=not self.__reversed, _cache=self.__cache) | |
2764 | ||
2765 | def __copy__(self): | |
2766 | return type(self)(self.__iterable, reverse=self.__reversed, _cache=self.__cache) | |
2767 | ||
2768 | def __repr__(self): | |
2769 | # repr and str should mimic a list. So we exhaust the iterable | |
2770 | return repr(self.exhaust()) | |
2771 | ||
2772 | def __str__(self): | |
2773 | return repr(self.exhaust()) | |
2774 | ||
2775 | ||
2776 | class PagedList: | |
2777 | ||
2778 | class IndexError(IndexError): | |
2779 | pass | |
2780 | ||
2781 | def __len__(self): | |
2782 | # This is only useful for tests | |
2783 | return len(self.getslice()) | |
2784 | ||
2785 | def __init__(self, pagefunc, pagesize, use_cache=True): | |
2786 | self._pagefunc = pagefunc | |
2787 | self._pagesize = pagesize | |
2788 | self._use_cache = use_cache | |
2789 | self._cache = {} | |
2790 | ||
2791 | def getpage(self, pagenum): | |
2792 | page_results = self._cache.get(pagenum) | |
2793 | if page_results is None: | |
2794 | page_results = list(self._pagefunc(pagenum)) | |
2795 | if self._use_cache: | |
2796 | self._cache[pagenum] = page_results | |
2797 | return page_results | |
2798 | ||
2799 | def getslice(self, start=0, end=None): | |
2800 | return list(self._getslice(start, end)) | |
2801 | ||
2802 | def _getslice(self, start, end): | |
2803 | raise NotImplementedError('This method must be implemented by subclasses') | |
2804 | ||
2805 | def __getitem__(self, idx): | |
2806 | # NOTE: cache must be enabled if this is used | |
2807 | if not isinstance(idx, int) or idx < 0: | |
2808 | raise TypeError('indices must be non-negative integers') | |
2809 | entries = self.getslice(idx, idx + 1) | |
2810 | if not entries: | |
2811 | raise self.IndexError() | |
2812 | return entries[0] | |
2813 | ||
2814 | ||
2815 | class OnDemandPagedList(PagedList): | |
2816 | def _getslice(self, start, end): | |
2817 | for pagenum in itertools.count(start // self._pagesize): | |
2818 | firstid = pagenum * self._pagesize | |
2819 | nextfirstid = pagenum * self._pagesize + self._pagesize | |
2820 | if start >= nextfirstid: | |
2821 | continue | |
2822 | ||
2823 | startv = ( | |
2824 | start % self._pagesize | |
2825 | if firstid <= start < nextfirstid | |
2826 | else 0) | |
2827 | endv = ( | |
2828 | ((end - 1) % self._pagesize) + 1 | |
2829 | if (end is not None and firstid <= end <= nextfirstid) | |
2830 | else None) | |
2831 | ||
2832 | page_results = self.getpage(pagenum) | |
2833 | if startv != 0 or endv is not None: | |
2834 | page_results = page_results[startv:endv] | |
2835 | yield from page_results | |
2836 | ||
2837 | # A little optimization - if current page is not "full", ie. does | |
2838 | # not contain page_size videos then we can assume that this page | |
2839 | # is the last one - there are no more ids on further pages - | |
2840 | # i.e. no need to query again. | |
2841 | if len(page_results) + startv < self._pagesize: | |
2842 | break | |
2843 | ||
2844 | # If we got the whole page, but the next page is not interesting, | |
2845 | # break out early as well | |
2846 | if end == nextfirstid: | |
2847 | break | |
2848 | ||
2849 | ||
2850 | class InAdvancePagedList(PagedList): | |
2851 | def __init__(self, pagefunc, pagecount, pagesize): | |
2852 | self._pagecount = pagecount | |
2853 | PagedList.__init__(self, pagefunc, pagesize, True) | |
2854 | ||
2855 | def _getslice(self, start, end): | |
2856 | start_page = start // self._pagesize | |
2857 | end_page = self._pagecount if end is None else min(self._pagecount, end // self._pagesize + 1) | |
2858 | skip_elems = start - start_page * self._pagesize | |
2859 | only_more = None if end is None else end - start | |
2860 | for pagenum in range(start_page, end_page): | |
2861 | page_results = self.getpage(pagenum) | |
2862 | if skip_elems: | |
2863 | page_results = page_results[skip_elems:] | |
2864 | skip_elems = None | |
2865 | if only_more is not None: | |
2866 | if len(page_results) < only_more: | |
2867 | only_more -= len(page_results) | |
2868 | else: | |
2869 | yield from page_results[:only_more] | |
2870 | break | |
2871 | yield from page_results | |
2872 | ||
2873 | ||
2874 | def uppercase_escape(s): | |
2875 | unicode_escape = codecs.getdecoder('unicode_escape') | |
2876 | return re.sub( | |
2877 | r'\\U[0-9a-fA-F]{8}', | |
2878 | lambda m: unicode_escape(m.group(0))[0], | |
2879 | s) | |
2880 | ||
2881 | ||
2882 | def lowercase_escape(s): | |
2883 | unicode_escape = codecs.getdecoder('unicode_escape') | |
2884 | return re.sub( | |
2885 | r'\\u[0-9a-fA-F]{4}', | |
2886 | lambda m: unicode_escape(m.group(0))[0], | |
2887 | s) | |
2888 | ||
2889 | ||
2890 | def escape_rfc3986(s): | |
2891 | """Escape non-ASCII characters as suggested by RFC 3986""" | |
2892 | if sys.version_info < (3, 0) and isinstance(s, compat_str): | |
2893 | s = s.encode('utf-8') | |
2894 | return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]") | |
2895 | ||
2896 | ||
2897 | def escape_url(url): | |
2898 | """Escape URL as suggested by RFC 3986""" | |
2899 | url_parsed = compat_urllib_parse_urlparse(url) | |
2900 | return url_parsed._replace( | |
2901 | netloc=url_parsed.netloc.encode('idna').decode('ascii'), | |
2902 | path=escape_rfc3986(url_parsed.path), | |
2903 | params=escape_rfc3986(url_parsed.params), | |
2904 | query=escape_rfc3986(url_parsed.query), | |
2905 | fragment=escape_rfc3986(url_parsed.fragment) | |
2906 | ).geturl() | |
2907 | ||
2908 | ||
2909 | def parse_qs(url): | |
2910 | return compat_parse_qs(compat_urllib_parse_urlparse(url).query) | |
2911 | ||
2912 | ||
2913 | def read_batch_urls(batch_fd): | |
2914 | def fixup(url): | |
2915 | if not isinstance(url, compat_str): | |
2916 | url = url.decode('utf-8', 'replace') | |
2917 | BOM_UTF8 = ('\xef\xbb\xbf', '\ufeff') | |
2918 | for bom in BOM_UTF8: | |
2919 | if url.startswith(bom): | |
2920 | url = url[len(bom):] | |
2921 | url = url.lstrip() | |
2922 | if not url or url.startswith(('#', ';', ']')): | |
2923 | return False | |
2924 | # "#" cannot be stripped out since it is part of the URI | |
2925 | # However, it can be safely stipped out if follwing a whitespace | |
2926 | return re.split(r'\s#', url, 1)[0].rstrip() | |
2927 | ||
2928 | with contextlib.closing(batch_fd) as fd: | |
2929 | return [url for url in map(fixup, fd) if url] | |
2930 | ||
2931 | ||
2932 | def urlencode_postdata(*args, **kargs): | |
2933 | return compat_urllib_parse_urlencode(*args, **kargs).encode('ascii') | |
2934 | ||
2935 | ||
2936 | def update_url_query(url, query): | |
2937 | if not query: | |
2938 | return url | |
2939 | parsed_url = compat_urlparse.urlparse(url) | |
2940 | qs = compat_parse_qs(parsed_url.query) | |
2941 | qs.update(query) | |
2942 | return compat_urlparse.urlunparse(parsed_url._replace( | |
2943 | query=compat_urllib_parse_urlencode(qs, True))) | |
2944 | ||
2945 | ||
2946 | def update_Request(req, url=None, data=None, headers={}, query={}): | |
2947 | req_headers = req.headers.copy() | |
2948 | req_headers.update(headers) | |
2949 | req_data = data or req.data | |
2950 | req_url = update_url_query(url or req.get_full_url(), query) | |
2951 | req_get_method = req.get_method() | |
2952 | if req_get_method == 'HEAD': | |
2953 | req_type = HEADRequest | |
2954 | elif req_get_method == 'PUT': | |
2955 | req_type = PUTRequest | |
2956 | else: | |
2957 | req_type = compat_urllib_request.Request | |
2958 | new_req = req_type( | |
2959 | req_url, data=req_data, headers=req_headers, | |
2960 | origin_req_host=req.origin_req_host, unverifiable=req.unverifiable) | |
2961 | if hasattr(req, 'timeout'): | |
2962 | new_req.timeout = req.timeout | |
2963 | return new_req | |
2964 | ||
2965 | ||
2966 | def _multipart_encode_impl(data, boundary): | |
2967 | content_type = 'multipart/form-data; boundary=%s' % boundary | |
2968 | ||
2969 | out = b'' | |
2970 | for k, v in data.items(): | |
2971 | out += b'--' + boundary.encode('ascii') + b'\r\n' | |
2972 | if isinstance(k, compat_str): | |
2973 | k = k.encode('utf-8') | |
2974 | if isinstance(v, compat_str): | |
2975 | v = v.encode('utf-8') | |
2976 | # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578 | |
2977 | # suggests sending UTF-8 directly. Firefox sends UTF-8, too | |
2978 | content = b'Content-Disposition: form-data; name="' + k + b'"\r\n\r\n' + v + b'\r\n' | |
2979 | if boundary.encode('ascii') in content: | |
2980 | raise ValueError('Boundary overlaps with data') | |
2981 | out += content | |
2982 | ||
2983 | out += b'--' + boundary.encode('ascii') + b'--\r\n' | |
2984 | ||
2985 | return out, content_type | |
2986 | ||
2987 | ||
2988 | def multipart_encode(data, boundary=None): | |
2989 | ''' | |
2990 | Encode a dict to RFC 7578-compliant form-data | |
2991 | ||
2992 | data: | |
2993 | A dict where keys and values can be either Unicode or bytes-like | |
2994 | objects. | |
2995 | boundary: | |
2996 | If specified a Unicode object, it's used as the boundary. Otherwise | |
2997 | a random boundary is generated. | |
2998 | ||
2999 | Reference: https://tools.ietf.org/html/rfc7578 | |
3000 | ''' | |
3001 | has_specified_boundary = boundary is not None | |
3002 | ||
3003 | while True: | |
3004 | if boundary is None: | |
3005 | boundary = '---------------' + str(random.randrange(0x0fffffff, 0xffffffff)) | |
3006 | ||
3007 | try: | |
3008 | out, content_type = _multipart_encode_impl(data, boundary) | |
3009 | break | |
3010 | except ValueError: | |
3011 | if has_specified_boundary: | |
3012 | raise | |
3013 | boundary = None | |
3014 | ||
3015 | return out, content_type | |
3016 | ||
3017 | ||
3018 | def dict_get(d, key_or_keys, default=None, skip_false_values=True): | |
3019 | if isinstance(key_or_keys, (list, tuple)): | |
3020 | for key in key_or_keys: | |
3021 | if key not in d or d[key] is None or skip_false_values and not d[key]: | |
3022 | continue | |
3023 | return d[key] | |
3024 | return default | |
3025 | return d.get(key_or_keys, default) | |
3026 | ||
3027 | ||
3028 | def try_get(src, getter, expected_type=None): | |
3029 | for get in variadic(getter): | |
3030 | try: | |
3031 | v = get(src) | |
3032 | except (AttributeError, KeyError, TypeError, IndexError): | |
3033 | pass | |
3034 | else: | |
3035 | if expected_type is None or isinstance(v, expected_type): | |
3036 | return v | |
3037 | ||
3038 | ||
3039 | def merge_dicts(*dicts): | |
3040 | merged = {} | |
3041 | for a_dict in dicts: | |
3042 | for k, v in a_dict.items(): | |
3043 | if v is None: | |
3044 | continue | |
3045 | if (k not in merged | |
3046 | or (isinstance(v, compat_str) and v | |
3047 | and isinstance(merged[k], compat_str) | |
3048 | and not merged[k])): | |
3049 | merged[k] = v | |
3050 | return merged | |
3051 | ||
3052 | ||
3053 | def encode_compat_str(string, encoding=preferredencoding(), errors='strict'): | |
3054 | return string if isinstance(string, compat_str) else compat_str(string, encoding, errors) | |
3055 | ||
3056 | ||
3057 | US_RATINGS = { | |
3058 | 'G': 0, | |
3059 | 'PG': 10, | |
3060 | 'PG-13': 13, | |
3061 | 'R': 16, | |
3062 | 'NC': 18, | |
3063 | } | |
3064 | ||
3065 | ||
3066 | TV_PARENTAL_GUIDELINES = { | |
3067 | 'TV-Y': 0, | |
3068 | 'TV-Y7': 7, | |
3069 | 'TV-G': 0, | |
3070 | 'TV-PG': 0, | |
3071 | 'TV-14': 14, | |
3072 | 'TV-MA': 17, | |
3073 | } | |
3074 | ||
3075 | ||
3076 | def parse_age_limit(s): | |
3077 | if type(s) == int: | |
3078 | return s if 0 <= s <= 21 else None | |
3079 | if not isinstance(s, compat_basestring): | |
3080 | return None | |
3081 | m = re.match(r'^(?P<age>\d{1,2})\+?$', s) | |
3082 | if m: | |
3083 | return int(m.group('age')) | |
3084 | s = s.upper() | |
3085 | if s in US_RATINGS: | |
3086 | return US_RATINGS[s] | |
3087 | m = re.match(r'^TV[_-]?(%s)$' % '|'.join(k[3:] for k in TV_PARENTAL_GUIDELINES), s) | |
3088 | if m: | |
3089 | return TV_PARENTAL_GUIDELINES['TV-' + m.group(1)] | |
3090 | return None | |
3091 | ||
3092 | ||
3093 | def strip_jsonp(code): | |
3094 | return re.sub( | |
3095 | r'''(?sx)^ | |
3096 | (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*) | |
3097 | (?:\s*&&\s*(?P=func_name))? | |
3098 | \s*\(\s*(?P<callback_data>.*)\);? | |
3099 | \s*?(?://[^\n]*)*$''', | |
3100 | r'\g<callback_data>', code) | |
3101 | ||
3102 | ||
3103 | def js_to_json(code, vars={}): | |
3104 | # vars is a dict of var, val pairs to substitute | |
3105 | COMMENT_RE = r'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n' | |
3106 | SKIP_RE = r'\s*(?:{comment})?\s*'.format(comment=COMMENT_RE) | |
3107 | INTEGER_TABLE = ( | |
3108 | (r'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip=SKIP_RE), 16), | |
3109 | (r'(?s)^(0+[0-7]+){skip}:?$'.format(skip=SKIP_RE), 8), | |
3110 | ) | |
3111 | ||
3112 | def fix_kv(m): | |
3113 | v = m.group(0) | |
3114 | if v in ('true', 'false', 'null'): | |
3115 | return v | |
3116 | elif v in ('undefined', 'void 0'): | |
3117 | return 'null' | |
3118 | elif v.startswith('/*') or v.startswith('//') or v.startswith('!') or v == ',': | |
3119 | return "" | |
3120 | ||
3121 | if v[0] in ("'", '"'): | |
3122 | v = re.sub(r'(?s)\\.|"', lambda m: { | |
3123 | '"': '\\"', | |
3124 | "\\'": "'", | |
3125 | '\\\n': '', | |
3126 | '\\x': '\\u00', | |
3127 | }.get(m.group(0), m.group(0)), v[1:-1]) | |
3128 | else: | |
3129 | for regex, base in INTEGER_TABLE: | |
3130 | im = re.match(regex, v) | |
3131 | if im: | |
3132 | i = int(im.group(1), base) | |
3133 | return '"%d":' % i if v.endswith(':') else '%d' % i | |
3134 | ||
3135 | if v in vars: | |
3136 | return vars[v] | |
3137 | ||
3138 | return '"%s"' % v | |
3139 | ||
3140 | return re.sub(r'''(?sx) | |
3141 | "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"| | |
3142 | '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'| | |
3143 | {comment}|,(?={skip}[\]}}])| | |
3144 | void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*| | |
3145 | \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?| | |
3146 | [0-9]+(?={skip}:)| | |
3147 | !+ | |
3148 | '''.format(comment=COMMENT_RE, skip=SKIP_RE), fix_kv, code) | |
3149 | ||
3150 | ||
3151 | def qualities(quality_ids): | |
3152 | """ Get a numeric quality value out of a list of possible values """ | |
3153 | def q(qid): | |
3154 | try: | |
3155 | return quality_ids.index(qid) | |
3156 | except ValueError: | |
3157 | return -1 | |
3158 | return q | |
3159 | ||
3160 | ||
3161 | POSTPROCESS_WHEN = {'pre_process', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist'} | |
3162 | ||
3163 | ||
3164 | DEFAULT_OUTTMPL = { | |
3165 | 'default': '%(title)s [%(id)s].%(ext)s', | |
3166 | 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s', | |
3167 | } | |
3168 | OUTTMPL_TYPES = { | |
3169 | 'chapter': None, | |
3170 | 'subtitle': None, | |
3171 | 'thumbnail': None, | |
3172 | 'description': 'description', | |
3173 | 'annotation': 'annotations.xml', | |
3174 | 'infojson': 'info.json', | |
3175 | 'link': None, | |
3176 | 'pl_video': None, | |
3177 | 'pl_thumbnail': None, | |
3178 | 'pl_description': 'description', | |
3179 | 'pl_infojson': 'info.json', | |
3180 | } | |
3181 | ||
3182 | # As of [1] format syntax is: | |
3183 | # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type | |
3184 | # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting | |
3185 | STR_FORMAT_RE_TMPL = r'''(?x) | |
3186 | (?<!%)(?P<prefix>(?:%%)*) | |
3187 | % | |
3188 | (?P<has_key>\((?P<key>{0})\))? | |
3189 | (?P<format> | |
3190 | (?P<conversion>[#0\-+ ]+)? | |
3191 | (?P<min_width>\d+)? | |
3192 | (?P<precision>\.\d+)? | |
3193 | (?P<len_mod>[hlL])? # unused in python | |
3194 | {1} # conversion type | |
3195 | ) | |
3196 | ''' | |
3197 | ||
3198 | ||
3199 | STR_FORMAT_TYPES = 'diouxXeEfFgGcrs' | |
3200 | ||
3201 | ||
3202 | def limit_length(s, length): | |
3203 | """ Add ellipses to overly long strings """ | |
3204 | if s is None: | |
3205 | return None | |
3206 | ELLIPSES = '...' | |
3207 | if len(s) > length: | |
3208 | return s[:length - len(ELLIPSES)] + ELLIPSES | |
3209 | return s | |
3210 | ||
3211 | ||
3212 | def version_tuple(v): | |
3213 | return tuple(int(e) for e in re.split(r'[-.]', v)) | |
3214 | ||
3215 | ||
3216 | def is_outdated_version(version, limit, assume_new=True): | |
3217 | if not version: | |
3218 | return not assume_new | |
3219 | try: | |
3220 | return version_tuple(version) < version_tuple(limit) | |
3221 | except ValueError: | |
3222 | return not assume_new | |
3223 | ||
3224 | ||
3225 | def ytdl_is_updateable(): | |
3226 | """ Returns if yt-dlp can be updated with -U """ | |
3227 | ||
3228 | from .update import is_non_updateable | |
3229 | ||
3230 | return not is_non_updateable() | |
3231 | ||
3232 | ||
3233 | def args_to_str(args): | |
3234 | # Get a short string representation for a subprocess command | |
3235 | return ' '.join(compat_shlex_quote(a) for a in args) | |
3236 | ||
3237 | ||
3238 | def error_to_compat_str(err): | |
3239 | err_str = str(err) | |
3240 | # On python 2 error byte string must be decoded with proper | |
3241 | # encoding rather than ascii | |
3242 | if sys.version_info[0] < 3: | |
3243 | err_str = err_str.decode(preferredencoding()) | |
3244 | return err_str | |
3245 | ||
3246 | ||
3247 | def mimetype2ext(mt): | |
3248 | if mt is None: | |
3249 | return None | |
3250 | ||
3251 | mt, _, params = mt.partition(';') | |
3252 | mt = mt.strip() | |
3253 | ||
3254 | FULL_MAP = { | |
3255 | 'audio/mp4': 'm4a', | |
3256 | # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as | |
3257 | # it's the most popular one | |
3258 | 'audio/mpeg': 'mp3', | |
3259 | 'audio/x-wav': 'wav', | |
3260 | 'audio/wav': 'wav', | |
3261 | 'audio/wave': 'wav', | |
3262 | } | |
3263 | ||
3264 | ext = FULL_MAP.get(mt) | |
3265 | if ext is not None: | |
3266 | return ext | |
3267 | ||
3268 | SUBTYPE_MAP = { | |
3269 | '3gpp': '3gp', | |
3270 | 'smptett+xml': 'tt', | |
3271 | 'ttaf+xml': 'dfxp', | |
3272 | 'ttml+xml': 'ttml', | |
3273 | 'x-flv': 'flv', | |
3274 | 'x-mp4-fragmented': 'mp4', | |
3275 | 'x-ms-sami': 'sami', | |
3276 | 'x-ms-wmv': 'wmv', | |
3277 | 'mpegurl': 'm3u8', | |
3278 | 'x-mpegurl': 'm3u8', | |
3279 | 'vnd.apple.mpegurl': 'm3u8', | |
3280 | 'dash+xml': 'mpd', | |
3281 | 'f4m+xml': 'f4m', | |
3282 | 'hds+xml': 'f4m', | |
3283 | 'vnd.ms-sstr+xml': 'ism', | |
3284 | 'quicktime': 'mov', | |
3285 | 'mp2t': 'ts', | |
3286 | 'x-wav': 'wav', | |
3287 | 'filmstrip+json': 'fs', | |
3288 | 'svg+xml': 'svg', | |
3289 | } | |
3290 | ||
3291 | _, _, subtype = mt.rpartition('/') | |
3292 | ext = SUBTYPE_MAP.get(subtype.lower()) | |
3293 | if ext is not None: | |
3294 | return ext | |
3295 | ||
3296 | SUFFIX_MAP = { | |
3297 | 'json': 'json', | |
3298 | 'xml': 'xml', | |
3299 | 'zip': 'zip', | |
3300 | 'gzip': 'gz', | |
3301 | } | |
3302 | ||
3303 | _, _, suffix = subtype.partition('+') | |
3304 | ext = SUFFIX_MAP.get(suffix) | |
3305 | if ext is not None: | |
3306 | return ext | |
3307 | ||
3308 | return subtype.replace('+', '.') | |
3309 | ||
3310 | ||
3311 | def ext2mimetype(ext_or_url): | |
3312 | if not ext_or_url: | |
3313 | return None | |
3314 | if '.' not in ext_or_url: | |
3315 | ext_or_url = f'file.{ext_or_url}' | |
3316 | return mimetypes.guess_type(ext_or_url)[0] | |
3317 | ||
3318 | ||
3319 | def parse_codecs(codecs_str): | |
3320 | # http://tools.ietf.org/html/rfc6381 | |
3321 | if not codecs_str: | |
3322 | return {} | |
3323 | split_codecs = list(filter(None, map( | |
3324 | str.strip, codecs_str.strip().strip(',').split(',')))) | |
3325 | vcodec, acodec, tcodec, hdr = None, None, None, None | |
3326 | for full_codec in split_codecs: | |
3327 | parts = full_codec.split('.') | |
3328 | codec = parts[0].replace('0', '') | |
3329 | if codec in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', | |
3330 | 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'): | |
3331 | if not vcodec: | |
3332 | vcodec = '.'.join(parts[:4]) if codec in ('vp9', 'av1', 'hvc1') else full_codec | |
3333 | if codec in ('dvh1', 'dvhe'): | |
3334 | hdr = 'DV' | |
3335 | elif codec == 'av1' and len(parts) > 3 and parts[3] == '10': | |
3336 | hdr = 'HDR10' | |
3337 | elif full_codec.replace('0', '').startswith('vp9.2'): | |
3338 | hdr = 'HDR10' | |
3339 | elif codec in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'): | |
3340 | if not acodec: | |
3341 | acodec = full_codec | |
3342 | elif codec in ('stpp', 'wvtt',): | |
3343 | if not tcodec: | |
3344 | tcodec = full_codec | |
3345 | else: | |
3346 | write_string('WARNING: Unknown codec %s\n' % full_codec, sys.stderr) | |
3347 | if vcodec or acodec or tcodec: | |
3348 | return { | |
3349 | 'vcodec': vcodec or 'none', | |
3350 | 'acodec': acodec or 'none', | |
3351 | 'dynamic_range': hdr, | |
3352 | **({'tcodec': tcodec} if tcodec is not None else {}), | |
3353 | } | |
3354 | elif len(split_codecs) == 2: | |
3355 | return { | |
3356 | 'vcodec': split_codecs[0], | |
3357 | 'acodec': split_codecs[1], | |
3358 | } | |
3359 | return {} | |
3360 | ||
3361 | ||
3362 | def urlhandle_detect_ext(url_handle): | |
3363 | getheader = url_handle.headers.get | |
3364 | ||
3365 | cd = getheader('Content-Disposition') | |
3366 | if cd: | |
3367 | m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd) | |
3368 | if m: | |
3369 | e = determine_ext(m.group('filename'), default_ext=None) | |
3370 | if e: | |
3371 | return e | |
3372 | ||
3373 | return mimetype2ext(getheader('Content-Type')) | |
3374 | ||
3375 | ||
3376 | def encode_data_uri(data, mime_type): | |
3377 | return 'data:%s;base64,%s' % (mime_type, base64.b64encode(data).decode('ascii')) | |
3378 | ||
3379 | ||
3380 | def age_restricted(content_limit, age_limit): | |
3381 | """ Returns True iff the content should be blocked """ | |
3382 | ||
3383 | if age_limit is None: # No limit set | |
3384 | return False | |
3385 | if content_limit is None: | |
3386 | return False # Content available for everyone | |
3387 | return age_limit < content_limit | |
3388 | ||
3389 | ||
3390 | def is_html(first_bytes): | |
3391 | """ Detect whether a file contains HTML by examining its first bytes. """ | |
3392 | ||
3393 | BOMS = [ | |
3394 | (b'\xef\xbb\xbf', 'utf-8'), | |
3395 | (b'\x00\x00\xfe\xff', 'utf-32-be'), | |
3396 | (b'\xff\xfe\x00\x00', 'utf-32-le'), | |
3397 | (b'\xff\xfe', 'utf-16-le'), | |
3398 | (b'\xfe\xff', 'utf-16-be'), | |
3399 | ] | |
3400 | for bom, enc in BOMS: | |
3401 | if first_bytes.startswith(bom): | |
3402 | s = first_bytes[len(bom):].decode(enc, 'replace') | |
3403 | break | |
3404 | else: | |
3405 | s = first_bytes.decode('utf-8', 'replace') | |
3406 | ||
3407 | return re.match(r'^\s*<', s) | |
3408 | ||
3409 | ||
3410 | def determine_protocol(info_dict): | |
3411 | protocol = info_dict.get('protocol') | |
3412 | if protocol is not None: | |
3413 | return protocol | |
3414 | ||
3415 | url = sanitize_url(info_dict['url']) | |
3416 | if url.startswith('rtmp'): | |
3417 | return 'rtmp' | |
3418 | elif url.startswith('mms'): | |
3419 | return 'mms' | |
3420 | elif url.startswith('rtsp'): | |
3421 | return 'rtsp' | |
3422 | ||
3423 | ext = determine_ext(url) | |
3424 | if ext == 'm3u8': | |
3425 | return 'm3u8' | |
3426 | elif ext == 'f4m': | |
3427 | return 'f4m' | |
3428 | ||
3429 | return compat_urllib_parse_urlparse(url).scheme | |
3430 | ||
3431 | ||
3432 | def render_table(header_row, data, delim=False, extra_gap=0, hide_empty=False): | |
3433 | """ Render a list of rows, each as a list of values. | |
3434 | Text after a \t will be right aligned """ | |
3435 | def width(string): | |
3436 | return len(remove_terminal_sequences(string).replace('\t', '')) | |
3437 | ||
3438 | def get_max_lens(table): | |
3439 | return [max(width(str(v)) for v in col) for col in zip(*table)] | |
3440 | ||
3441 | def filter_using_list(row, filterArray): | |
3442 | return [col for take, col in itertools.zip_longest(filterArray, row, fillvalue=True) if take] | |
3443 | ||
3444 | max_lens = get_max_lens(data) if hide_empty else [] | |
3445 | header_row = filter_using_list(header_row, max_lens) | |
3446 | data = [filter_using_list(row, max_lens) for row in data] | |
3447 | ||
3448 | table = [header_row] + data | |
3449 | max_lens = get_max_lens(table) | |
3450 | extra_gap += 1 | |
3451 | if delim: | |
3452 | table = [header_row, [delim * (ml + extra_gap) for ml in max_lens]] + data | |
3453 | table[1][-1] = table[1][-1][:-extra_gap] # Remove extra_gap from end of delimiter | |
3454 | for row in table: | |
3455 | for pos, text in enumerate(map(str, row)): | |
3456 | if '\t' in text: | |
3457 | row[pos] = text.replace('\t', ' ' * (max_lens[pos] - width(text))) + ' ' * extra_gap | |
3458 | else: | |
3459 | row[pos] = text + ' ' * (max_lens[pos] - width(text) + extra_gap) | |
3460 | ret = '\n'.join(''.join(row).rstrip() for row in table) | |
3461 | return ret | |
3462 | ||
3463 | ||
3464 | def _match_one(filter_part, dct, incomplete): | |
3465 | # TODO: Generalize code with YoutubeDL._build_format_filter | |
3466 | STRING_OPERATORS = { | |
3467 | '*=': operator.contains, | |
3468 | '^=': lambda attr, value: attr.startswith(value), | |
3469 | '$=': lambda attr, value: attr.endswith(value), | |
3470 | '~=': lambda attr, value: re.search(value, attr), | |
3471 | } | |
3472 | COMPARISON_OPERATORS = { | |
3473 | **STRING_OPERATORS, | |
3474 | '<=': operator.le, # "<=" must be defined above "<" | |
3475 | '<': operator.lt, | |
3476 | '>=': operator.ge, | |
3477 | '>': operator.gt, | |
3478 | '=': operator.eq, | |
3479 | } | |
3480 | ||
3481 | operator_rex = re.compile(r'''(?x)\s* | |
3482 | (?P<key>[a-z_]+) | |
3483 | \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s* | |
3484 | (?: | |
3485 | (?P<quote>["\'])(?P<quotedstrval>.+?)(?P=quote)| | |
3486 | (?P<strval>.+?) | |
3487 | ) | |
3488 | \s*$ | |
3489 | ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys()))) | |
3490 | m = operator_rex.search(filter_part) | |
3491 | if m: | |
3492 | m = m.groupdict() | |
3493 | unnegated_op = COMPARISON_OPERATORS[m['op']] | |
3494 | if m['negation']: | |
3495 | op = lambda attr, value: not unnegated_op(attr, value) | |
3496 | else: | |
3497 | op = unnegated_op | |
3498 | comparison_value = m['quotedstrval'] or m['strval'] or m['intval'] | |
3499 | if m['quote']: | |
3500 | comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote']) | |
3501 | actual_value = dct.get(m['key']) | |
3502 | numeric_comparison = None | |
3503 | if isinstance(actual_value, compat_numeric_types): | |
3504 | # If the original field is a string and matching comparisonvalue is | |
3505 | # a number we should respect the origin of the original field | |
3506 | # and process comparison value as a string (see | |
3507 | # https://github.com/ytdl-org/youtube-dl/issues/11082) | |
3508 | try: | |
3509 | numeric_comparison = int(comparison_value) | |
3510 | except ValueError: | |
3511 | numeric_comparison = parse_filesize(comparison_value) | |
3512 | if numeric_comparison is None: | |
3513 | numeric_comparison = parse_filesize(f'{comparison_value}B') | |
3514 | if numeric_comparison is None: | |
3515 | numeric_comparison = parse_duration(comparison_value) | |
3516 | if numeric_comparison is not None and m['op'] in STRING_OPERATORS: | |
3517 | raise ValueError('Operator %s only supports string values!' % m['op']) | |
3518 | if actual_value is None: | |
3519 | return incomplete or m['none_inclusive'] | |
3520 | return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison) | |
3521 | ||
3522 | UNARY_OPERATORS = { | |
3523 | '': lambda v: (v is True) if isinstance(v, bool) else (v is not None), | |
3524 | '!': lambda v: (v is False) if isinstance(v, bool) else (v is None), | |
3525 | } | |
3526 | operator_rex = re.compile(r'''(?x)\s* | |
3527 | (?P<op>%s)\s*(?P<key>[a-z_]+) | |
3528 | \s*$ | |
3529 | ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys()))) | |
3530 | m = operator_rex.search(filter_part) | |
3531 | if m: | |
3532 | op = UNARY_OPERATORS[m.group('op')] | |
3533 | actual_value = dct.get(m.group('key')) | |
3534 | if incomplete and actual_value is None: | |
3535 | return True | |
3536 | return op(actual_value) | |
3537 | ||
3538 | raise ValueError('Invalid filter part %r' % filter_part) | |
3539 | ||
3540 | ||
3541 | def match_str(filter_str, dct, incomplete=False): | |
3542 | """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false | |
3543 | When incomplete, all conditions passes on missing fields | |
3544 | """ | |
3545 | return all( | |
3546 | _match_one(filter_part.replace(r'\&', '&'), dct, incomplete) | |
3547 | for filter_part in re.split(r'(?<!\\)&', filter_str)) | |
3548 | ||
3549 | ||
3550 | def match_filter_func(filter_str): | |
3551 | def _match_func(info_dict, *args, **kwargs): | |
3552 | if match_str(filter_str, info_dict, *args, **kwargs): | |
3553 | return None | |
3554 | else: | |
3555 | video_title = info_dict.get('title', info_dict.get('id', 'video')) | |
3556 | return '%s does not pass filter %s, skipping ..' % (video_title, filter_str) | |
3557 | return _match_func | |
3558 | ||
3559 | ||
3560 | def parse_dfxp_time_expr(time_expr): | |
3561 | if not time_expr: | |
3562 | return | |
3563 | ||
3564 | mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr) | |
3565 | if mobj: | |
3566 | return float(mobj.group('time_offset')) | |
3567 | ||
3568 | mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr) | |
3569 | if mobj: | |
3570 | return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.')) | |
3571 | ||
3572 | ||
3573 | def srt_subtitles_timecode(seconds): | |
3574 | return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000) | |
3575 | ||
3576 | ||
3577 | def ass_subtitles_timecode(seconds): | |
3578 | time = timetuple_from_msec(seconds * 1000) | |
3579 | return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10) | |
3580 | ||
3581 | ||
3582 | def dfxp2srt(dfxp_data): | |
3583 | ''' | |
3584 | @param dfxp_data A bytes-like object containing DFXP data | |
3585 | @returns A unicode object containing converted SRT data | |
3586 | ''' | |
3587 | LEGACY_NAMESPACES = ( | |
3588 | (b'http://www.w3.org/ns/ttml', [ | |
3589 | b'http://www.w3.org/2004/11/ttaf1', | |
3590 | b'http://www.w3.org/2006/04/ttaf1', | |
3591 | b'http://www.w3.org/2006/10/ttaf1', | |
3592 | ]), | |
3593 | (b'http://www.w3.org/ns/ttml#styling', [ | |
3594 | b'http://www.w3.org/ns/ttml#style', | |
3595 | ]), | |
3596 | ) | |
3597 | ||
3598 | SUPPORTED_STYLING = [ | |
3599 | 'color', | |
3600 | 'fontFamily', | |
3601 | 'fontSize', | |
3602 | 'fontStyle', | |
3603 | 'fontWeight', | |
3604 | 'textDecoration' | |
3605 | ] | |
3606 | ||
3607 | _x = functools.partial(xpath_with_ns, ns_map={ | |
3608 | 'xml': 'http://www.w3.org/XML/1998/namespace', | |
3609 | 'ttml': 'http://www.w3.org/ns/ttml', | |
3610 | 'tts': 'http://www.w3.org/ns/ttml#styling', | |
3611 | }) | |
3612 | ||
3613 | styles = {} | |
3614 | default_style = {} | |
3615 | ||
3616 | class TTMLPElementParser(object): | |
3617 | _out = '' | |
3618 | _unclosed_elements = [] | |
3619 | _applied_styles = [] | |
3620 | ||
3621 | def start(self, tag, attrib): | |
3622 | if tag in (_x('ttml:br'), 'br'): | |
3623 | self._out += '\n' | |
3624 | else: | |
3625 | unclosed_elements = [] | |
3626 | style = {} | |
3627 | element_style_id = attrib.get('style') | |
3628 | if default_style: | |
3629 | style.update(default_style) | |
3630 | if element_style_id: | |
3631 | style.update(styles.get(element_style_id, {})) | |
3632 | for prop in SUPPORTED_STYLING: | |
3633 | prop_val = attrib.get(_x('tts:' + prop)) | |
3634 | if prop_val: | |
3635 | style[prop] = prop_val | |
3636 | if style: | |
3637 | font = '' | |
3638 | for k, v in sorted(style.items()): | |
3639 | if self._applied_styles and self._applied_styles[-1].get(k) == v: | |
3640 | continue | |
3641 | if k == 'color': | |
3642 | font += ' color="%s"' % v | |
3643 | elif k == 'fontSize': | |
3644 | font += ' size="%s"' % v | |
3645 | elif k == 'fontFamily': | |
3646 | font += ' face="%s"' % v | |
3647 | elif k == 'fontWeight' and v == 'bold': | |
3648 | self._out += '<b>' | |
3649 | unclosed_elements.append('b') | |
3650 | elif k == 'fontStyle' and v == 'italic': | |
3651 | self._out += '<i>' | |
3652 | unclosed_elements.append('i') | |
3653 | elif k == 'textDecoration' and v == 'underline': | |
3654 | self._out += '<u>' | |
3655 | unclosed_elements.append('u') | |
3656 | if font: | |
3657 | self._out += '<font' + font + '>' | |
3658 | unclosed_elements.append('font') | |
3659 | applied_style = {} | |
3660 | if self._applied_styles: | |
3661 | applied_style.update(self._applied_styles[-1]) | |
3662 | applied_style.update(style) | |
3663 | self._applied_styles.append(applied_style) | |
3664 | self._unclosed_elements.append(unclosed_elements) | |
3665 | ||
3666 | def end(self, tag): | |
3667 | if tag not in (_x('ttml:br'), 'br'): | |
3668 | unclosed_elements = self._unclosed_elements.pop() | |
3669 | for element in reversed(unclosed_elements): | |
3670 | self._out += '</%s>' % element | |
3671 | if unclosed_elements and self._applied_styles: | |
3672 | self._applied_styles.pop() | |
3673 | ||
3674 | def data(self, data): | |
3675 | self._out += data | |
3676 | ||
3677 | def close(self): | |
3678 | return self._out.strip() | |
3679 | ||
3680 | def parse_node(node): | |
3681 | target = TTMLPElementParser() | |
3682 | parser = xml.etree.ElementTree.XMLParser(target=target) | |
3683 | parser.feed(xml.etree.ElementTree.tostring(node)) | |
3684 | return parser.close() | |
3685 | ||
3686 | for k, v in LEGACY_NAMESPACES: | |
3687 | for ns in v: | |
3688 | dfxp_data = dfxp_data.replace(ns, k) | |
3689 | ||
3690 | dfxp = compat_etree_fromstring(dfxp_data) | |
3691 | out = [] | |
3692 | paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p') | |
3693 | ||
3694 | if not paras: | |
3695 | raise ValueError('Invalid dfxp/TTML subtitle') | |
3696 | ||
3697 | repeat = False | |
3698 | while True: | |
3699 | for style in dfxp.findall(_x('.//ttml:style')): | |
3700 | style_id = style.get('id') or style.get(_x('xml:id')) | |
3701 | if not style_id: | |
3702 | continue | |
3703 | parent_style_id = style.get('style') | |
3704 | if parent_style_id: | |
3705 | if parent_style_id not in styles: | |
3706 | repeat = True | |
3707 | continue | |
3708 | styles[style_id] = styles[parent_style_id].copy() | |
3709 | for prop in SUPPORTED_STYLING: | |
3710 | prop_val = style.get(_x('tts:' + prop)) | |
3711 | if prop_val: | |
3712 | styles.setdefault(style_id, {})[prop] = prop_val | |
3713 | if repeat: | |
3714 | repeat = False | |
3715 | else: | |
3716 | break | |
3717 | ||
3718 | for p in ('body', 'div'): | |
3719 | ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p]) | |
3720 | if ele is None: | |
3721 | continue | |
3722 | style = styles.get(ele.get('style')) | |
3723 | if not style: | |
3724 | continue | |
3725 | default_style.update(style) | |
3726 | ||
3727 | for para, index in zip(paras, itertools.count(1)): | |
3728 | begin_time = parse_dfxp_time_expr(para.attrib.get('begin')) | |
3729 | end_time = parse_dfxp_time_expr(para.attrib.get('end')) | |
3730 | dur = parse_dfxp_time_expr(para.attrib.get('dur')) | |
3731 | if begin_time is None: | |
3732 | continue | |
3733 | if not end_time: | |
3734 | if not dur: | |
3735 | continue | |
3736 | end_time = begin_time + dur | |
3737 | out.append('%d\n%s --> %s\n%s\n\n' % ( | |
3738 | index, | |
3739 | srt_subtitles_timecode(begin_time), | |
3740 | srt_subtitles_timecode(end_time), | |
3741 | parse_node(para))) | |
3742 | ||
3743 | return ''.join(out) | |
3744 | ||
3745 | ||
3746 | def cli_option(params, command_option, param): | |
3747 | param = params.get(param) | |
3748 | if param: | |
3749 | param = compat_str(param) | |
3750 | return [command_option, param] if param is not None else [] | |
3751 | ||
3752 | ||
3753 | def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None): | |
3754 | param = params.get(param) | |
3755 | if param is None: | |
3756 | return [] | |
3757 | assert isinstance(param, bool) | |
3758 | if separator: | |
3759 | return [command_option + separator + (true_value if param else false_value)] | |
3760 | return [command_option, true_value if param else false_value] | |
3761 | ||
3762 | ||
3763 | def cli_valueless_option(params, command_option, param, expected_value=True): | |
3764 | param = params.get(param) | |
3765 | return [command_option] if param == expected_value else [] | |
3766 | ||
3767 | ||
3768 | def cli_configuration_args(argdict, keys, default=[], use_compat=True): | |
3769 | if isinstance(argdict, (list, tuple)): # for backward compatibility | |
3770 | if use_compat: | |
3771 | return argdict | |
3772 | else: | |
3773 | argdict = None | |
3774 | if argdict is None: | |
3775 | return default | |
3776 | assert isinstance(argdict, dict) | |
3777 | ||
3778 | assert isinstance(keys, (list, tuple)) | |
3779 | for key_list in keys: | |
3780 | arg_list = list(filter( | |
3781 | lambda x: x is not None, | |
3782 | [argdict.get(key.lower()) for key in variadic(key_list)])) | |
3783 | if arg_list: | |
3784 | return [arg for args in arg_list for arg in args] | |
3785 | return default | |
3786 | ||
3787 | ||
3788 | def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True): | |
3789 | main_key, exe = main_key.lower(), exe.lower() | |
3790 | root_key = exe if main_key == exe else f'{main_key}+{exe}' | |
3791 | keys = [f'{root_key}{k}' for k in (keys or [''])] | |
3792 | if root_key in keys: | |
3793 | if main_key != exe: | |
3794 | keys.append((main_key, exe)) | |
3795 | keys.append('default') | |
3796 | else: | |
3797 | use_compat = False | |
3798 | return cli_configuration_args(argdict, keys, default, use_compat) | |
3799 | ||
3800 | ||
3801 | class ISO639Utils(object): | |
3802 | # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt | |
3803 | _lang_map = { | |
3804 | 'aa': 'aar', | |
3805 | 'ab': 'abk', | |
3806 | 'ae': 'ave', | |
3807 | 'af': 'afr', | |
3808 | 'ak': 'aka', | |
3809 | 'am': 'amh', | |
3810 | 'an': 'arg', | |
3811 | 'ar': 'ara', | |
3812 | 'as': 'asm', | |
3813 | 'av': 'ava', | |
3814 | 'ay': 'aym', | |
3815 | 'az': 'aze', | |
3816 | 'ba': 'bak', | |
3817 | 'be': 'bel', | |
3818 | 'bg': 'bul', | |
3819 | 'bh': 'bih', | |
3820 | 'bi': 'bis', | |
3821 | 'bm': 'bam', | |
3822 | 'bn': 'ben', | |
3823 | 'bo': 'bod', | |
3824 | 'br': 'bre', | |
3825 | 'bs': 'bos', | |
3826 | 'ca': 'cat', | |
3827 | 'ce': 'che', | |
3828 | 'ch': 'cha', | |
3829 | 'co': 'cos', | |
3830 | 'cr': 'cre', | |
3831 | 'cs': 'ces', | |
3832 | 'cu': 'chu', | |
3833 | 'cv': 'chv', | |
3834 | 'cy': 'cym', | |
3835 | 'da': 'dan', | |
3836 | 'de': 'deu', | |
3837 | 'dv': 'div', | |
3838 | 'dz': 'dzo', | |
3839 | 'ee': 'ewe', | |
3840 | 'el': 'ell', | |
3841 | 'en': 'eng', | |
3842 | 'eo': 'epo', | |
3843 | 'es': 'spa', | |
3844 | 'et': 'est', | |
3845 | 'eu': 'eus', | |
3846 | 'fa': 'fas', | |
3847 | 'ff': 'ful', | |
3848 | 'fi': 'fin', | |
3849 | 'fj': 'fij', | |
3850 | 'fo': 'fao', | |
3851 | 'fr': 'fra', | |
3852 | 'fy': 'fry', | |
3853 | 'ga': 'gle', | |
3854 | 'gd': 'gla', | |
3855 | 'gl': 'glg', | |
3856 | 'gn': 'grn', | |
3857 | 'gu': 'guj', | |
3858 | 'gv': 'glv', | |
3859 | 'ha': 'hau', | |
3860 | 'he': 'heb', | |
3861 | 'iw': 'heb', # Replaced by he in 1989 revision | |
3862 | 'hi': 'hin', | |
3863 | 'ho': 'hmo', | |
3864 | 'hr': 'hrv', | |
3865 | 'ht': 'hat', | |
3866 | 'hu': 'hun', | |
3867 | 'hy': 'hye', | |
3868 | 'hz': 'her', | |
3869 | 'ia': 'ina', | |
3870 | 'id': 'ind', | |
3871 | 'in': 'ind', # Replaced by id in 1989 revision | |
3872 | 'ie': 'ile', | |
3873 | 'ig': 'ibo', | |
3874 | 'ii': 'iii', | |
3875 | 'ik': 'ipk', | |
3876 | 'io': 'ido', | |
3877 | 'is': 'isl', | |
3878 | 'it': 'ita', | |
3879 | 'iu': 'iku', | |
3880 | 'ja': 'jpn', | |
3881 | 'jv': 'jav', | |
3882 | 'ka': 'kat', | |
3883 | 'kg': 'kon', | |
3884 | 'ki': 'kik', | |
3885 | 'kj': 'kua', | |
3886 | 'kk': 'kaz', | |
3887 | 'kl': 'kal', | |
3888 | 'km': 'khm', | |
3889 | 'kn': 'kan', | |
3890 | 'ko': 'kor', | |
3891 | 'kr': 'kau', | |
3892 | 'ks': 'kas', | |
3893 | 'ku': 'kur', | |
3894 | 'kv': 'kom', | |
3895 | 'kw': 'cor', | |
3896 | 'ky': 'kir', | |
3897 | 'la': 'lat', | |
3898 | 'lb': 'ltz', | |
3899 | 'lg': 'lug', | |
3900 | 'li': 'lim', | |
3901 | 'ln': 'lin', | |
3902 | 'lo': 'lao', | |
3903 | 'lt': 'lit', | |
3904 | 'lu': 'lub', | |
3905 | 'lv': 'lav', | |
3906 | 'mg': 'mlg', | |
3907 | 'mh': 'mah', | |
3908 | 'mi': 'mri', | |
3909 | 'mk': 'mkd', | |
3910 | 'ml': 'mal', | |
3911 | 'mn': 'mon', | |
3912 | 'mr': 'mar', | |
3913 | 'ms': 'msa', | |
3914 | 'mt': 'mlt', | |
3915 | 'my': 'mya', | |
3916 | 'na': 'nau', | |
3917 | 'nb': 'nob', | |
3918 | 'nd': 'nde', | |
3919 | 'ne': 'nep', | |
3920 | 'ng': 'ndo', | |
3921 | 'nl': 'nld', | |
3922 | 'nn': 'nno', | |
3923 | 'no': 'nor', | |
3924 | 'nr': 'nbl', | |
3925 | 'nv': 'nav', | |
3926 | 'ny': 'nya', | |
3927 | 'oc': 'oci', | |
3928 | 'oj': 'oji', | |
3929 | 'om': 'orm', | |
3930 | 'or': 'ori', | |
3931 | 'os': 'oss', | |
3932 | 'pa': 'pan', | |
3933 | 'pi': 'pli', | |
3934 | 'pl': 'pol', | |
3935 | 'ps': 'pus', | |
3936 | 'pt': 'por', | |
3937 | 'qu': 'que', | |
3938 | 'rm': 'roh', | |
3939 | 'rn': 'run', | |
3940 | 'ro': 'ron', | |
3941 | 'ru': 'rus', | |
3942 | 'rw': 'kin', | |
3943 | 'sa': 'san', | |
3944 | 'sc': 'srd', | |
3945 | 'sd': 'snd', | |
3946 | 'se': 'sme', | |
3947 | 'sg': 'sag', | |
3948 | 'si': 'sin', | |
3949 | 'sk': 'slk', | |
3950 | 'sl': 'slv', | |
3951 | 'sm': 'smo', | |
3952 | 'sn': 'sna', | |
3953 | 'so': 'som', | |
3954 | 'sq': 'sqi', | |
3955 | 'sr': 'srp', | |
3956 | 'ss': 'ssw', | |
3957 | 'st': 'sot', | |
3958 | 'su': 'sun', | |
3959 | 'sv': 'swe', | |
3960 | 'sw': 'swa', | |
3961 | 'ta': 'tam', | |
3962 | 'te': 'tel', | |
3963 | 'tg': 'tgk', | |
3964 | 'th': 'tha', | |
3965 | 'ti': 'tir', | |
3966 | 'tk': 'tuk', | |
3967 | 'tl': 'tgl', | |
3968 | 'tn': 'tsn', | |
3969 | 'to': 'ton', | |
3970 | 'tr': 'tur', | |
3971 | 'ts': 'tso', | |
3972 | 'tt': 'tat', | |
3973 | 'tw': 'twi', | |
3974 | 'ty': 'tah', | |
3975 | 'ug': 'uig', | |
3976 | 'uk': 'ukr', | |
3977 | 'ur': 'urd', | |
3978 | 'uz': 'uzb', | |
3979 | 've': 'ven', | |
3980 | 'vi': 'vie', | |
3981 | 'vo': 'vol', | |
3982 | 'wa': 'wln', | |
3983 | 'wo': 'wol', | |
3984 | 'xh': 'xho', | |
3985 | 'yi': 'yid', | |
3986 | 'ji': 'yid', # Replaced by yi in 1989 revision | |
3987 | 'yo': 'yor', | |
3988 | 'za': 'zha', | |
3989 | 'zh': 'zho', | |
3990 | 'zu': 'zul', | |
3991 | } | |
3992 | ||
3993 | @classmethod | |
3994 | def short2long(cls, code): | |
3995 | """Convert language code from ISO 639-1 to ISO 639-2/T""" | |
3996 | return cls._lang_map.get(code[:2]) | |
3997 | ||
3998 | @classmethod | |
3999 | def long2short(cls, code): | |
4000 | """Convert language code from ISO 639-2/T to ISO 639-1""" | |
4001 | for short_name, long_name in cls._lang_map.items(): | |
4002 | if long_name == code: | |
4003 | return short_name | |
4004 | ||
4005 | ||
4006 | class ISO3166Utils(object): | |
4007 | # From http://data.okfn.org/data/core/country-list | |
4008 | _country_map = { | |
4009 | 'AF': 'Afghanistan', | |
4010 | 'AX': 'Åland Islands', | |
4011 | 'AL': 'Albania', | |
4012 | 'DZ': 'Algeria', | |
4013 | 'AS': 'American Samoa', | |
4014 | 'AD': 'Andorra', | |
4015 | 'AO': 'Angola', | |
4016 | 'AI': 'Anguilla', | |
4017 | 'AQ': 'Antarctica', | |
4018 | 'AG': 'Antigua and Barbuda', | |
4019 | 'AR': 'Argentina', | |
4020 | 'AM': 'Armenia', | |
4021 | 'AW': 'Aruba', | |
4022 | 'AU': 'Australia', | |
4023 | 'AT': 'Austria', | |
4024 | 'AZ': 'Azerbaijan', | |
4025 | 'BS': 'Bahamas', | |
4026 | 'BH': 'Bahrain', | |
4027 | 'BD': 'Bangladesh', | |
4028 | 'BB': 'Barbados', | |
4029 | 'BY': 'Belarus', | |
4030 | 'BE': 'Belgium', | |
4031 | 'BZ': 'Belize', | |
4032 | 'BJ': 'Benin', | |
4033 | 'BM': 'Bermuda', | |
4034 | 'BT': 'Bhutan', | |
4035 | 'BO': 'Bolivia, Plurinational State of', | |
4036 | 'BQ': 'Bonaire, Sint Eustatius and Saba', | |
4037 | 'BA': 'Bosnia and Herzegovina', | |
4038 | 'BW': 'Botswana', | |
4039 | 'BV': 'Bouvet Island', | |
4040 | 'BR': 'Brazil', | |
4041 | 'IO': 'British Indian Ocean Territory', | |
4042 | 'BN': 'Brunei Darussalam', | |
4043 | 'BG': 'Bulgaria', | |
4044 | 'BF': 'Burkina Faso', | |
4045 | 'BI': 'Burundi', | |
4046 | 'KH': 'Cambodia', | |
4047 | 'CM': 'Cameroon', | |
4048 | 'CA': 'Canada', | |
4049 | 'CV': 'Cape Verde', | |
4050 | 'KY': 'Cayman Islands', | |
4051 | 'CF': 'Central African Republic', | |
4052 | 'TD': 'Chad', | |
4053 | 'CL': 'Chile', | |
4054 | 'CN': 'China', | |
4055 | 'CX': 'Christmas Island', | |
4056 | 'CC': 'Cocos (Keeling) Islands', | |
4057 | 'CO': 'Colombia', | |
4058 | 'KM': 'Comoros', | |
4059 | 'CG': 'Congo', | |
4060 | 'CD': 'Congo, the Democratic Republic of the', | |
4061 | 'CK': 'Cook Islands', | |
4062 | 'CR': 'Costa Rica', | |
4063 | 'CI': 'Côte d\'Ivoire', | |
4064 | 'HR': 'Croatia', | |
4065 | 'CU': 'Cuba', | |
4066 | 'CW': 'Curaçao', | |
4067 | 'CY': 'Cyprus', | |
4068 | 'CZ': 'Czech Republic', | |
4069 | 'DK': 'Denmark', | |
4070 | 'DJ': 'Djibouti', | |
4071 | 'DM': 'Dominica', | |
4072 | 'DO': 'Dominican Republic', | |
4073 | 'EC': 'Ecuador', | |
4074 | 'EG': 'Egypt', | |
4075 | 'SV': 'El Salvador', | |
4076 | 'GQ': 'Equatorial Guinea', | |
4077 | 'ER': 'Eritrea', | |
4078 | 'EE': 'Estonia', | |
4079 | 'ET': 'Ethiopia', | |
4080 | 'FK': 'Falkland Islands (Malvinas)', | |
4081 | 'FO': 'Faroe Islands', | |
4082 | 'FJ': 'Fiji', | |
4083 | 'FI': 'Finland', | |
4084 | 'FR': 'France', | |
4085 | 'GF': 'French Guiana', | |
4086 | 'PF': 'French Polynesia', | |
4087 | 'TF': 'French Southern Territories', | |
4088 | 'GA': 'Gabon', | |
4089 | 'GM': 'Gambia', | |
4090 | 'GE': 'Georgia', | |
4091 | 'DE': 'Germany', | |
4092 | 'GH': 'Ghana', | |
4093 | 'GI': 'Gibraltar', | |
4094 | 'GR': 'Greece', | |
4095 | 'GL': 'Greenland', | |
4096 | 'GD': 'Grenada', | |
4097 | 'GP': 'Guadeloupe', | |
4098 | 'GU': 'Guam', | |
4099 | 'GT': 'Guatemala', | |
4100 | 'GG': 'Guernsey', | |
4101 | 'GN': 'Guinea', | |
4102 | 'GW': 'Guinea-Bissau', | |
4103 | 'GY': 'Guyana', | |
4104 | 'HT': 'Haiti', | |
4105 | 'HM': 'Heard Island and McDonald Islands', | |
4106 | 'VA': 'Holy See (Vatican City State)', | |
4107 | 'HN': 'Honduras', | |
4108 | 'HK': 'Hong Kong', | |
4109 | 'HU': 'Hungary', | |
4110 | 'IS': 'Iceland', | |
4111 | 'IN': 'India', | |
4112 | 'ID': 'Indonesia', | |
4113 | 'IR': 'Iran, Islamic Republic of', | |
4114 | 'IQ': 'Iraq', | |
4115 | 'IE': 'Ireland', | |
4116 | 'IM': 'Isle of Man', | |
4117 | 'IL': 'Israel', | |
4118 | 'IT': 'Italy', | |
4119 | 'JM': 'Jamaica', | |
4120 | 'JP': 'Japan', | |
4121 | 'JE': 'Jersey', | |
4122 | 'JO': 'Jordan', | |
4123 | 'KZ': 'Kazakhstan', | |
4124 | 'KE': 'Kenya', | |
4125 | 'KI': 'Kiribati', | |
4126 | 'KP': 'Korea, Democratic People\'s Republic of', | |
4127 | 'KR': 'Korea, Republic of', | |
4128 | 'KW': 'Kuwait', | |
4129 | 'KG': 'Kyrgyzstan', | |
4130 | 'LA': 'Lao People\'s Democratic Republic', | |
4131 | 'LV': 'Latvia', | |
4132 | 'LB': 'Lebanon', | |
4133 | 'LS': 'Lesotho', | |
4134 | 'LR': 'Liberia', | |
4135 | 'LY': 'Libya', | |
4136 | 'LI': 'Liechtenstein', | |
4137 | 'LT': 'Lithuania', | |
4138 | 'LU': 'Luxembourg', | |
4139 | 'MO': 'Macao', | |
4140 | 'MK': 'Macedonia, the Former Yugoslav Republic of', | |
4141 | 'MG': 'Madagascar', | |
4142 | 'MW': 'Malawi', | |
4143 | 'MY': 'Malaysia', | |
4144 | 'MV': 'Maldives', | |
4145 | 'ML': 'Mali', | |
4146 | 'MT': 'Malta', | |
4147 | 'MH': 'Marshall Islands', | |
4148 | 'MQ': 'Martinique', | |
4149 | 'MR': 'Mauritania', | |
4150 | 'MU': 'Mauritius', | |
4151 | 'YT': 'Mayotte', | |
4152 | 'MX': 'Mexico', | |
4153 | 'FM': 'Micronesia, Federated States of', | |
4154 | 'MD': 'Moldova, Republic of', | |
4155 | 'MC': 'Monaco', | |
4156 | 'MN': 'Mongolia', | |
4157 | 'ME': 'Montenegro', | |
4158 | 'MS': 'Montserrat', | |
4159 | 'MA': 'Morocco', | |
4160 | 'MZ': 'Mozambique', | |
4161 | 'MM': 'Myanmar', | |
4162 | 'NA': 'Namibia', | |
4163 | 'NR': 'Nauru', | |
4164 | 'NP': 'Nepal', | |
4165 | 'NL': 'Netherlands', | |
4166 | 'NC': 'New Caledonia', | |
4167 | 'NZ': 'New Zealand', | |
4168 | 'NI': 'Nicaragua', | |
4169 | 'NE': 'Niger', | |
4170 | 'NG': 'Nigeria', | |
4171 | 'NU': 'Niue', | |
4172 | 'NF': 'Norfolk Island', | |
4173 | 'MP': 'Northern Mariana Islands', | |
4174 | 'NO': 'Norway', | |
4175 | 'OM': 'Oman', | |
4176 | 'PK': 'Pakistan', | |
4177 | 'PW': 'Palau', | |
4178 | 'PS': 'Palestine, State of', | |
4179 | 'PA': 'Panama', | |
4180 | 'PG': 'Papua New Guinea', | |
4181 | 'PY': 'Paraguay', | |
4182 | 'PE': 'Peru', | |
4183 | 'PH': 'Philippines', | |
4184 | 'PN': 'Pitcairn', | |
4185 | 'PL': 'Poland', | |
4186 | 'PT': 'Portugal', | |
4187 | 'PR': 'Puerto Rico', | |
4188 | 'QA': 'Qatar', | |
4189 | 'RE': 'Réunion', | |
4190 | 'RO': 'Romania', | |
4191 | 'RU': 'Russian Federation', | |
4192 | 'RW': 'Rwanda', | |
4193 | 'BL': 'Saint Barthélemy', | |
4194 | 'SH': 'Saint Helena, Ascension and Tristan da Cunha', | |
4195 | 'KN': 'Saint Kitts and Nevis', | |
4196 | 'LC': 'Saint Lucia', | |
4197 | 'MF': 'Saint Martin (French part)', | |
4198 | 'PM': 'Saint Pierre and Miquelon', | |
4199 | 'VC': 'Saint Vincent and the Grenadines', | |
4200 | 'WS': 'Samoa', | |
4201 | 'SM': 'San Marino', | |
4202 | 'ST': 'Sao Tome and Principe', | |
4203 | 'SA': 'Saudi Arabia', | |
4204 | 'SN': 'Senegal', | |
4205 | 'RS': 'Serbia', | |
4206 | 'SC': 'Seychelles', | |
4207 | 'SL': 'Sierra Leone', | |
4208 | 'SG': 'Singapore', | |
4209 | 'SX': 'Sint Maarten (Dutch part)', | |
4210 | 'SK': 'Slovakia', | |
4211 | 'SI': 'Slovenia', | |
4212 | 'SB': 'Solomon Islands', | |
4213 | 'SO': 'Somalia', | |
4214 | 'ZA': 'South Africa', | |
4215 | 'GS': 'South Georgia and the South Sandwich Islands', | |
4216 | 'SS': 'South Sudan', | |
4217 | 'ES': 'Spain', | |
4218 | 'LK': 'Sri Lanka', | |
4219 | 'SD': 'Sudan', | |
4220 | 'SR': 'Suriname', | |
4221 | 'SJ': 'Svalbard and Jan Mayen', | |
4222 | 'SZ': 'Swaziland', | |
4223 | 'SE': 'Sweden', | |
4224 | 'CH': 'Switzerland', | |
4225 | 'SY': 'Syrian Arab Republic', | |
4226 | 'TW': 'Taiwan, Province of China', | |
4227 | 'TJ': 'Tajikistan', | |
4228 | 'TZ': 'Tanzania, United Republic of', | |
4229 | 'TH': 'Thailand', | |
4230 | 'TL': 'Timor-Leste', | |
4231 | 'TG': 'Togo', | |
4232 | 'TK': 'Tokelau', | |
4233 | 'TO': 'Tonga', | |
4234 | 'TT': 'Trinidad and Tobago', | |
4235 | 'TN': 'Tunisia', | |
4236 | 'TR': 'Turkey', | |
4237 | 'TM': 'Turkmenistan', | |
4238 | 'TC': 'Turks and Caicos Islands', | |
4239 | 'TV': 'Tuvalu', | |
4240 | 'UG': 'Uganda', | |
4241 | 'UA': 'Ukraine', | |
4242 | 'AE': 'United Arab Emirates', | |
4243 | 'GB': 'United Kingdom', | |
4244 | 'US': 'United States', | |
4245 | 'UM': 'United States Minor Outlying Islands', | |
4246 | 'UY': 'Uruguay', | |
4247 | 'UZ': 'Uzbekistan', | |
4248 | 'VU': 'Vanuatu', | |
4249 | 'VE': 'Venezuela, Bolivarian Republic of', | |
4250 | 'VN': 'Viet Nam', | |
4251 | 'VG': 'Virgin Islands, British', | |
4252 | 'VI': 'Virgin Islands, U.S.', | |
4253 | 'WF': 'Wallis and Futuna', | |
4254 | 'EH': 'Western Sahara', | |
4255 | 'YE': 'Yemen', | |
4256 | 'ZM': 'Zambia', | |
4257 | 'ZW': 'Zimbabwe', | |
4258 | } | |
4259 | ||
4260 | @classmethod | |
4261 | def short2full(cls, code): | |
4262 | """Convert an ISO 3166-2 country code to the corresponding full name""" | |
4263 | return cls._country_map.get(code.upper()) | |
4264 | ||
4265 | ||
4266 | class GeoUtils(object): | |
4267 | # Major IPv4 address blocks per country | |
4268 | _country_ip_map = { | |
4269 | 'AD': '46.172.224.0/19', | |
4270 | 'AE': '94.200.0.0/13', | |
4271 | 'AF': '149.54.0.0/17', | |
4272 | 'AG': '209.59.64.0/18', | |
4273 | 'AI': '204.14.248.0/21', | |
4274 | 'AL': '46.99.0.0/16', | |
4275 | 'AM': '46.70.0.0/15', | |
4276 | 'AO': '105.168.0.0/13', | |
4277 | 'AP': '182.50.184.0/21', | |
4278 | 'AQ': '23.154.160.0/24', | |
4279 | 'AR': '181.0.0.0/12', | |
4280 | 'AS': '202.70.112.0/20', | |
4281 | 'AT': '77.116.0.0/14', | |
4282 | 'AU': '1.128.0.0/11', | |
4283 | 'AW': '181.41.0.0/18', | |
4284 | 'AX': '185.217.4.0/22', | |
4285 | 'AZ': '5.197.0.0/16', | |
4286 | 'BA': '31.176.128.0/17', | |
4287 | 'BB': '65.48.128.0/17', | |
4288 | 'BD': '114.130.0.0/16', | |
4289 | 'BE': '57.0.0.0/8', | |
4290 | 'BF': '102.178.0.0/15', | |
4291 | 'BG': '95.42.0.0/15', | |
4292 | 'BH': '37.131.0.0/17', | |
4293 | 'BI': '154.117.192.0/18', | |
4294 | 'BJ': '137.255.0.0/16', | |
4295 | 'BL': '185.212.72.0/23', | |
4296 | 'BM': '196.12.64.0/18', | |
4297 | 'BN': '156.31.0.0/16', | |
4298 | 'BO': '161.56.0.0/16', | |
4299 | 'BQ': '161.0.80.0/20', | |
4300 | 'BR': '191.128.0.0/12', | |
4301 | 'BS': '24.51.64.0/18', | |
4302 | 'BT': '119.2.96.0/19', | |
4303 | 'BW': '168.167.0.0/16', | |
4304 | 'BY': '178.120.0.0/13', | |
4305 | 'BZ': '179.42.192.0/18', | |
4306 | 'CA': '99.224.0.0/11', | |
4307 | 'CD': '41.243.0.0/16', | |
4308 | 'CF': '197.242.176.0/21', | |
4309 | 'CG': '160.113.0.0/16', | |
4310 | 'CH': '85.0.0.0/13', | |
4311 | 'CI': '102.136.0.0/14', | |
4312 | 'CK': '202.65.32.0/19', | |
4313 | 'CL': '152.172.0.0/14', | |
4314 | 'CM': '102.244.0.0/14', | |
4315 | 'CN': '36.128.0.0/10', | |
4316 | 'CO': '181.240.0.0/12', | |
4317 | 'CR': '201.192.0.0/12', | |
4318 | 'CU': '152.206.0.0/15', | |
4319 | 'CV': '165.90.96.0/19', | |
4320 | 'CW': '190.88.128.0/17', | |
4321 | 'CY': '31.153.0.0/16', | |
4322 | 'CZ': '88.100.0.0/14', | |
4323 | 'DE': '53.0.0.0/8', | |
4324 | 'DJ': '197.241.0.0/17', | |
4325 | 'DK': '87.48.0.0/12', | |
4326 | 'DM': '192.243.48.0/20', | |
4327 | 'DO': '152.166.0.0/15', | |
4328 | 'DZ': '41.96.0.0/12', | |
4329 | 'EC': '186.68.0.0/15', | |
4330 | 'EE': '90.190.0.0/15', | |
4331 | 'EG': '156.160.0.0/11', | |
4332 | 'ER': '196.200.96.0/20', | |
4333 | 'ES': '88.0.0.0/11', | |
4334 | 'ET': '196.188.0.0/14', | |
4335 | 'EU': '2.16.0.0/13', | |
4336 | 'FI': '91.152.0.0/13', | |
4337 | 'FJ': '144.120.0.0/16', | |
4338 | 'FK': '80.73.208.0/21', | |
4339 | 'FM': '119.252.112.0/20', | |
4340 | 'FO': '88.85.32.0/19', | |
4341 | 'FR': '90.0.0.0/9', | |
4342 | 'GA': '41.158.0.0/15', | |
4343 | 'GB': '25.0.0.0/8', | |
4344 | 'GD': '74.122.88.0/21', | |
4345 | 'GE': '31.146.0.0/16', | |
4346 | 'GF': '161.22.64.0/18', | |
4347 | 'GG': '62.68.160.0/19', | |
4348 | 'GH': '154.160.0.0/12', | |
4349 | 'GI': '95.164.0.0/16', | |
4350 | 'GL': '88.83.0.0/19', | |
4351 | 'GM': '160.182.0.0/15', | |
4352 | 'GN': '197.149.192.0/18', | |
4353 | 'GP': '104.250.0.0/19', | |
4354 | 'GQ': '105.235.224.0/20', | |
4355 | 'GR': '94.64.0.0/13', | |
4356 | 'GT': '168.234.0.0/16', | |
4357 | 'GU': '168.123.0.0/16', | |
4358 | 'GW': '197.214.80.0/20', | |
4359 | 'GY': '181.41.64.0/18', | |
4360 | 'HK': '113.252.0.0/14', | |
4361 | 'HN': '181.210.0.0/16', | |
4362 | 'HR': '93.136.0.0/13', | |
4363 | 'HT': '148.102.128.0/17', | |
4364 | 'HU': '84.0.0.0/14', | |
4365 | 'ID': '39.192.0.0/10', | |
4366 | 'IE': '87.32.0.0/12', | |
4367 | 'IL': '79.176.0.0/13', | |
4368 | 'IM': '5.62.80.0/20', | |
4369 | 'IN': '117.192.0.0/10', | |
4370 | 'IO': '203.83.48.0/21', | |
4371 | 'IQ': '37.236.0.0/14', | |
4372 | 'IR': '2.176.0.0/12', | |
4373 | 'IS': '82.221.0.0/16', | |
4374 | 'IT': '79.0.0.0/10', | |
4375 | 'JE': '87.244.64.0/18', | |
4376 | 'JM': '72.27.0.0/17', | |
4377 | 'JO': '176.29.0.0/16', | |
4378 | 'JP': '133.0.0.0/8', | |
4379 | 'KE': '105.48.0.0/12', | |
4380 | 'KG': '158.181.128.0/17', | |
4381 | 'KH': '36.37.128.0/17', | |
4382 | 'KI': '103.25.140.0/22', | |
4383 | 'KM': '197.255.224.0/20', | |
4384 | 'KN': '198.167.192.0/19', | |
4385 | 'KP': '175.45.176.0/22', | |
4386 | 'KR': '175.192.0.0/10', | |
4387 | 'KW': '37.36.0.0/14', | |
4388 | 'KY': '64.96.0.0/15', | |
4389 | 'KZ': '2.72.0.0/13', | |
4390 | 'LA': '115.84.64.0/18', | |
4391 | 'LB': '178.135.0.0/16', | |
4392 | 'LC': '24.92.144.0/20', | |
4393 | 'LI': '82.117.0.0/19', | |
4394 | 'LK': '112.134.0.0/15', | |
4395 | 'LR': '102.183.0.0/16', | |
4396 | 'LS': '129.232.0.0/17', | |
4397 | 'LT': '78.56.0.0/13', | |
4398 | 'LU': '188.42.0.0/16', | |
4399 | 'LV': '46.109.0.0/16', | |
4400 | 'LY': '41.252.0.0/14', | |
4401 | 'MA': '105.128.0.0/11', | |
4402 | 'MC': '88.209.64.0/18', | |
4403 | 'MD': '37.246.0.0/16', | |
4404 | 'ME': '178.175.0.0/17', | |
4405 | 'MF': '74.112.232.0/21', | |
4406 | 'MG': '154.126.0.0/17', | |
4407 | 'MH': '117.103.88.0/21', | |
4408 | 'MK': '77.28.0.0/15', | |
4409 | 'ML': '154.118.128.0/18', | |
4410 | 'MM': '37.111.0.0/17', | |
4411 | 'MN': '49.0.128.0/17', | |
4412 | 'MO': '60.246.0.0/16', | |
4413 | 'MP': '202.88.64.0/20', | |
4414 | 'MQ': '109.203.224.0/19', | |
4415 | 'MR': '41.188.64.0/18', | |
4416 | 'MS': '208.90.112.0/22', | |
4417 | 'MT': '46.11.0.0/16', | |
4418 | 'MU': '105.16.0.0/12', | |
4419 | 'MV': '27.114.128.0/18', | |
4420 | 'MW': '102.70.0.0/15', | |
4421 | 'MX': '187.192.0.0/11', | |
4422 | 'MY': '175.136.0.0/13', | |
4423 | 'MZ': '197.218.0.0/15', | |
4424 | 'NA': '41.182.0.0/16', | |
4425 | 'NC': '101.101.0.0/18', | |
4426 | 'NE': '197.214.0.0/18', | |
4427 | 'NF': '203.17.240.0/22', | |
4428 | 'NG': '105.112.0.0/12', | |
4429 | 'NI': '186.76.0.0/15', | |
4430 | 'NL': '145.96.0.0/11', | |
4431 | 'NO': '84.208.0.0/13', | |
4432 | 'NP': '36.252.0.0/15', | |
4433 | 'NR': '203.98.224.0/19', | |
4434 | 'NU': '49.156.48.0/22', | |
4435 | 'NZ': '49.224.0.0/14', | |
4436 | 'OM': '5.36.0.0/15', | |
4437 | 'PA': '186.72.0.0/15', | |
4438 | 'PE': '186.160.0.0/14', | |
4439 | 'PF': '123.50.64.0/18', | |
4440 | 'PG': '124.240.192.0/19', | |
4441 | 'PH': '49.144.0.0/13', | |
4442 | 'PK': '39.32.0.0/11', | |
4443 | 'PL': '83.0.0.0/11', | |
4444 | 'PM': '70.36.0.0/20', | |
4445 | 'PR': '66.50.0.0/16', | |
4446 | 'PS': '188.161.0.0/16', | |
4447 | 'PT': '85.240.0.0/13', | |
4448 | 'PW': '202.124.224.0/20', | |
4449 | 'PY': '181.120.0.0/14', | |
4450 | 'QA': '37.210.0.0/15', | |
4451 | 'RE': '102.35.0.0/16', | |
4452 | 'RO': '79.112.0.0/13', | |
4453 | 'RS': '93.86.0.0/15', | |
4454 | 'RU': '5.136.0.0/13', | |
4455 | 'RW': '41.186.0.0/16', | |
4456 | 'SA': '188.48.0.0/13', | |
4457 | 'SB': '202.1.160.0/19', | |
4458 | 'SC': '154.192.0.0/11', | |
4459 | 'SD': '102.120.0.0/13', | |
4460 | 'SE': '78.64.0.0/12', | |
4461 | 'SG': '8.128.0.0/10', | |
4462 | 'SI': '188.196.0.0/14', | |
4463 | 'SK': '78.98.0.0/15', | |
4464 | 'SL': '102.143.0.0/17', | |
4465 | 'SM': '89.186.32.0/19', | |
4466 | 'SN': '41.82.0.0/15', | |
4467 | 'SO': '154.115.192.0/18', | |
4468 | 'SR': '186.179.128.0/17', | |
4469 | 'SS': '105.235.208.0/21', | |
4470 | 'ST': '197.159.160.0/19', | |
4471 | 'SV': '168.243.0.0/16', | |
4472 | 'SX': '190.102.0.0/20', | |
4473 | 'SY': '5.0.0.0/16', | |
4474 | 'SZ': '41.84.224.0/19', | |
4475 | 'TC': '65.255.48.0/20', | |
4476 | 'TD': '154.68.128.0/19', | |
4477 | 'TG': '196.168.0.0/14', | |
4478 | 'TH': '171.96.0.0/13', | |
4479 | 'TJ': '85.9.128.0/18', | |
4480 | 'TK': '27.96.24.0/21', | |
4481 | 'TL': '180.189.160.0/20', | |
4482 | 'TM': '95.85.96.0/19', | |
4483 | 'TN': '197.0.0.0/11', | |
4484 | 'TO': '175.176.144.0/21', | |
4485 | 'TR': '78.160.0.0/11', | |
4486 | 'TT': '186.44.0.0/15', | |
4487 | 'TV': '202.2.96.0/19', | |
4488 | 'TW': '120.96.0.0/11', | |
4489 | 'TZ': '156.156.0.0/14', | |
4490 | 'UA': '37.52.0.0/14', | |
4491 | 'UG': '102.80.0.0/13', | |
4492 | 'US': '6.0.0.0/8', | |
4493 | 'UY': '167.56.0.0/13', | |
4494 | 'UZ': '84.54.64.0/18', | |
4495 | 'VA': '212.77.0.0/19', | |
4496 | 'VC': '207.191.240.0/21', | |
4497 | 'VE': '186.88.0.0/13', | |
4498 | 'VG': '66.81.192.0/20', | |
4499 | 'VI': '146.226.0.0/16', | |
4500 | 'VN': '14.160.0.0/11', | |
4501 | 'VU': '202.80.32.0/20', | |
4502 | 'WF': '117.20.32.0/21', | |
4503 | 'WS': '202.4.32.0/19', | |
4504 | 'YE': '134.35.0.0/16', | |
4505 | 'YT': '41.242.116.0/22', | |
4506 | 'ZA': '41.0.0.0/11', | |
4507 | 'ZM': '102.144.0.0/13', | |
4508 | 'ZW': '102.177.192.0/18', | |
4509 | } | |
4510 | ||
4511 | @classmethod | |
4512 | def random_ipv4(cls, code_or_block): | |
4513 | if len(code_or_block) == 2: | |
4514 | block = cls._country_ip_map.get(code_or_block.upper()) | |
4515 | if not block: | |
4516 | return None | |
4517 | else: | |
4518 | block = code_or_block | |
4519 | addr, preflen = block.split('/') | |
4520 | addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0] | |
4521 | addr_max = addr_min | (0xffffffff >> int(preflen)) | |
4522 | return compat_str(socket.inet_ntoa( | |
4523 | compat_struct_pack('!L', random.randint(addr_min, addr_max)))) | |
4524 | ||
4525 | ||
4526 | class PerRequestProxyHandler(compat_urllib_request.ProxyHandler): | |
4527 | def __init__(self, proxies=None): | |
4528 | # Set default handlers | |
4529 | for type in ('http', 'https'): | |
4530 | setattr(self, '%s_open' % type, | |
4531 | lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open: | |
4532 | meth(r, proxy, type)) | |
4533 | compat_urllib_request.ProxyHandler.__init__(self, proxies) | |
4534 | ||
4535 | def proxy_open(self, req, proxy, type): | |
4536 | req_proxy = req.headers.get('Ytdl-request-proxy') | |
4537 | if req_proxy is not None: | |
4538 | proxy = req_proxy | |
4539 | del req.headers['Ytdl-request-proxy'] | |
4540 | ||
4541 | if proxy == '__noproxy__': | |
4542 | return None # No Proxy | |
4543 | if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'): | |
4544 | req.add_header('Ytdl-socks-proxy', proxy) | |
4545 | # yt-dlp's http/https handlers do wrapping the socket with socks | |
4546 | return None | |
4547 | return compat_urllib_request.ProxyHandler.proxy_open( | |
4548 | self, req, proxy, type) | |
4549 | ||
4550 | ||
4551 | # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is | |
4552 | # released into Public Domain | |
4553 | # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387 | |
4554 | ||
4555 | def long_to_bytes(n, blocksize=0): | |
4556 | """long_to_bytes(n:long, blocksize:int) : string | |
4557 | Convert a long integer to a byte string. | |
4558 | ||
4559 | If optional blocksize is given and greater than zero, pad the front of the | |
4560 | byte string with binary zeros so that the length is a multiple of | |
4561 | blocksize. | |
4562 | """ | |
4563 | # after much testing, this algorithm was deemed to be the fastest | |
4564 | s = b'' | |
4565 | n = int(n) | |
4566 | while n > 0: | |
4567 | s = compat_struct_pack('>I', n & 0xffffffff) + s | |
4568 | n = n >> 32 | |
4569 | # strip off leading zeros | |
4570 | for i in range(len(s)): | |
4571 | if s[i] != b'\000'[0]: | |
4572 | break | |
4573 | else: | |
4574 | # only happens when n == 0 | |
4575 | s = b'\000' | |
4576 | i = 0 | |
4577 | s = s[i:] | |
4578 | # add back some pad bytes. this could be done more efficiently w.r.t. the | |
4579 | # de-padding being done above, but sigh... | |
4580 | if blocksize > 0 and len(s) % blocksize: | |
4581 | s = (blocksize - len(s) % blocksize) * b'\000' + s | |
4582 | return s | |
4583 | ||
4584 | ||
4585 | def bytes_to_long(s): | |
4586 | """bytes_to_long(string) : long | |
4587 | Convert a byte string to a long integer. | |
4588 | ||
4589 | This is (essentially) the inverse of long_to_bytes(). | |
4590 | """ | |
4591 | acc = 0 | |
4592 | length = len(s) | |
4593 | if length % 4: | |
4594 | extra = (4 - length % 4) | |
4595 | s = b'\000' * extra + s | |
4596 | length = length + extra | |
4597 | for i in range(0, length, 4): | |
4598 | acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0] | |
4599 | return acc | |
4600 | ||
4601 | ||
4602 | def ohdave_rsa_encrypt(data, exponent, modulus): | |
4603 | ''' | |
4604 | Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/ | |
4605 | ||
4606 | Input: | |
4607 | data: data to encrypt, bytes-like object | |
4608 | exponent, modulus: parameter e and N of RSA algorithm, both integer | |
4609 | Output: hex string of encrypted data | |
4610 | ||
4611 | Limitation: supports one block encryption only | |
4612 | ''' | |
4613 | ||
4614 | payload = int(binascii.hexlify(data[::-1]), 16) | |
4615 | encrypted = pow(payload, exponent, modulus) | |
4616 | return '%x' % encrypted | |
4617 | ||
4618 | ||
4619 | def pkcs1pad(data, length): | |
4620 | """ | |
4621 | Padding input data with PKCS#1 scheme | |
4622 | ||
4623 | @param {int[]} data input data | |
4624 | @param {int} length target length | |
4625 | @returns {int[]} padded data | |
4626 | """ | |
4627 | if len(data) > length - 11: | |
4628 | raise ValueError('Input data too long for PKCS#1 padding') | |
4629 | ||
4630 | pseudo_random = [random.randint(0, 254) for _ in range(length - len(data) - 3)] | |
4631 | return [0, 2] + pseudo_random + [0] + data | |
4632 | ||
4633 | ||
4634 | def encode_base_n(num, n, table=None): | |
4635 | FULL_TABLE = '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ' | |
4636 | if not table: | |
4637 | table = FULL_TABLE[:n] | |
4638 | ||
4639 | if n > len(table): | |
4640 | raise ValueError('base %d exceeds table length %d' % (n, len(table))) | |
4641 | ||
4642 | if num == 0: | |
4643 | return table[0] | |
4644 | ||
4645 | ret = '' | |
4646 | while num: | |
4647 | ret = table[num % n] + ret | |
4648 | num = num // n | |
4649 | return ret | |
4650 | ||
4651 | ||
4652 | def decode_packed_codes(code): | |
4653 | mobj = re.search(PACKED_CODES_RE, code) | |
4654 | obfuscated_code, base, count, symbols = mobj.groups() | |
4655 | base = int(base) | |
4656 | count = int(count) | |
4657 | symbols = symbols.split('|') | |
4658 | symbol_table = {} | |
4659 | ||
4660 | while count: | |
4661 | count -= 1 | |
4662 | base_n_count = encode_base_n(count, base) | |
4663 | symbol_table[base_n_count] = symbols[count] or base_n_count | |
4664 | ||
4665 | return re.sub( | |
4666 | r'\b(\w+)\b', lambda mobj: symbol_table[mobj.group(0)], | |
4667 | obfuscated_code) | |
4668 | ||
4669 | ||
4670 | def caesar(s, alphabet, shift): | |
4671 | if shift == 0: | |
4672 | return s | |
4673 | l = len(alphabet) | |
4674 | return ''.join( | |
4675 | alphabet[(alphabet.index(c) + shift) % l] if c in alphabet else c | |
4676 | for c in s) | |
4677 | ||
4678 | ||
4679 | def rot47(s): | |
4680 | return caesar(s, r'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47) | |
4681 | ||
4682 | ||
4683 | def parse_m3u8_attributes(attrib): | |
4684 | info = {} | |
4685 | for (key, val) in re.findall(r'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib): | |
4686 | if val.startswith('"'): | |
4687 | val = val[1:-1] | |
4688 | info[key] = val | |
4689 | return info | |
4690 | ||
4691 | ||
4692 | def urshift(val, n): | |
4693 | return val >> n if val >= 0 else (val + 0x100000000) >> n | |
4694 | ||
4695 | ||
4696 | # Based on png2str() written by @gdkchan and improved by @yokrysty | |
4697 | # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706 | |
4698 | def decode_png(png_data): | |
4699 | # Reference: https://www.w3.org/TR/PNG/ | |
4700 | header = png_data[8:] | |
4701 | ||
4702 | if png_data[:8] != b'\x89PNG\x0d\x0a\x1a\x0a' or header[4:8] != b'IHDR': | |
4703 | raise IOError('Not a valid PNG file.') | |
4704 | ||
4705 | int_map = {1: '>B', 2: '>H', 4: '>I'} | |
4706 | unpack_integer = lambda x: compat_struct_unpack(int_map[len(x)], x)[0] | |
4707 | ||
4708 | chunks = [] | |
4709 | ||
4710 | while header: | |
4711 | length = unpack_integer(header[:4]) | |
4712 | header = header[4:] | |
4713 | ||
4714 | chunk_type = header[:4] | |
4715 | header = header[4:] | |
4716 | ||
4717 | chunk_data = header[:length] | |
4718 | header = header[length:] | |
4719 | ||
4720 | header = header[4:] # Skip CRC | |
4721 | ||
4722 | chunks.append({ | |
4723 | 'type': chunk_type, | |
4724 | 'length': length, | |
4725 | 'data': chunk_data | |
4726 | }) | |
4727 | ||
4728 | ihdr = chunks[0]['data'] | |
4729 | ||
4730 | width = unpack_integer(ihdr[:4]) | |
4731 | height = unpack_integer(ihdr[4:8]) | |
4732 | ||
4733 | idat = b'' | |
4734 | ||
4735 | for chunk in chunks: | |
4736 | if chunk['type'] == b'IDAT': | |
4737 | idat += chunk['data'] | |
4738 | ||
4739 | if not idat: | |
4740 | raise IOError('Unable to read PNG data.') | |
4741 | ||
4742 | decompressed_data = bytearray(zlib.decompress(idat)) | |
4743 | ||
4744 | stride = width * 3 | |
4745 | pixels = [] | |
4746 | ||
4747 | def _get_pixel(idx): | |
4748 | x = idx % stride | |
4749 | y = idx // stride | |
4750 | return pixels[y][x] | |
4751 | ||
4752 | for y in range(height): | |
4753 | basePos = y * (1 + stride) | |
4754 | filter_type = decompressed_data[basePos] | |
4755 | ||
4756 | current_row = [] | |
4757 | ||
4758 | pixels.append(current_row) | |
4759 | ||
4760 | for x in range(stride): | |
4761 | color = decompressed_data[1 + basePos + x] | |
4762 | basex = y * stride + x | |
4763 | left = 0 | |
4764 | up = 0 | |
4765 | ||
4766 | if x > 2: | |
4767 | left = _get_pixel(basex - 3) | |
4768 | if y > 0: | |
4769 | up = _get_pixel(basex - stride) | |
4770 | ||
4771 | if filter_type == 1: # Sub | |
4772 | color = (color + left) & 0xff | |
4773 | elif filter_type == 2: # Up | |
4774 | color = (color + up) & 0xff | |
4775 | elif filter_type == 3: # Average | |
4776 | color = (color + ((left + up) >> 1)) & 0xff | |
4777 | elif filter_type == 4: # Paeth | |
4778 | a = left | |
4779 | b = up | |
4780 | c = 0 | |
4781 | ||
4782 | if x > 2 and y > 0: | |
4783 | c = _get_pixel(basex - stride - 3) | |
4784 | ||
4785 | p = a + b - c | |
4786 | ||
4787 | pa = abs(p - a) | |
4788 | pb = abs(p - b) | |
4789 | pc = abs(p - c) | |
4790 | ||
4791 | if pa <= pb and pa <= pc: | |
4792 | color = (color + a) & 0xff | |
4793 | elif pb <= pc: | |
4794 | color = (color + b) & 0xff | |
4795 | else: | |
4796 | color = (color + c) & 0xff | |
4797 | ||
4798 | current_row.append(color) | |
4799 | ||
4800 | return width, height, pixels | |
4801 | ||
4802 | ||
4803 | def write_xattr(path, key, value): | |
4804 | # This mess below finds the best xattr tool for the job | |
4805 | try: | |
4806 | # try the pyxattr module... | |
4807 | import xattr | |
4808 | ||
4809 | if hasattr(xattr, 'set'): # pyxattr | |
4810 | # Unicode arguments are not supported in python-pyxattr until | |
4811 | # version 0.5.0 | |
4812 | # See https://github.com/ytdl-org/youtube-dl/issues/5498 | |
4813 | pyxattr_required_version = '0.5.0' | |
4814 | if version_tuple(xattr.__version__) < version_tuple(pyxattr_required_version): | |
4815 | # TODO: fallback to CLI tools | |
4816 | raise XAttrUnavailableError( | |
4817 | 'python-pyxattr is detected but is too old. ' | |
4818 | 'yt-dlp requires %s or above while your version is %s. ' | |
4819 | 'Falling back to other xattr implementations' % ( | |
4820 | pyxattr_required_version, xattr.__version__)) | |
4821 | ||
4822 | setxattr = xattr.set | |
4823 | else: # xattr | |
4824 | setxattr = xattr.setxattr | |
4825 | ||
4826 | try: | |
4827 | setxattr(path, key, value) | |
4828 | except EnvironmentError as e: | |
4829 | raise XAttrMetadataError(e.errno, e.strerror) | |
4830 | ||
4831 | except ImportError: | |
4832 | if compat_os_name == 'nt': | |
4833 | # Write xattrs to NTFS Alternate Data Streams: | |
4834 | # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29 | |
4835 | assert ':' not in key | |
4836 | assert os.path.exists(path) | |
4837 | ||
4838 | ads_fn = path + ':' + key | |
4839 | try: | |
4840 | with open(ads_fn, 'wb') as f: | |
4841 | f.write(value) | |
4842 | except EnvironmentError as e: | |
4843 | raise XAttrMetadataError(e.errno, e.strerror) | |
4844 | else: | |
4845 | user_has_setfattr = check_executable('setfattr', ['--version']) | |
4846 | user_has_xattr = check_executable('xattr', ['-h']) | |
4847 | ||
4848 | if user_has_setfattr or user_has_xattr: | |
4849 | ||
4850 | value = value.decode('utf-8') | |
4851 | if user_has_setfattr: | |
4852 | executable = 'setfattr' | |
4853 | opts = ['-n', key, '-v', value] | |
4854 | elif user_has_xattr: | |
4855 | executable = 'xattr' | |
4856 | opts = ['-w', key, value] | |
4857 | ||
4858 | cmd = ([encodeFilename(executable, True)] | |
4859 | + [encodeArgument(o) for o in opts] | |
4860 | + [encodeFilename(path, True)]) | |
4861 | ||
4862 | try: | |
4863 | p = Popen( | |
4864 | cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, stdin=subprocess.PIPE) | |
4865 | except EnvironmentError as e: | |
4866 | raise XAttrMetadataError(e.errno, e.strerror) | |
4867 | stdout, stderr = p.communicate_or_kill() | |
4868 | stderr = stderr.decode('utf-8', 'replace') | |
4869 | if p.returncode != 0: | |
4870 | raise XAttrMetadataError(p.returncode, stderr) | |
4871 | ||
4872 | else: | |
4873 | # On Unix, and can't find pyxattr, setfattr, or xattr. | |
4874 | if sys.platform.startswith('linux'): | |
4875 | raise XAttrUnavailableError( | |
4876 | "Couldn't find a tool to set the xattrs. " | |
4877 | "Install either the python 'pyxattr' or 'xattr' " | |
4878 | "modules, or the GNU 'attr' package " | |
4879 | "(which contains the 'setfattr' tool).") | |
4880 | else: | |
4881 | raise XAttrUnavailableError( | |
4882 | "Couldn't find a tool to set the xattrs. " | |
4883 | "Install either the python 'xattr' module, " | |
4884 | "or the 'xattr' binary.") | |
4885 | ||
4886 | ||
4887 | def random_birthday(year_field, month_field, day_field): | |
4888 | start_date = datetime.date(1950, 1, 1) | |
4889 | end_date = datetime.date(1995, 12, 31) | |
4890 | offset = random.randint(0, (end_date - start_date).days) | |
4891 | random_date = start_date + datetime.timedelta(offset) | |
4892 | return { | |
4893 | year_field: str(random_date.year), | |
4894 | month_field: str(random_date.month), | |
4895 | day_field: str(random_date.day), | |
4896 | } | |
4897 | ||
4898 | ||
4899 | # Templates for internet shortcut files, which are plain text files. | |
4900 | DOT_URL_LINK_TEMPLATE = ''' | |
4901 | [InternetShortcut] | |
4902 | URL=%(url)s | |
4903 | '''.lstrip() | |
4904 | ||
4905 | DOT_WEBLOC_LINK_TEMPLATE = ''' | |
4906 | <?xml version="1.0" encoding="UTF-8"?> | |
4907 | <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd"> | |
4908 | <plist version="1.0"> | |
4909 | <dict> | |
4910 | \t<key>URL</key> | |
4911 | \t<string>%(url)s</string> | |
4912 | </dict> | |
4913 | </plist> | |
4914 | '''.lstrip() | |
4915 | ||
4916 | DOT_DESKTOP_LINK_TEMPLATE = ''' | |
4917 | [Desktop Entry] | |
4918 | Encoding=UTF-8 | |
4919 | Name=%(filename)s | |
4920 | Type=Link | |
4921 | URL=%(url)s | |
4922 | Icon=text-html | |
4923 | '''.lstrip() | |
4924 | ||
4925 | LINK_TEMPLATES = { | |
4926 | 'url': DOT_URL_LINK_TEMPLATE, | |
4927 | 'desktop': DOT_DESKTOP_LINK_TEMPLATE, | |
4928 | 'webloc': DOT_WEBLOC_LINK_TEMPLATE, | |
4929 | } | |
4930 | ||
4931 | ||
4932 | def iri_to_uri(iri): | |
4933 | """ | |
4934 | Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only). | |
4935 | ||
4936 | The function doesn't add an additional layer of escaping; e.g., it doesn't escape `%3C` as `%253C`. Instead, it percent-escapes characters with an underlying UTF-8 encoding *besides* those already escaped, leaving the URI intact. | |
4937 | """ | |
4938 | ||
4939 | iri_parts = compat_urllib_parse_urlparse(iri) | |
4940 | ||
4941 | if '[' in iri_parts.netloc: | |
4942 | raise ValueError('IPv6 URIs are not, yet, supported.') | |
4943 | # Querying `.netloc`, when there's only one bracket, also raises a ValueError. | |
4944 | ||
4945 | # The `safe` argument values, that the following code uses, contain the characters that should not be percent-encoded. Everything else but letters, digits and '_.-' will be percent-encoded with an underlying UTF-8 encoding. Everything already percent-encoded will be left as is. | |
4946 | ||
4947 | net_location = '' | |
4948 | if iri_parts.username: | |
4949 | net_location += compat_urllib_parse_quote(iri_parts.username, safe=r"!$%&'()*+,~") | |
4950 | if iri_parts.password is not None: | |
4951 | net_location += ':' + compat_urllib_parse_quote(iri_parts.password, safe=r"!$%&'()*+,~") | |
4952 | net_location += '@' | |
4953 | ||
4954 | net_location += iri_parts.hostname.encode('idna').decode('utf-8') # Punycode for Unicode hostnames. | |
4955 | # The 'idna' encoding produces ASCII text. | |
4956 | if iri_parts.port is not None and iri_parts.port != 80: | |
4957 | net_location += ':' + str(iri_parts.port) | |
4958 | ||
4959 | return compat_urllib_parse_urlunparse( | |
4960 | (iri_parts.scheme, | |
4961 | net_location, | |
4962 | ||
4963 | compat_urllib_parse_quote_plus(iri_parts.path, safe=r"!$%&'()*+,/:;=@|~"), | |
4964 | ||
4965 | # Unsure about the `safe` argument, since this is a legacy way of handling parameters. | |
4966 | compat_urllib_parse_quote_plus(iri_parts.params, safe=r"!$%&'()*+,/:;=@|~"), | |
4967 | ||
4968 | # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component. | |
4969 | compat_urllib_parse_quote_plus(iri_parts.query, safe=r"!$%&'()*+,/:;=?@{|}~"), | |
4970 | ||
4971 | compat_urllib_parse_quote_plus(iri_parts.fragment, safe=r"!#$%&'()*+,/:;=?@{|}~"))) | |
4972 | ||
4973 | # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes. | |
4974 | ||
4975 | ||
4976 | def to_high_limit_path(path): | |
4977 | if sys.platform in ['win32', 'cygwin']: | |
4978 | # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited. | |
4979 | return r'\\?\ '.rstrip() + os.path.abspath(path) | |
4980 | ||
4981 | return path | |
4982 | ||
4983 | ||
4984 | def format_field(obj, field=None, template='%s', ignore=(None, ''), default='', func=None): | |
4985 | val = traverse_obj(obj, *variadic(field)) | |
4986 | if val in ignore: | |
4987 | return default | |
4988 | return template % (func(val) if func else val) | |
4989 | ||
4990 | ||
4991 | def clean_podcast_url(url): | |
4992 | return re.sub(r'''(?x) | |
4993 | (?: | |
4994 | (?: | |
4995 | chtbl\.com/track| | |
4996 | media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/ | |
4997 | play\.podtrac\.com | |
4998 | )/[^/]+| | |
4999 | (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure | |
5000 | flex\.acast\.com| | |
5001 | pd(?: | |
5002 | cn\.co| # https://podcorn.com/analytics-prefix/ | |
5003 | st\.fm # https://podsights.com/docs/ | |
5004 | )/e | |
5005 | )/''', '', url) | |
5006 | ||
5007 | ||
5008 | _HEX_TABLE = '0123456789abcdef' | |
5009 | ||
5010 | ||
5011 | def random_uuidv4(): | |
5012 | return re.sub(r'[xy]', lambda x: _HEX_TABLE[random.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx') | |
5013 | ||
5014 | ||
5015 | def make_dir(path, to_screen=None): | |
5016 | try: | |
5017 | dn = os.path.dirname(path) | |
5018 | if dn and not os.path.exists(dn): | |
5019 | os.makedirs(dn) | |
5020 | return True | |
5021 | except (OSError, IOError) as err: | |
5022 | if callable(to_screen) is not None: | |
5023 | to_screen('unable to create directory ' + error_to_compat_str(err)) | |
5024 | return False | |
5025 | ||
5026 | ||
5027 | def get_executable_path(): | |
5028 | from zipimport import zipimporter | |
5029 | if hasattr(sys, 'frozen'): # Running from PyInstaller | |
5030 | path = os.path.dirname(sys.executable) | |
5031 | elif isinstance(globals().get('__loader__'), zipimporter): # Running from ZIP | |
5032 | path = os.path.join(os.path.dirname(__file__), '../..') | |
5033 | else: | |
5034 | path = os.path.join(os.path.dirname(__file__), '..') | |
5035 | return os.path.abspath(path) | |
5036 | ||
5037 | ||
5038 | def load_plugins(name, suffix, namespace): | |
5039 | classes = {} | |
5040 | try: | |
5041 | plugins_spec = importlib.util.spec_from_file_location( | |
5042 | name, os.path.join(get_executable_path(), 'ytdlp_plugins', name, '__init__.py')) | |
5043 | plugins = importlib.util.module_from_spec(plugins_spec) | |
5044 | sys.modules[plugins_spec.name] = plugins | |
5045 | plugins_spec.loader.exec_module(plugins) | |
5046 | for name in dir(plugins): | |
5047 | if name in namespace: | |
5048 | continue | |
5049 | if not name.endswith(suffix): | |
5050 | continue | |
5051 | klass = getattr(plugins, name) | |
5052 | classes[name] = namespace[name] = klass | |
5053 | except FileNotFoundError: | |
5054 | pass | |
5055 | return classes | |
5056 | ||
5057 | ||
5058 | def traverse_obj( | |
5059 | obj, *path_list, default=None, expected_type=None, get_all=True, | |
5060 | casesense=True, is_user_input=False, traverse_string=False): | |
5061 | ''' Traverse nested list/dict/tuple | |
5062 | @param path_list A list of paths which are checked one by one. | |
5063 | Each path is a list of keys where each key is a string, | |
5064 | a function, a tuple of strings/None or "...". | |
5065 | When a fuction is given, it takes the key as argument and | |
5066 | returns whether the key matches or not. When a tuple is given, | |
5067 | all the keys given in the tuple are traversed, and | |
5068 | "..." traverses all the keys in the object | |
5069 | "None" returns the object without traversal | |
5070 | @param default Default value to return | |
5071 | @param expected_type Only accept final value of this type (Can also be any callable) | |
5072 | @param get_all Return all the values obtained from a path or only the first one | |
5073 | @param casesense Whether to consider dictionary keys as case sensitive | |
5074 | @param is_user_input Whether the keys are generated from user input. If True, | |
5075 | strings are converted to int/slice if necessary | |
5076 | @param traverse_string Whether to traverse inside strings. If True, any | |
5077 | non-compatible object will also be converted into a string | |
5078 | # TODO: Write tests | |
5079 | ''' | |
5080 | if not casesense: | |
5081 | _lower = lambda k: (k.lower() if isinstance(k, str) else k) | |
5082 | path_list = (map(_lower, variadic(path)) for path in path_list) | |
5083 | ||
5084 | def _traverse_obj(obj, path, _current_depth=0): | |
5085 | nonlocal depth | |
5086 | path = tuple(variadic(path)) | |
5087 | for i, key in enumerate(path): | |
5088 | if None in (key, obj): | |
5089 | return obj | |
5090 | if isinstance(key, (list, tuple)): | |
5091 | obj = [_traverse_obj(obj, sub_key, _current_depth) for sub_key in key] | |
5092 | key = ... | |
5093 | if key is ...: | |
5094 | obj = (obj.values() if isinstance(obj, dict) | |
5095 | else obj if isinstance(obj, (list, tuple, LazyList)) | |
5096 | else str(obj) if traverse_string else []) | |
5097 | _current_depth += 1 | |
5098 | depth = max(depth, _current_depth) | |
5099 | return [_traverse_obj(inner_obj, path[i + 1:], _current_depth) for inner_obj in obj] | |
5100 | elif callable(key): | |
5101 | if isinstance(obj, (list, tuple, LazyList)): | |
5102 | obj = enumerate(obj) | |
5103 | elif isinstance(obj, dict): | |
5104 | obj = obj.items() | |
5105 | else: | |
5106 | if not traverse_string: | |
5107 | return None | |
5108 | obj = str(obj) | |
5109 | _current_depth += 1 | |
5110 | depth = max(depth, _current_depth) | |
5111 | return [_traverse_obj(v, path[i + 1:], _current_depth) for k, v in obj if key(k)] | |
5112 | elif isinstance(obj, dict) and not (is_user_input and key == ':'): | |
5113 | obj = (obj.get(key) if casesense or (key in obj) | |
5114 | else next((v for k, v in obj.items() if _lower(k) == key), None)) | |
5115 | else: | |
5116 | if is_user_input: | |
5117 | key = (int_or_none(key) if ':' not in key | |
5118 | else slice(*map(int_or_none, key.split(':')))) | |
5119 | if key == slice(None): | |
5120 | return _traverse_obj(obj, (..., *path[i + 1:]), _current_depth) | |
5121 | if not isinstance(key, (int, slice)): | |
5122 | return None | |
5123 | if not isinstance(obj, (list, tuple, LazyList)): | |
5124 | if not traverse_string: | |
5125 | return None | |
5126 | obj = str(obj) | |
5127 | try: | |
5128 | obj = obj[key] | |
5129 | except IndexError: | |
5130 | return None | |
5131 | return obj | |
5132 | ||
5133 | if isinstance(expected_type, type): | |
5134 | type_test = lambda val: val if isinstance(val, expected_type) else None | |
5135 | elif expected_type is not None: | |
5136 | type_test = expected_type | |
5137 | else: | |
5138 | type_test = lambda val: val | |
5139 | ||
5140 | for path in path_list: | |
5141 | depth = 0 | |
5142 | val = _traverse_obj(obj, path) | |
5143 | if val is not None: | |
5144 | if depth: | |
5145 | for _ in range(depth - 1): | |
5146 | val = itertools.chain.from_iterable(v for v in val if v is not None) | |
5147 | val = [v for v in map(type_test, val) if v is not None] | |
5148 | if val: | |
5149 | return val if get_all else val[0] | |
5150 | else: | |
5151 | val = type_test(val) | |
5152 | if val is not None: | |
5153 | return val | |
5154 | return default | |
5155 | ||
5156 | ||
5157 | def traverse_dict(dictn, keys, casesense=True): | |
5158 | write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated ' | |
5159 | 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead') | |
5160 | return traverse_obj(dictn, keys, casesense=casesense, is_user_input=True, traverse_string=True) | |
5161 | ||
5162 | ||
5163 | def variadic(x, allowed_types=(str, bytes, dict)): | |
5164 | return x if isinstance(x, collections.abc.Iterable) and not isinstance(x, allowed_types) else (x,) | |
5165 | ||
5166 | ||
5167 | # create a JSON Web Signature (jws) with HS256 algorithm | |
5168 | # the resulting format is in JWS Compact Serialization | |
5169 | # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html | |
5170 | # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html | |
5171 | def jwt_encode_hs256(payload_data, key, headers={}): | |
5172 | header_data = { | |
5173 | 'alg': 'HS256', | |
5174 | 'typ': 'JWT', | |
5175 | } | |
5176 | if headers: | |
5177 | header_data.update(headers) | |
5178 | header_b64 = base64.b64encode(json.dumps(header_data).encode('utf-8')) | |
5179 | payload_b64 = base64.b64encode(json.dumps(payload_data).encode('utf-8')) | |
5180 | h = hmac.new(key.encode('utf-8'), header_b64 + b'.' + payload_b64, hashlib.sha256) | |
5181 | signature_b64 = base64.b64encode(h.digest()) | |
5182 | token = header_b64 + b'.' + payload_b64 + b'.' + signature_b64 | |
5183 | return token | |
5184 | ||
5185 | ||
5186 | # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256 | |
5187 | def jwt_decode_hs256(jwt): | |
5188 | header_b64, payload_b64, signature_b64 = jwt.split('.') | |
5189 | payload_data = json.loads(base64.urlsafe_b64decode(payload_b64)) | |
5190 | return payload_data | |
5191 | ||
5192 | ||
5193 | def supports_terminal_sequences(stream): | |
5194 | if compat_os_name == 'nt': | |
5195 | from .compat import WINDOWS_VT_MODE # Must be imported locally | |
5196 | if not WINDOWS_VT_MODE or get_windows_version() < (10, 0, 10586): | |
5197 | return False | |
5198 | elif not os.getenv('TERM'): | |
5199 | return False | |
5200 | try: | |
5201 | return stream.isatty() | |
5202 | except BaseException: | |
5203 | return False | |
5204 | ||
5205 | ||
5206 | _terminal_sequences_re = re.compile('\033\\[[^m]+m') | |
5207 | ||
5208 | ||
5209 | def remove_terminal_sequences(string): | |
5210 | return _terminal_sequences_re.sub('', string) | |
5211 | ||
5212 | ||
5213 | def number_of_digits(number): | |
5214 | return len('%d' % number) | |
5215 | ||
5216 | ||
5217 | def join_nonempty(*values, delim='-', from_dict=None): | |
5218 | if from_dict is not None: | |
5219 | values = map(from_dict.get, values) | |
5220 | return delim.join(map(str, filter(None, values))) | |
5221 | ||
5222 | ||
5223 | class Config: | |
5224 | own_args = None | |
5225 | filename = None | |
5226 | __initialized = False | |
5227 | ||
5228 | def __init__(self, parser, label=None): | |
5229 | self._parser, self.label = parser, label | |
5230 | self._loaded_paths, self.configs = set(), [] | |
5231 | ||
5232 | def init(self, args=None, filename=None): | |
5233 | assert not self.__initialized | |
5234 | directory = '' | |
5235 | if filename: | |
5236 | location = os.path.realpath(filename) | |
5237 | directory = os.path.dirname(location) | |
5238 | if location in self._loaded_paths: | |
5239 | return False | |
5240 | self._loaded_paths.add(location) | |
5241 | ||
5242 | self.__initialized = True | |
5243 | self.own_args, self.filename = args, filename | |
5244 | for location in self._parser.parse_args(args)[0].config_locations or []: | |
5245 | location = os.path.join(directory, expand_path(location)) | |
5246 | if os.path.isdir(location): | |
5247 | location = os.path.join(location, 'yt-dlp.conf') | |
5248 | if not os.path.exists(location): | |
5249 | self._parser.error(f'config location {location} does not exist') | |
5250 | self.append_config(self.read_file(location), location) | |
5251 | return True | |
5252 | ||
5253 | def __str__(self): | |
5254 | label = join_nonempty( | |
5255 | self.label, 'config', f'"{self.filename}"' if self.filename else '', | |
5256 | delim=' ') | |
5257 | return join_nonempty( | |
5258 | self.own_args is not None and f'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}', | |
5259 | *(f'\n{c}'.replace('\n', '\n| ')[1:] for c in self.configs), | |
5260 | delim='\n') | |
5261 | ||
5262 | @staticmethod | |
5263 | def read_file(filename, default=[]): | |
5264 | try: | |
5265 | optionf = open(filename) | |
5266 | except IOError: | |
5267 | return default # silently skip if file is not present | |
5268 | try: | |
5269 | # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56 | |
5270 | contents = optionf.read() | |
5271 | if sys.version_info < (3,): | |
5272 | contents = contents.decode(preferredencoding()) | |
5273 | res = compat_shlex_split(contents, comments=True) | |
5274 | finally: | |
5275 | optionf.close() | |
5276 | return res | |
5277 | ||
5278 | @staticmethod | |
5279 | def hide_login_info(opts): | |
5280 | PRIVATE_OPTS = set(['-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username']) | |
5281 | eqre = re.compile('^(?P<key>' + ('|'.join(re.escape(po) for po in PRIVATE_OPTS)) + ')=.+$') | |
5282 | ||
5283 | def _scrub_eq(o): | |
5284 | m = eqre.match(o) | |
5285 | if m: | |
5286 | return m.group('key') + '=PRIVATE' | |
5287 | else: | |
5288 | return o | |
5289 | ||
5290 | opts = list(map(_scrub_eq, opts)) | |
5291 | for idx, opt in enumerate(opts): | |
5292 | if opt in PRIVATE_OPTS and idx + 1 < len(opts): | |
5293 | opts[idx + 1] = 'PRIVATE' | |
5294 | return opts | |
5295 | ||
5296 | def append_config(self, *args, label=None): | |
5297 | config = type(self)(self._parser, label) | |
5298 | config._loaded_paths = self._loaded_paths | |
5299 | if config.init(*args): | |
5300 | self.configs.append(config) | |
5301 | ||
5302 | @property | |
5303 | def all_args(self): | |
5304 | for config in reversed(self.configs): | |
5305 | yield from config.all_args | |
5306 | yield from self.own_args or [] | |
5307 | ||
5308 | def parse_args(self): | |
5309 | return self._parser.parse_args(list(self.all_args)) |