]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/utils.py
2fe9cf585db817e1d86831c71ef28502b4a16ee5
[youtubedl] / youtube_dl / utils.py
1 #!/usr/bin/env python
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 contextlib
11 import ctypes
12 import datetime
13 import email.utils
14 import email.header
15 import errno
16 import functools
17 import gzip
18 import io
19 import itertools
20 import json
21 import locale
22 import math
23 import operator
24 import os
25 import platform
26 import random
27 import re
28 import socket
29 import ssl
30 import subprocess
31 import sys
32 import tempfile
33 import traceback
34 import xml.etree.ElementTree
35 import zlib
36
37 from .compat import (
38 compat_HTMLParseError,
39 compat_HTMLParser,
40 compat_basestring,
41 compat_chr,
42 compat_ctypes_WINFUNCTYPE,
43 compat_etree_fromstring,
44 compat_expanduser,
45 compat_html_entities,
46 compat_html_entities_html5,
47 compat_http_client,
48 compat_kwargs,
49 compat_os_name,
50 compat_parse_qs,
51 compat_shlex_quote,
52 compat_socket_create_connection,
53 compat_str,
54 compat_struct_pack,
55 compat_struct_unpack,
56 compat_urllib_error,
57 compat_urllib_parse,
58 compat_urllib_parse_urlencode,
59 compat_urllib_parse_urlparse,
60 compat_urllib_parse_unquote_plus,
61 compat_urllib_request,
62 compat_urlparse,
63 compat_xpath,
64 )
65
66 from .socks import (
67 ProxyType,
68 sockssocket,
69 )
70
71
72 def register_socks_protocols():
73 # "Register" SOCKS protocols
74 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
75 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
76 for scheme in ('socks', 'socks4', 'socks4a', 'socks5'):
77 if scheme not in compat_urlparse.uses_netloc:
78 compat_urlparse.uses_netloc.append(scheme)
79
80
81 # This is not clearly defined otherwise
82 compiled_regex_type = type(re.compile(''))
83
84 std_headers = {
85 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/47.0 (Chrome)',
86 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
87 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
88 'Accept-Encoding': 'gzip, deflate',
89 'Accept-Language': 'en-us,en;q=0.5',
90 }
91
92
93 USER_AGENTS = {
94 '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',
95 }
96
97
98 NO_DEFAULT = object()
99
100 ENGLISH_MONTH_NAMES = [
101 'January', 'February', 'March', 'April', 'May', 'June',
102 'July', 'August', 'September', 'October', 'November', 'December']
103
104 MONTH_NAMES = {
105 'en': ENGLISH_MONTH_NAMES,
106 'fr': [
107 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
108 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
109 }
110
111 KNOWN_EXTENSIONS = (
112 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
113 'flv', 'f4v', 'f4a', 'f4b',
114 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
115 'mkv', 'mka', 'mk3d',
116 'avi', 'divx',
117 'mov',
118 'asf', 'wmv', 'wma',
119 '3gp', '3g2',
120 'mp3',
121 'flac',
122 'ape',
123 'wav',
124 'f4f', 'f4m', 'm3u8', 'smil')
125
126 # needed for sanitizing filenames in restricted mode
127 ACCENT_CHARS = dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
128 itertools.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUYP', ['ss'],
129 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuypy')))
130
131 DATE_FORMATS = (
132 '%d %B %Y',
133 '%d %b %Y',
134 '%B %d %Y',
135 '%B %dst %Y',
136 '%B %dnd %Y',
137 '%B %dth %Y',
138 '%b %d %Y',
139 '%b %dst %Y',
140 '%b %dnd %Y',
141 '%b %dth %Y',
142 '%b %dst %Y %I:%M',
143 '%b %dnd %Y %I:%M',
144 '%b %dth %Y %I:%M',
145 '%Y %m %d',
146 '%Y-%m-%d',
147 '%Y/%m/%d',
148 '%Y/%m/%d %H:%M',
149 '%Y/%m/%d %H:%M:%S',
150 '%Y-%m-%d %H:%M',
151 '%Y-%m-%d %H:%M:%S',
152 '%Y-%m-%d %H:%M:%S.%f',
153 '%d.%m.%Y %H:%M',
154 '%d.%m.%Y %H.%M',
155 '%Y-%m-%dT%H:%M:%SZ',
156 '%Y-%m-%dT%H:%M:%S.%fZ',
157 '%Y-%m-%dT%H:%M:%S.%f0Z',
158 '%Y-%m-%dT%H:%M:%S',
159 '%Y-%m-%dT%H:%M:%S.%f',
160 '%Y-%m-%dT%H:%M',
161 '%b %d %Y at %H:%M',
162 '%b %d %Y at %H:%M:%S',
163 '%B %d %Y at %H:%M',
164 '%B %d %Y at %H:%M:%S',
165 )
166
167 DATE_FORMATS_DAY_FIRST = list(DATE_FORMATS)
168 DATE_FORMATS_DAY_FIRST.extend([
169 '%d-%m-%Y',
170 '%d.%m.%Y',
171 '%d.%m.%y',
172 '%d/%m/%Y',
173 '%d/%m/%y',
174 '%d/%m/%Y %H:%M:%S',
175 ])
176
177 DATE_FORMATS_MONTH_FIRST = list(DATE_FORMATS)
178 DATE_FORMATS_MONTH_FIRST.extend([
179 '%m-%d-%Y',
180 '%m.%d.%Y',
181 '%m/%d/%Y',
182 '%m/%d/%y',
183 '%m/%d/%Y %H:%M:%S',
184 ])
185
186 PACKED_CODES_RE = r"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
187
188
189 def preferredencoding():
190 """Get preferred encoding.
191
192 Returns the best encoding scheme for the system, based on
193 locale.getpreferredencoding() and some further tweaks.
194 """
195 try:
196 pref = locale.getpreferredencoding()
197 'TEST'.encode(pref)
198 except Exception:
199 pref = 'UTF-8'
200
201 return pref
202
203
204 def write_json_file(obj, fn):
205 """ Encode obj as JSON and write it to fn, atomically if possible """
206
207 fn = encodeFilename(fn)
208 if sys.version_info < (3, 0) and sys.platform != 'win32':
209 encoding = get_filesystem_encoding()
210 # os.path.basename returns a bytes object, but NamedTemporaryFile
211 # will fail if the filename contains non ascii characters unless we
212 # use a unicode object
213 path_basename = lambda f: os.path.basename(fn).decode(encoding)
214 # the same for os.path.dirname
215 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
216 else:
217 path_basename = os.path.basename
218 path_dirname = os.path.dirname
219
220 args = {
221 'suffix': '.tmp',
222 'prefix': path_basename(fn) + '.',
223 'dir': path_dirname(fn),
224 'delete': False,
225 }
226
227 # In Python 2.x, json.dump expects a bytestream.
228 # In Python 3.x, it writes to a character stream
229 if sys.version_info < (3, 0):
230 args['mode'] = 'wb'
231 else:
232 args.update({
233 'mode': 'w',
234 'encoding': 'utf-8',
235 })
236
237 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
238
239 try:
240 with tf:
241 json.dump(obj, tf)
242 if sys.platform == 'win32':
243 # Need to remove existing file on Windows, else os.rename raises
244 # WindowsError or FileExistsError.
245 try:
246 os.unlink(fn)
247 except OSError:
248 pass
249 os.rename(tf.name, fn)
250 except Exception:
251 try:
252 os.remove(tf.name)
253 except OSError:
254 pass
255 raise
256
257
258 if sys.version_info >= (2, 7):
259 def find_xpath_attr(node, xpath, key, val=None):
260 """ Find the xpath xpath[@key=val] """
261 assert re.match(r'^[a-zA-Z_-]+$', key)
262 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
263 return node.find(expr)
264 else:
265 def find_xpath_attr(node, xpath, key, val=None):
266 for f in node.findall(compat_xpath(xpath)):
267 if key not in f.attrib:
268 continue
269 if val is None or f.attrib.get(key) == val:
270 return f
271 return None
272
273 # On python2.6 the xml.etree.ElementTree.Element methods don't support
274 # the namespace parameter
275
276
277 def xpath_with_ns(path, ns_map):
278 components = [c.split(':') for c in path.split('/')]
279 replaced = []
280 for c in components:
281 if len(c) == 1:
282 replaced.append(c[0])
283 else:
284 ns, tag = c
285 replaced.append('{%s}%s' % (ns_map[ns], tag))
286 return '/'.join(replaced)
287
288
289 def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
290 def _find_xpath(xpath):
291 return node.find(compat_xpath(xpath))
292
293 if isinstance(xpath, (str, compat_str)):
294 n = _find_xpath(xpath)
295 else:
296 for xp in xpath:
297 n = _find_xpath(xp)
298 if n is not None:
299 break
300
301 if n is None:
302 if default is not NO_DEFAULT:
303 return default
304 elif fatal:
305 name = xpath if name is None else name
306 raise ExtractorError('Could not find XML element %s' % name)
307 else:
308 return None
309 return n
310
311
312 def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
313 n = xpath_element(node, xpath, name, fatal=fatal, default=default)
314 if n is None or n == default:
315 return n
316 if n.text is None:
317 if default is not NO_DEFAULT:
318 return default
319 elif fatal:
320 name = xpath if name is None else name
321 raise ExtractorError('Could not find XML element\'s text %s' % name)
322 else:
323 return None
324 return n.text
325
326
327 def xpath_attr(node, xpath, key, name=None, fatal=False, default=NO_DEFAULT):
328 n = find_xpath_attr(node, xpath, key)
329 if n is None:
330 if default is not NO_DEFAULT:
331 return default
332 elif fatal:
333 name = '%s[@%s]' % (xpath, key) if name is None else name
334 raise ExtractorError('Could not find XML attribute %s' % name)
335 else:
336 return None
337 return n.attrib[key]
338
339
340 def get_element_by_id(id, html):
341 """Return the content of the tag with the specified ID in the passed HTML document"""
342 return get_element_by_attribute('id', id, html)
343
344
345 def get_element_by_class(class_name, html):
346 """Return the content of the first tag with the specified class in the passed HTML document"""
347 retval = get_elements_by_class(class_name, html)
348 return retval[0] if retval else None
349
350
351 def get_element_by_attribute(attribute, value, html, escape_value=True):
352 retval = get_elements_by_attribute(attribute, value, html, escape_value)
353 return retval[0] if retval else None
354
355
356 def get_elements_by_class(class_name, html):
357 """Return the content of all tags with the specified class in the passed HTML document as a list"""
358 return get_elements_by_attribute(
359 'class', r'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
360 html, escape_value=False)
361
362
363 def get_elements_by_attribute(attribute, value, html, escape_value=True):
364 """Return the content of the tag with the specified attribute in the passed HTML document"""
365
366 value = re.escape(value) if escape_value else value
367
368 retlist = []
369 for m in re.finditer(r'''(?xs)
370 <([a-zA-Z0-9:._-]+)
371 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?
372 \s+%s=['"]?%s['"]?
373 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?
374 \s*>
375 (?P<content>.*?)
376 </\1>
377 ''' % (re.escape(attribute), value), html):
378 res = m.group('content')
379
380 if res.startswith('"') or res.startswith("'"):
381 res = res[1:-1]
382
383 retlist.append(unescapeHTML(res))
384
385 return retlist
386
387
388 class HTMLAttributeParser(compat_HTMLParser):
389 """Trivial HTML parser to gather the attributes for a single element"""
390 def __init__(self):
391 self.attrs = {}
392 compat_HTMLParser.__init__(self)
393
394 def handle_starttag(self, tag, attrs):
395 self.attrs = dict(attrs)
396
397
398 def extract_attributes(html_element):
399 """Given a string for an HTML element such as
400 <el
401 a="foo" B="bar" c="&98;az" d=boz
402 empty= noval entity="&amp;"
403 sq='"' dq="'"
404 >
405 Decode and return a dictionary of attributes.
406 {
407 'a': 'foo', 'b': 'bar', c: 'baz', d: 'boz',
408 'empty': '', 'noval': None, 'entity': '&',
409 'sq': '"', 'dq': '\''
410 }.
411 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
412 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
413 """
414 parser = HTMLAttributeParser()
415 try:
416 parser.feed(html_element)
417 parser.close()
418 # Older Python may throw HTMLParseError in case of malformed HTML
419 except compat_HTMLParseError:
420 pass
421 return parser.attrs
422
423
424 def clean_html(html):
425 """Clean an HTML snippet into a readable string"""
426
427 if html is None: # Convenience for sanitizing descriptions etc.
428 return html
429
430 # Newline vs <br />
431 html = html.replace('\n', ' ')
432 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
433 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
434 # Strip html tags
435 html = re.sub('<.*?>', '', html)
436 # Replace html entities
437 html = unescapeHTML(html)
438 return html.strip()
439
440
441 def sanitize_open(filename, open_mode):
442 """Try to open the given filename, and slightly tweak it if this fails.
443
444 Attempts to open the given filename. If this fails, it tries to change
445 the filename slightly, step by step, until it's either able to open it
446 or it fails and raises a final exception, like the standard open()
447 function.
448
449 It returns the tuple (stream, definitive_file_name).
450 """
451 try:
452 if filename == '-':
453 if sys.platform == 'win32':
454 import msvcrt
455 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
456 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
457 stream = open(encodeFilename(filename), open_mode)
458 return (stream, filename)
459 except (IOError, OSError) as err:
460 if err.errno in (errno.EACCES,):
461 raise
462
463 # In case of error, try to remove win32 forbidden chars
464 alt_filename = sanitize_path(filename)
465 if alt_filename == filename:
466 raise
467 else:
468 # An exception here should be caught in the caller
469 stream = open(encodeFilename(alt_filename), open_mode)
470 return (stream, alt_filename)
471
472
473 def timeconvert(timestr):
474 """Convert RFC 2822 defined time string into system timestamp"""
475 timestamp = None
476 timetuple = email.utils.parsedate_tz(timestr)
477 if timetuple is not None:
478 timestamp = email.utils.mktime_tz(timetuple)
479 return timestamp
480
481
482 def sanitize_filename(s, restricted=False, is_id=False):
483 """Sanitizes a string so it could be used as part of a filename.
484 If restricted is set, use a stricter subset of allowed characters.
485 Set is_id if this is not an arbitrary string, but an ID that should be kept
486 if possible.
487 """
488 def replace_insane(char):
489 if restricted and char in ACCENT_CHARS:
490 return ACCENT_CHARS[char]
491 if char == '?' or ord(char) < 32 or ord(char) == 127:
492 return ''
493 elif char == '"':
494 return '' if restricted else '\''
495 elif char == ':':
496 return '_-' if restricted else ' -'
497 elif char in '\\/|*<>':
498 return '_'
499 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
500 return '_'
501 if restricted and ord(char) > 127:
502 return '_'
503 return char
504
505 # Handle timestamps
506 s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
507 result = ''.join(map(replace_insane, s))
508 if not is_id:
509 while '__' in result:
510 result = result.replace('__', '_')
511 result = result.strip('_')
512 # Common case of "Foreign band name - English song title"
513 if restricted and result.startswith('-_'):
514 result = result[2:]
515 if result.startswith('-'):
516 result = '_' + result[len('-'):]
517 result = result.lstrip('.')
518 if not result:
519 result = '_'
520 return result
521
522
523 def sanitize_path(s):
524 """Sanitizes and normalizes path on Windows"""
525 if sys.platform != 'win32':
526 return s
527 drive_or_unc, _ = os.path.splitdrive(s)
528 if sys.version_info < (2, 7) and not drive_or_unc:
529 drive_or_unc, _ = os.path.splitunc(s)
530 norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
531 if drive_or_unc:
532 norm_path.pop(0)
533 sanitized_path = [
534 path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part)
535 for path_part in norm_path]
536 if drive_or_unc:
537 sanitized_path.insert(0, drive_or_unc + os.path.sep)
538 return os.path.join(*sanitized_path)
539
540
541 # Prepend protocol-less URLs with `http:` scheme in order to mitigate the number of
542 # unwanted failures due to missing protocol
543 def sanitize_url(url):
544 return 'http:%s' % url if url.startswith('//') else url
545
546
547 def sanitized_Request(url, *args, **kwargs):
548 return compat_urllib_request.Request(sanitize_url(url), *args, **kwargs)
549
550
551 def expand_path(s):
552 """Expand shell variables and ~"""
553 return os.path.expandvars(compat_expanduser(s))
554
555
556 def orderedSet(iterable):
557 """ Remove all duplicates from the input iterable """
558 res = []
559 for el in iterable:
560 if el not in res:
561 res.append(el)
562 return res
563
564
565 def _htmlentity_transform(entity_with_semicolon):
566 """Transforms an HTML entity to a character."""
567 entity = entity_with_semicolon[:-1]
568
569 # Known non-numeric HTML entity
570 if entity in compat_html_entities.name2codepoint:
571 return compat_chr(compat_html_entities.name2codepoint[entity])
572
573 # TODO: HTML5 allows entities without a semicolon. For example,
574 # '&Eacuteric' should be decoded as 'Éric'.
575 if entity_with_semicolon in compat_html_entities_html5:
576 return compat_html_entities_html5[entity_with_semicolon]
577
578 mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
579 if mobj is not None:
580 numstr = mobj.group(1)
581 if numstr.startswith('x'):
582 base = 16
583 numstr = '0%s' % numstr
584 else:
585 base = 10
586 # See https://github.com/rg3/youtube-dl/issues/7518
587 try:
588 return compat_chr(int(numstr, base))
589 except ValueError:
590 pass
591
592 # Unknown entity in name, return its literal representation
593 return '&%s;' % entity
594
595
596 def unescapeHTML(s):
597 if s is None:
598 return None
599 assert type(s) == compat_str
600
601 return re.sub(
602 r'&([^&;]+;)', lambda m: _htmlentity_transform(m.group(1)), s)
603
604
605 def get_subprocess_encoding():
606 if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
607 # For subprocess calls, encode with locale encoding
608 # Refer to http://stackoverflow.com/a/9951851/35070
609 encoding = preferredencoding()
610 else:
611 encoding = sys.getfilesystemencoding()
612 if encoding is None:
613 encoding = 'utf-8'
614 return encoding
615
616
617 def encodeFilename(s, for_subprocess=False):
618 """
619 @param s The name of the file
620 """
621
622 assert type(s) == compat_str
623
624 # Python 3 has a Unicode API
625 if sys.version_info >= (3, 0):
626 return s
627
628 # Pass '' directly to use Unicode APIs on Windows 2000 and up
629 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
630 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
631 if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
632 return s
633
634 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
635 if sys.platform.startswith('java'):
636 return s
637
638 return s.encode(get_subprocess_encoding(), 'ignore')
639
640
641 def decodeFilename(b, for_subprocess=False):
642
643 if sys.version_info >= (3, 0):
644 return b
645
646 if not isinstance(b, bytes):
647 return b
648
649 return b.decode(get_subprocess_encoding(), 'ignore')
650
651
652 def encodeArgument(s):
653 if not isinstance(s, compat_str):
654 # Legacy code that uses byte strings
655 # Uncomment the following line after fixing all post processors
656 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
657 s = s.decode('ascii')
658 return encodeFilename(s, True)
659
660
661 def decodeArgument(b):
662 return decodeFilename(b, True)
663
664
665 def decodeOption(optval):
666 if optval is None:
667 return optval
668 if isinstance(optval, bytes):
669 optval = optval.decode(preferredencoding())
670
671 assert isinstance(optval, compat_str)
672 return optval
673
674
675 def formatSeconds(secs):
676 if secs > 3600:
677 return '%d:%02d:%02d' % (secs // 3600, (secs % 3600) // 60, secs % 60)
678 elif secs > 60:
679 return '%d:%02d' % (secs // 60, secs % 60)
680 else:
681 return '%d' % secs
682
683
684 def make_HTTPS_handler(params, **kwargs):
685 opts_no_check_certificate = params.get('nocheckcertificate', False)
686 if hasattr(ssl, 'create_default_context'): # Python >= 3.4 or 2.7.9
687 context = ssl.create_default_context(ssl.Purpose.SERVER_AUTH)
688 if opts_no_check_certificate:
689 context.check_hostname = False
690 context.verify_mode = ssl.CERT_NONE
691 try:
692 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
693 except TypeError:
694 # Python 2.7.8
695 # (create_default_context present but HTTPSHandler has no context=)
696 pass
697
698 if sys.version_info < (3, 2):
699 return YoutubeDLHTTPSHandler(params, **kwargs)
700 else: # Python < 3.4
701 context = ssl.SSLContext(ssl.PROTOCOL_TLSv1)
702 context.verify_mode = (ssl.CERT_NONE
703 if opts_no_check_certificate
704 else ssl.CERT_REQUIRED)
705 context.set_default_verify_paths()
706 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
707
708
709 def bug_reports_message():
710 if ytdl_is_updateable():
711 update_cmd = 'type youtube-dl -U to update'
712 else:
713 update_cmd = 'see https://yt-dl.org/update on how to update'
714 msg = '; please report this issue on https://yt-dl.org/bug .'
715 msg += ' Make sure you are using the latest version; %s.' % update_cmd
716 msg += ' Be sure to call youtube-dl with the --verbose flag and include its complete output.'
717 return msg
718
719
720 class YoutubeDLError(Exception):
721 """Base exception for YoutubeDL errors."""
722 pass
723
724
725 class ExtractorError(YoutubeDLError):
726 """Error during info extraction."""
727
728 def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):
729 """ tb, if given, is the original traceback (so that it can be printed out).
730 If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
731 """
732
733 if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError):
734 expected = True
735 if video_id is not None:
736 msg = video_id + ': ' + msg
737 if cause:
738 msg += ' (caused by %r)' % cause
739 if not expected:
740 msg += bug_reports_message()
741 super(ExtractorError, self).__init__(msg)
742
743 self.traceback = tb
744 self.exc_info = sys.exc_info() # preserve original exception
745 self.cause = cause
746 self.video_id = video_id
747
748 def format_traceback(self):
749 if self.traceback is None:
750 return None
751 return ''.join(traceback.format_tb(self.traceback))
752
753
754 class UnsupportedError(ExtractorError):
755 def __init__(self, url):
756 super(UnsupportedError, self).__init__(
757 'Unsupported URL: %s' % url, expected=True)
758 self.url = url
759
760
761 class RegexNotFoundError(ExtractorError):
762 """Error when a regex didn't match"""
763 pass
764
765
766 class GeoRestrictedError(ExtractorError):
767 """Geographic restriction Error exception.
768
769 This exception may be thrown when a video is not available from your
770 geographic location due to geographic restrictions imposed by a website.
771 """
772 def __init__(self, msg, countries=None):
773 super(GeoRestrictedError, self).__init__(msg, expected=True)
774 self.msg = msg
775 self.countries = countries
776
777
778 class DownloadError(YoutubeDLError):
779 """Download Error exception.
780
781 This exception may be thrown by FileDownloader objects if they are not
782 configured to continue on errors. They will contain the appropriate
783 error message.
784 """
785
786 def __init__(self, msg, exc_info=None):
787 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
788 super(DownloadError, self).__init__(msg)
789 self.exc_info = exc_info
790
791
792 class SameFileError(YoutubeDLError):
793 """Same File exception.
794
795 This exception will be thrown by FileDownloader objects if they detect
796 multiple files would have to be downloaded to the same file on disk.
797 """
798 pass
799
800
801 class PostProcessingError(YoutubeDLError):
802 """Post Processing exception.
803
804 This exception may be raised by PostProcessor's .run() method to
805 indicate an error in the postprocessing task.
806 """
807
808 def __init__(self, msg):
809 super(PostProcessingError, self).__init__(msg)
810 self.msg = msg
811
812
813 class MaxDownloadsReached(YoutubeDLError):
814 """ --max-downloads limit has been reached. """
815 pass
816
817
818 class UnavailableVideoError(YoutubeDLError):
819 """Unavailable Format exception.
820
821 This exception will be thrown when a video is requested
822 in a format that is not available for that video.
823 """
824 pass
825
826
827 class ContentTooShortError(YoutubeDLError):
828 """Content Too Short exception.
829
830 This exception may be raised by FileDownloader objects when a file they
831 download is too small for what the server announced first, indicating
832 the connection was probably interrupted.
833 """
834
835 def __init__(self, downloaded, expected):
836 super(ContentTooShortError, self).__init__(
837 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded, expected)
838 )
839 # Both in bytes
840 self.downloaded = downloaded
841 self.expected = expected
842
843
844 class XAttrMetadataError(YoutubeDLError):
845 def __init__(self, code=None, msg='Unknown error'):
846 super(XAttrMetadataError, self).__init__(msg)
847 self.code = code
848 self.msg = msg
849
850 # Parsing code and msg
851 if (self.code in (errno.ENOSPC, errno.EDQUOT) or
852 'No space left' in self.msg or 'Disk quota excedded' in self.msg):
853 self.reason = 'NO_SPACE'
854 elif self.code == errno.E2BIG or 'Argument list too long' in self.msg:
855 self.reason = 'VALUE_TOO_LONG'
856 else:
857 self.reason = 'NOT_SUPPORTED'
858
859
860 class XAttrUnavailableError(YoutubeDLError):
861 pass
862
863
864 def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
865 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
866 # expected HTTP responses to meet HTTP/1.0 or later (see also
867 # https://github.com/rg3/youtube-dl/issues/6727)
868 if sys.version_info < (3, 0):
869 kwargs[b'strict'] = True
870 hc = http_class(*args, **kwargs)
871 source_address = ydl_handler._params.get('source_address')
872 if source_address is not None:
873 sa = (source_address, 0)
874 if hasattr(hc, 'source_address'): # Python 2.7+
875 hc.source_address = sa
876 else: # Python 2.6
877 def _hc_connect(self, *args, **kwargs):
878 sock = compat_socket_create_connection(
879 (self.host, self.port), self.timeout, sa)
880 if is_https:
881 self.sock = ssl.wrap_socket(
882 sock, self.key_file, self.cert_file,
883 ssl_version=ssl.PROTOCOL_TLSv1)
884 else:
885 self.sock = sock
886 hc.connect = functools.partial(_hc_connect, hc)
887
888 return hc
889
890
891 def handle_youtubedl_headers(headers):
892 filtered_headers = headers
893
894 if 'Youtubedl-no-compression' in filtered_headers:
895 filtered_headers = dict((k, v) for k, v in filtered_headers.items() if k.lower() != 'accept-encoding')
896 del filtered_headers['Youtubedl-no-compression']
897
898 return filtered_headers
899
900
901 class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
902 """Handler for HTTP requests and responses.
903
904 This class, when installed with an OpenerDirector, automatically adds
905 the standard headers to every HTTP request and handles gzipped and
906 deflated responses from web servers. If compression is to be avoided in
907 a particular request, the original request in the program code only has
908 to include the HTTP header "Youtubedl-no-compression", which will be
909 removed before making the real request.
910
911 Part of this code was copied from:
912
913 http://techknack.net/python-urllib2-handlers/
914
915 Andrew Rowls, the author of that code, agreed to release it to the
916 public domain.
917 """
918
919 def __init__(self, params, *args, **kwargs):
920 compat_urllib_request.HTTPHandler.__init__(self, *args, **kwargs)
921 self._params = params
922
923 def http_open(self, req):
924 conn_class = compat_http_client.HTTPConnection
925
926 socks_proxy = req.headers.get('Ytdl-socks-proxy')
927 if socks_proxy:
928 conn_class = make_socks_conn_class(conn_class, socks_proxy)
929 del req.headers['Ytdl-socks-proxy']
930
931 return self.do_open(functools.partial(
932 _create_http_connection, self, conn_class, False),
933 req)
934
935 @staticmethod
936 def deflate(data):
937 try:
938 return zlib.decompress(data, -zlib.MAX_WBITS)
939 except zlib.error:
940 return zlib.decompress(data)
941
942 def http_request(self, req):
943 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
944 # always respected by websites, some tend to give out URLs with non percent-encoded
945 # non-ASCII characters (see telemb.py, ard.py [#3412])
946 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
947 # To work around aforementioned issue we will replace request's original URL with
948 # percent-encoded one
949 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
950 # the code of this workaround has been moved here from YoutubeDL.urlopen()
951 url = req.get_full_url()
952 url_escaped = escape_url(url)
953
954 # Substitute URL if any change after escaping
955 if url != url_escaped:
956 req = update_Request(req, url=url_escaped)
957
958 for h, v in std_headers.items():
959 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
960 # The dict keys are capitalized because of this bug by urllib
961 if h.capitalize() not in req.headers:
962 req.add_header(h, v)
963
964 req.headers = handle_youtubedl_headers(req.headers)
965
966 if sys.version_info < (2, 7) and '#' in req.get_full_url():
967 # Python 2.6 is brain-dead when it comes to fragments
968 req._Request__original = req._Request__original.partition('#')[0]
969 req._Request__r_type = req._Request__r_type.partition('#')[0]
970
971 return req
972
973 def http_response(self, req, resp):
974 old_resp = resp
975 # gzip
976 if resp.headers.get('Content-encoding', '') == 'gzip':
977 content = resp.read()
978 gz = gzip.GzipFile(fileobj=io.BytesIO(content), mode='rb')
979 try:
980 uncompressed = io.BytesIO(gz.read())
981 except IOError as original_ioerror:
982 # There may be junk add the end of the file
983 # See http://stackoverflow.com/q/4928560/35070 for details
984 for i in range(1, 1024):
985 try:
986 gz = gzip.GzipFile(fileobj=io.BytesIO(content[:-i]), mode='rb')
987 uncompressed = io.BytesIO(gz.read())
988 except IOError:
989 continue
990 break
991 else:
992 raise original_ioerror
993 resp = compat_urllib_request.addinfourl(uncompressed, old_resp.headers, old_resp.url, old_resp.code)
994 resp.msg = old_resp.msg
995 del resp.headers['Content-encoding']
996 # deflate
997 if resp.headers.get('Content-encoding', '') == 'deflate':
998 gz = io.BytesIO(self.deflate(resp.read()))
999 resp = compat_urllib_request.addinfourl(gz, old_resp.headers, old_resp.url, old_resp.code)
1000 resp.msg = old_resp.msg
1001 del resp.headers['Content-encoding']
1002 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1003 # https://github.com/rg3/youtube-dl/issues/6457).
1004 if 300 <= resp.code < 400:
1005 location = resp.headers.get('Location')
1006 if location:
1007 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1008 if sys.version_info >= (3, 0):
1009 location = location.encode('iso-8859-1').decode('utf-8')
1010 else:
1011 location = location.decode('utf-8')
1012 location_escaped = escape_url(location)
1013 if location != location_escaped:
1014 del resp.headers['Location']
1015 if sys.version_info < (3, 0):
1016 location_escaped = location_escaped.encode('utf-8')
1017 resp.headers['Location'] = location_escaped
1018 return resp
1019
1020 https_request = http_request
1021 https_response = http_response
1022
1023
1024 def make_socks_conn_class(base_class, socks_proxy):
1025 assert issubclass(base_class, (
1026 compat_http_client.HTTPConnection, compat_http_client.HTTPSConnection))
1027
1028 url_components = compat_urlparse.urlparse(socks_proxy)
1029 if url_components.scheme.lower() == 'socks5':
1030 socks_type = ProxyType.SOCKS5
1031 elif url_components.scheme.lower() in ('socks', 'socks4'):
1032 socks_type = ProxyType.SOCKS4
1033 elif url_components.scheme.lower() == 'socks4a':
1034 socks_type = ProxyType.SOCKS4A
1035
1036 def unquote_if_non_empty(s):
1037 if not s:
1038 return s
1039 return compat_urllib_parse_unquote_plus(s)
1040
1041 proxy_args = (
1042 socks_type,
1043 url_components.hostname, url_components.port or 1080,
1044 True, # Remote DNS
1045 unquote_if_non_empty(url_components.username),
1046 unquote_if_non_empty(url_components.password),
1047 )
1048
1049 class SocksConnection(base_class):
1050 def connect(self):
1051 self.sock = sockssocket()
1052 self.sock.setproxy(*proxy_args)
1053 if type(self.timeout) in (int, float):
1054 self.sock.settimeout(self.timeout)
1055 self.sock.connect((self.host, self.port))
1056
1057 if isinstance(self, compat_http_client.HTTPSConnection):
1058 if hasattr(self, '_context'): # Python > 2.6
1059 self.sock = self._context.wrap_socket(
1060 self.sock, server_hostname=self.host)
1061 else:
1062 self.sock = ssl.wrap_socket(self.sock)
1063
1064 return SocksConnection
1065
1066
1067 class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
1068 def __init__(self, params, https_conn_class=None, *args, **kwargs):
1069 compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)
1070 self._https_conn_class = https_conn_class or compat_http_client.HTTPSConnection
1071 self._params = params
1072
1073 def https_open(self, req):
1074 kwargs = {}
1075 conn_class = self._https_conn_class
1076
1077 if hasattr(self, '_context'): # python > 2.6
1078 kwargs['context'] = self._context
1079 if hasattr(self, '_check_hostname'): # python 3.x
1080 kwargs['check_hostname'] = self._check_hostname
1081
1082 socks_proxy = req.headers.get('Ytdl-socks-proxy')
1083 if socks_proxy:
1084 conn_class = make_socks_conn_class(conn_class, socks_proxy)
1085 del req.headers['Ytdl-socks-proxy']
1086
1087 return self.do_open(functools.partial(
1088 _create_http_connection, self, conn_class, True),
1089 req, **kwargs)
1090
1091
1092 class YoutubeDLCookieProcessor(compat_urllib_request.HTTPCookieProcessor):
1093 def __init__(self, cookiejar=None):
1094 compat_urllib_request.HTTPCookieProcessor.__init__(self, cookiejar)
1095
1096 def http_response(self, request, response):
1097 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1098 # characters in Set-Cookie HTTP header of last response (see
1099 # https://github.com/rg3/youtube-dl/issues/6769).
1100 # In order to at least prevent crashing we will percent encode Set-Cookie
1101 # header before HTTPCookieProcessor starts processing it.
1102 # if sys.version_info < (3, 0) and response.headers:
1103 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1104 # set_cookie = response.headers.get(set_cookie_header)
1105 # if set_cookie:
1106 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1107 # if set_cookie != set_cookie_escaped:
1108 # del response.headers[set_cookie_header]
1109 # response.headers[set_cookie_header] = set_cookie_escaped
1110 return compat_urllib_request.HTTPCookieProcessor.http_response(self, request, response)
1111
1112 https_request = compat_urllib_request.HTTPCookieProcessor.http_request
1113 https_response = http_response
1114
1115
1116 def extract_timezone(date_str):
1117 m = re.search(
1118 r'^.{8,}?(?P<tz>Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
1119 date_str)
1120 if not m:
1121 timezone = datetime.timedelta()
1122 else:
1123 date_str = date_str[:-len(m.group('tz'))]
1124 if not m.group('sign'):
1125 timezone = datetime.timedelta()
1126 else:
1127 sign = 1 if m.group('sign') == '+' else -1
1128 timezone = datetime.timedelta(
1129 hours=sign * int(m.group('hours')),
1130 minutes=sign * int(m.group('minutes')))
1131 return timezone, date_str
1132
1133
1134 def parse_iso8601(date_str, delimiter='T', timezone=None):
1135 """ Return a UNIX timestamp from the given date """
1136
1137 if date_str is None:
1138 return None
1139
1140 date_str = re.sub(r'\.[0-9]+', '', date_str)
1141
1142 if timezone is None:
1143 timezone, date_str = extract_timezone(date_str)
1144
1145 try:
1146 date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
1147 dt = datetime.datetime.strptime(date_str, date_format) - timezone
1148 return calendar.timegm(dt.timetuple())
1149 except ValueError:
1150 pass
1151
1152
1153 def date_formats(day_first=True):
1154 return DATE_FORMATS_DAY_FIRST if day_first else DATE_FORMATS_MONTH_FIRST
1155
1156
1157 def unified_strdate(date_str, day_first=True):
1158 """Return a string with the date in the format YYYYMMDD"""
1159
1160 if date_str is None:
1161 return None
1162 upload_date = None
1163 # Replace commas
1164 date_str = date_str.replace(',', ' ')
1165 # Remove AM/PM + timezone
1166 date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
1167 _, date_str = extract_timezone(date_str)
1168
1169 for expression in date_formats(day_first):
1170 try:
1171 upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
1172 except ValueError:
1173 pass
1174 if upload_date is None:
1175 timetuple = email.utils.parsedate_tz(date_str)
1176 if timetuple:
1177 try:
1178 upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
1179 except ValueError:
1180 pass
1181 if upload_date is not None:
1182 return compat_str(upload_date)
1183
1184
1185 def unified_timestamp(date_str, day_first=True):
1186 if date_str is None:
1187 return None
1188
1189 date_str = re.sub(r'[,|]', '', date_str)
1190
1191 pm_delta = 12 if re.search(r'(?i)PM', date_str) else 0
1192 timezone, date_str = extract_timezone(date_str)
1193
1194 # Remove AM/PM + timezone
1195 date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
1196
1197 # Remove unrecognized timezones from ISO 8601 alike timestamps
1198 m = re.search(r'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str)
1199 if m:
1200 date_str = date_str[:-len(m.group('tz'))]
1201
1202 for expression in date_formats(day_first):
1203 try:
1204 dt = datetime.datetime.strptime(date_str, expression) - timezone + datetime.timedelta(hours=pm_delta)
1205 return calendar.timegm(dt.timetuple())
1206 except ValueError:
1207 pass
1208 timetuple = email.utils.parsedate_tz(date_str)
1209 if timetuple:
1210 return calendar.timegm(timetuple) + pm_delta * 3600
1211
1212
1213 def determine_ext(url, default_ext='unknown_video'):
1214 if url is None:
1215 return default_ext
1216 guess = url.partition('?')[0].rpartition('.')[2]
1217 if re.match(r'^[A-Za-z0-9]+$', guess):
1218 return guess
1219 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1220 elif guess.rstrip('/') in KNOWN_EXTENSIONS:
1221 return guess.rstrip('/')
1222 else:
1223 return default_ext
1224
1225
1226 def subtitles_filename(filename, sub_lang, sub_format):
1227 return filename.rsplit('.', 1)[0] + '.' + sub_lang + '.' + sub_format
1228
1229
1230 def date_from_str(date_str):
1231 """
1232 Return a datetime object from a string in the format YYYYMMDD or
1233 (now|today)[+-][0-9](day|week|month|year)(s)?"""
1234 today = datetime.date.today()
1235 if date_str in ('now', 'today'):
1236 return today
1237 if date_str == 'yesterday':
1238 return today - datetime.timedelta(days=1)
1239 match = re.match(r'(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
1240 if match is not None:
1241 sign = match.group('sign')
1242 time = int(match.group('time'))
1243 if sign == '-':
1244 time = -time
1245 unit = match.group('unit')
1246 # A bad approximation?
1247 if unit == 'month':
1248 unit = 'day'
1249 time *= 30
1250 elif unit == 'year':
1251 unit = 'day'
1252 time *= 365
1253 unit += 's'
1254 delta = datetime.timedelta(**{unit: time})
1255 return today + delta
1256 return datetime.datetime.strptime(date_str, '%Y%m%d').date()
1257
1258
1259 def hyphenate_date(date_str):
1260 """
1261 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1262 match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str)
1263 if match is not None:
1264 return '-'.join(match.groups())
1265 else:
1266 return date_str
1267
1268
1269 class DateRange(object):
1270 """Represents a time interval between two dates"""
1271
1272 def __init__(self, start=None, end=None):
1273 """start and end must be strings in the format accepted by date"""
1274 if start is not None:
1275 self.start = date_from_str(start)
1276 else:
1277 self.start = datetime.datetime.min.date()
1278 if end is not None:
1279 self.end = date_from_str(end)
1280 else:
1281 self.end = datetime.datetime.max.date()
1282 if self.start > self.end:
1283 raise ValueError('Date range: "%s" , the start date must be before the end date' % self)
1284
1285 @classmethod
1286 def day(cls, day):
1287 """Returns a range that only contains the given day"""
1288 return cls(day, day)
1289
1290 def __contains__(self, date):
1291 """Check if the date is in the range"""
1292 if not isinstance(date, datetime.date):
1293 date = date_from_str(date)
1294 return self.start <= date <= self.end
1295
1296 def __str__(self):
1297 return '%s - %s' % (self.start.isoformat(), self.end.isoformat())
1298
1299
1300 def platform_name():
1301 """ Returns the platform name as a compat_str """
1302 res = platform.platform()
1303 if isinstance(res, bytes):
1304 res = res.decode(preferredencoding())
1305
1306 assert isinstance(res, compat_str)
1307 return res
1308
1309
1310 def _windows_write_string(s, out):
1311 """ Returns True if the string was written using special methods,
1312 False if it has yet to be written out."""
1313 # Adapted from http://stackoverflow.com/a/3259271/35070
1314
1315 import ctypes
1316 import ctypes.wintypes
1317
1318 WIN_OUTPUT_IDS = {
1319 1: -11,
1320 2: -12,
1321 }
1322
1323 try:
1324 fileno = out.fileno()
1325 except AttributeError:
1326 # If the output stream doesn't have a fileno, it's virtual
1327 return False
1328 except io.UnsupportedOperation:
1329 # Some strange Windows pseudo files?
1330 return False
1331 if fileno not in WIN_OUTPUT_IDS:
1332 return False
1333
1334 GetStdHandle = compat_ctypes_WINFUNCTYPE(
1335 ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
1336 ('GetStdHandle', ctypes.windll.kernel32))
1337 h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
1338
1339 WriteConsoleW = compat_ctypes_WINFUNCTYPE(
1340 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
1341 ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
1342 ctypes.wintypes.LPVOID)(('WriteConsoleW', ctypes.windll.kernel32))
1343 written = ctypes.wintypes.DWORD(0)
1344
1345 GetFileType = compat_ctypes_WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)(('GetFileType', ctypes.windll.kernel32))
1346 FILE_TYPE_CHAR = 0x0002
1347 FILE_TYPE_REMOTE = 0x8000
1348 GetConsoleMode = compat_ctypes_WINFUNCTYPE(
1349 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
1350 ctypes.POINTER(ctypes.wintypes.DWORD))(
1351 ('GetConsoleMode', ctypes.windll.kernel32))
1352 INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
1353
1354 def not_a_console(handle):
1355 if handle == INVALID_HANDLE_VALUE or handle is None:
1356 return True
1357 return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR or
1358 GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0)
1359
1360 if not_a_console(h):
1361 return False
1362
1363 def next_nonbmp_pos(s):
1364 try:
1365 return next(i for i, c in enumerate(s) if ord(c) > 0xffff)
1366 except StopIteration:
1367 return len(s)
1368
1369 while s:
1370 count = min(next_nonbmp_pos(s), 1024)
1371
1372 ret = WriteConsoleW(
1373 h, s, count if count else 2, ctypes.byref(written), None)
1374 if ret == 0:
1375 raise OSError('Failed to write string')
1376 if not count: # We just wrote a non-BMP character
1377 assert written.value == 2
1378 s = s[1:]
1379 else:
1380 assert written.value > 0
1381 s = s[written.value:]
1382 return True
1383
1384
1385 def write_string(s, out=None, encoding=None):
1386 if out is None:
1387 out = sys.stderr
1388 assert type(s) == compat_str
1389
1390 if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'):
1391 if _windows_write_string(s, out):
1392 return
1393
1394 if ('b' in getattr(out, 'mode', '') or
1395 sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr
1396 byt = s.encode(encoding or preferredencoding(), 'ignore')
1397 out.write(byt)
1398 elif hasattr(out, 'buffer'):
1399 enc = encoding or getattr(out, 'encoding', None) or preferredencoding()
1400 byt = s.encode(enc, 'ignore')
1401 out.buffer.write(byt)
1402 else:
1403 out.write(s)
1404 out.flush()
1405
1406
1407 def bytes_to_intlist(bs):
1408 if not bs:
1409 return []
1410 if isinstance(bs[0], int): # Python 3
1411 return list(bs)
1412 else:
1413 return [ord(c) for c in bs]
1414
1415
1416 def intlist_to_bytes(xs):
1417 if not xs:
1418 return b''
1419 return compat_struct_pack('%dB' % len(xs), *xs)
1420
1421
1422 # Cross-platform file locking
1423 if sys.platform == 'win32':
1424 import ctypes.wintypes
1425 import msvcrt
1426
1427 class OVERLAPPED(ctypes.Structure):
1428 _fields_ = [
1429 ('Internal', ctypes.wintypes.LPVOID),
1430 ('InternalHigh', ctypes.wintypes.LPVOID),
1431 ('Offset', ctypes.wintypes.DWORD),
1432 ('OffsetHigh', ctypes.wintypes.DWORD),
1433 ('hEvent', ctypes.wintypes.HANDLE),
1434 ]
1435
1436 kernel32 = ctypes.windll.kernel32
1437 LockFileEx = kernel32.LockFileEx
1438 LockFileEx.argtypes = [
1439 ctypes.wintypes.HANDLE, # hFile
1440 ctypes.wintypes.DWORD, # dwFlags
1441 ctypes.wintypes.DWORD, # dwReserved
1442 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1443 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1444 ctypes.POINTER(OVERLAPPED) # Overlapped
1445 ]
1446 LockFileEx.restype = ctypes.wintypes.BOOL
1447 UnlockFileEx = kernel32.UnlockFileEx
1448 UnlockFileEx.argtypes = [
1449 ctypes.wintypes.HANDLE, # hFile
1450 ctypes.wintypes.DWORD, # dwReserved
1451 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
1452 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
1453 ctypes.POINTER(OVERLAPPED) # Overlapped
1454 ]
1455 UnlockFileEx.restype = ctypes.wintypes.BOOL
1456 whole_low = 0xffffffff
1457 whole_high = 0x7fffffff
1458
1459 def _lock_file(f, exclusive):
1460 overlapped = OVERLAPPED()
1461 overlapped.Offset = 0
1462 overlapped.OffsetHigh = 0
1463 overlapped.hEvent = 0
1464 f._lock_file_overlapped_p = ctypes.pointer(overlapped)
1465 handle = msvcrt.get_osfhandle(f.fileno())
1466 if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0,
1467 whole_low, whole_high, f._lock_file_overlapped_p):
1468 raise OSError('Locking file failed: %r' % ctypes.FormatError())
1469
1470 def _unlock_file(f):
1471 assert f._lock_file_overlapped_p
1472 handle = msvcrt.get_osfhandle(f.fileno())
1473 if not UnlockFileEx(handle, 0,
1474 whole_low, whole_high, f._lock_file_overlapped_p):
1475 raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
1476
1477 else:
1478 # Some platforms, such as Jython, is missing fcntl
1479 try:
1480 import fcntl
1481
1482 def _lock_file(f, exclusive):
1483 fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
1484
1485 def _unlock_file(f):
1486 fcntl.flock(f, fcntl.LOCK_UN)
1487 except ImportError:
1488 UNSUPPORTED_MSG = 'file locking is not supported on this platform'
1489
1490 def _lock_file(f, exclusive):
1491 raise IOError(UNSUPPORTED_MSG)
1492
1493 def _unlock_file(f):
1494 raise IOError(UNSUPPORTED_MSG)
1495
1496
1497 class locked_file(object):
1498 def __init__(self, filename, mode, encoding=None):
1499 assert mode in ['r', 'a', 'w']
1500 self.f = io.open(filename, mode, encoding=encoding)
1501 self.mode = mode
1502
1503 def __enter__(self):
1504 exclusive = self.mode != 'r'
1505 try:
1506 _lock_file(self.f, exclusive)
1507 except IOError:
1508 self.f.close()
1509 raise
1510 return self
1511
1512 def __exit__(self, etype, value, traceback):
1513 try:
1514 _unlock_file(self.f)
1515 finally:
1516 self.f.close()
1517
1518 def __iter__(self):
1519 return iter(self.f)
1520
1521 def write(self, *args):
1522 return self.f.write(*args)
1523
1524 def read(self, *args):
1525 return self.f.read(*args)
1526
1527
1528 def get_filesystem_encoding():
1529 encoding = sys.getfilesystemencoding()
1530 return encoding if encoding is not None else 'utf-8'
1531
1532
1533 def shell_quote(args):
1534 quoted_args = []
1535 encoding = get_filesystem_encoding()
1536 for a in args:
1537 if isinstance(a, bytes):
1538 # We may get a filename encoded with 'encodeFilename'
1539 a = a.decode(encoding)
1540 quoted_args.append(compat_shlex_quote(a))
1541 return ' '.join(quoted_args)
1542
1543
1544 def smuggle_url(url, data):
1545 """ Pass additional data in a URL for internal use. """
1546
1547 url, idata = unsmuggle_url(url, {})
1548 data.update(idata)
1549 sdata = compat_urllib_parse_urlencode(
1550 {'__youtubedl_smuggle': json.dumps(data)})
1551 return url + '#' + sdata
1552
1553
1554 def unsmuggle_url(smug_url, default=None):
1555 if '#__youtubedl_smuggle' not in smug_url:
1556 return smug_url, default
1557 url, _, sdata = smug_url.rpartition('#')
1558 jsond = compat_parse_qs(sdata)['__youtubedl_smuggle'][0]
1559 data = json.loads(jsond)
1560 return url, data
1561
1562
1563 def format_bytes(bytes):
1564 if bytes is None:
1565 return 'N/A'
1566 if type(bytes) is str:
1567 bytes = float(bytes)
1568 if bytes == 0.0:
1569 exponent = 0
1570 else:
1571 exponent = int(math.log(bytes, 1024.0))
1572 suffix = ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent]
1573 converted = float(bytes) / float(1024 ** exponent)
1574 return '%.2f%s' % (converted, suffix)
1575
1576
1577 def lookup_unit_table(unit_table, s):
1578 units_re = '|'.join(re.escape(u) for u in unit_table)
1579 m = re.match(
1580 r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re, s)
1581 if not m:
1582 return None
1583 num_str = m.group('num').replace(',', '.')
1584 mult = unit_table[m.group('unit')]
1585 return int(float(num_str) * mult)
1586
1587
1588 def parse_filesize(s):
1589 if s is None:
1590 return None
1591
1592 # The lower-case forms are of course incorrect and unofficial,
1593 # but we support those too
1594 _UNIT_TABLE = {
1595 'B': 1,
1596 'b': 1,
1597 'bytes': 1,
1598 'KiB': 1024,
1599 'KB': 1000,
1600 'kB': 1024,
1601 'Kb': 1000,
1602 'kb': 1000,
1603 'kilobytes': 1000,
1604 'kibibytes': 1024,
1605 'MiB': 1024 ** 2,
1606 'MB': 1000 ** 2,
1607 'mB': 1024 ** 2,
1608 'Mb': 1000 ** 2,
1609 'mb': 1000 ** 2,
1610 'megabytes': 1000 ** 2,
1611 'mebibytes': 1024 ** 2,
1612 'GiB': 1024 ** 3,
1613 'GB': 1000 ** 3,
1614 'gB': 1024 ** 3,
1615 'Gb': 1000 ** 3,
1616 'gb': 1000 ** 3,
1617 'gigabytes': 1000 ** 3,
1618 'gibibytes': 1024 ** 3,
1619 'TiB': 1024 ** 4,
1620 'TB': 1000 ** 4,
1621 'tB': 1024 ** 4,
1622 'Tb': 1000 ** 4,
1623 'tb': 1000 ** 4,
1624 'terabytes': 1000 ** 4,
1625 'tebibytes': 1024 ** 4,
1626 'PiB': 1024 ** 5,
1627 'PB': 1000 ** 5,
1628 'pB': 1024 ** 5,
1629 'Pb': 1000 ** 5,
1630 'pb': 1000 ** 5,
1631 'petabytes': 1000 ** 5,
1632 'pebibytes': 1024 ** 5,
1633 'EiB': 1024 ** 6,
1634 'EB': 1000 ** 6,
1635 'eB': 1024 ** 6,
1636 'Eb': 1000 ** 6,
1637 'eb': 1000 ** 6,
1638 'exabytes': 1000 ** 6,
1639 'exbibytes': 1024 ** 6,
1640 'ZiB': 1024 ** 7,
1641 'ZB': 1000 ** 7,
1642 'zB': 1024 ** 7,
1643 'Zb': 1000 ** 7,
1644 'zb': 1000 ** 7,
1645 'zettabytes': 1000 ** 7,
1646 'zebibytes': 1024 ** 7,
1647 'YiB': 1024 ** 8,
1648 'YB': 1000 ** 8,
1649 'yB': 1024 ** 8,
1650 'Yb': 1000 ** 8,
1651 'yb': 1000 ** 8,
1652 'yottabytes': 1000 ** 8,
1653 'yobibytes': 1024 ** 8,
1654 }
1655
1656 return lookup_unit_table(_UNIT_TABLE, s)
1657
1658
1659 def parse_count(s):
1660 if s is None:
1661 return None
1662
1663 s = s.strip()
1664
1665 if re.match(r'^[\d,.]+$', s):
1666 return str_to_int(s)
1667
1668 _UNIT_TABLE = {
1669 'k': 1000,
1670 'K': 1000,
1671 'm': 1000 ** 2,
1672 'M': 1000 ** 2,
1673 'kk': 1000 ** 2,
1674 'KK': 1000 ** 2,
1675 }
1676
1677 return lookup_unit_table(_UNIT_TABLE, s)
1678
1679
1680 def month_by_name(name, lang='en'):
1681 """ Return the number of a month by (locale-independently) English name """
1682
1683 month_names = MONTH_NAMES.get(lang, MONTH_NAMES['en'])
1684
1685 try:
1686 return month_names.index(name) + 1
1687 except ValueError:
1688 return None
1689
1690
1691 def month_by_abbreviation(abbrev):
1692 """ Return the number of a month by (locale-independently) English
1693 abbreviations """
1694
1695 try:
1696 return [s[:3] for s in ENGLISH_MONTH_NAMES].index(abbrev) + 1
1697 except ValueError:
1698 return None
1699
1700
1701 def fix_xml_ampersands(xml_str):
1702 """Replace all the '&' by '&amp;' in XML"""
1703 return re.sub(
1704 r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1705 '&amp;',
1706 xml_str)
1707
1708
1709 def setproctitle(title):
1710 assert isinstance(title, compat_str)
1711
1712 # ctypes in Jython is not complete
1713 # http://bugs.jython.org/issue2148
1714 if sys.platform.startswith('java'):
1715 return
1716
1717 try:
1718 libc = ctypes.cdll.LoadLibrary('libc.so.6')
1719 except OSError:
1720 return
1721 except TypeError:
1722 # LoadLibrary in Windows Python 2.7.13 only expects
1723 # a bytestring, but since unicode_literals turns
1724 # every string into a unicode string, it fails.
1725 return
1726 title_bytes = title.encode('utf-8')
1727 buf = ctypes.create_string_buffer(len(title_bytes))
1728 buf.value = title_bytes
1729 try:
1730 libc.prctl(15, buf, 0, 0, 0)
1731 except AttributeError:
1732 return # Strange libc, just skip this
1733
1734
1735 def remove_start(s, start):
1736 return s[len(start):] if s is not None and s.startswith(start) else s
1737
1738
1739 def remove_end(s, end):
1740 return s[:-len(end)] if s is not None and s.endswith(end) else s
1741
1742
1743 def remove_quotes(s):
1744 if s is None or len(s) < 2:
1745 return s
1746 for quote in ('"', "'", ):
1747 if s[0] == quote and s[-1] == quote:
1748 return s[1:-1]
1749 return s
1750
1751
1752 def url_basename(url):
1753 path = compat_urlparse.urlparse(url).path
1754 return path.strip('/').split('/')[-1]
1755
1756
1757 def base_url(url):
1758 return re.match(r'https?://[^?#&]+/', url).group()
1759
1760
1761 def urljoin(base, path):
1762 if isinstance(path, bytes):
1763 path = path.decode('utf-8')
1764 if not isinstance(path, compat_str) or not path:
1765 return None
1766 if re.match(r'^(?:https?:)?//', path):
1767 return path
1768 if isinstance(base, bytes):
1769 base = base.decode('utf-8')
1770 if not isinstance(base, compat_str) or not re.match(
1771 r'^(?:https?:)?//', base):
1772 return None
1773 return compat_urlparse.urljoin(base, path)
1774
1775
1776 class HEADRequest(compat_urllib_request.Request):
1777 def get_method(self):
1778 return 'HEAD'
1779
1780
1781 class PUTRequest(compat_urllib_request.Request):
1782 def get_method(self):
1783 return 'PUT'
1784
1785
1786 def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
1787 if get_attr:
1788 if v is not None:
1789 v = getattr(v, get_attr, None)
1790 if v == '':
1791 v = None
1792 if v is None:
1793 return default
1794 try:
1795 return int(v) * invscale // scale
1796 except ValueError:
1797 return default
1798
1799
1800 def str_or_none(v, default=None):
1801 return default if v is None else compat_str(v)
1802
1803
1804 def str_to_int(int_str):
1805 """ A more relaxed version of int_or_none """
1806 if int_str is None:
1807 return None
1808 int_str = re.sub(r'[,\.\+]', '', int_str)
1809 return int(int_str)
1810
1811
1812 def float_or_none(v, scale=1, invscale=1, default=None):
1813 if v is None:
1814 return default
1815 try:
1816 return float(v) * invscale / scale
1817 except ValueError:
1818 return default
1819
1820
1821 def bool_or_none(v, default=None):
1822 return v if isinstance(v, bool) else default
1823
1824
1825 def strip_or_none(v):
1826 return None if v is None else v.strip()
1827
1828
1829 def parse_duration(s):
1830 if not isinstance(s, compat_basestring):
1831 return None
1832
1833 s = s.strip()
1834
1835 days, hours, mins, secs, ms = [None] * 5
1836 m = re.match(r'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?Z?$', s)
1837 if m:
1838 days, hours, mins, secs, ms = m.groups()
1839 else:
1840 m = re.match(
1841 r'''(?ix)(?:P?
1842 (?:
1843 [0-9]+\s*y(?:ears?)?\s*
1844 )?
1845 (?:
1846 [0-9]+\s*m(?:onths?)?\s*
1847 )?
1848 (?:
1849 [0-9]+\s*w(?:eeks?)?\s*
1850 )?
1851 (?:
1852 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
1853 )?
1854 T)?
1855 (?:
1856 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
1857 )?
1858 (?:
1859 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
1860 )?
1861 (?:
1862 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
1863 )?Z?$''', s)
1864 if m:
1865 days, hours, mins, secs, ms = m.groups()
1866 else:
1867 m = re.match(r'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s)
1868 if m:
1869 hours, mins = m.groups()
1870 else:
1871 return None
1872
1873 duration = 0
1874 if secs:
1875 duration += float(secs)
1876 if mins:
1877 duration += float(mins) * 60
1878 if hours:
1879 duration += float(hours) * 60 * 60
1880 if days:
1881 duration += float(days) * 24 * 60 * 60
1882 if ms:
1883 duration += float(ms)
1884 return duration
1885
1886
1887 def prepend_extension(filename, ext, expected_real_ext=None):
1888 name, real_ext = os.path.splitext(filename)
1889 return (
1890 '{0}.{1}{2}'.format(name, ext, real_ext)
1891 if not expected_real_ext or real_ext[1:] == expected_real_ext
1892 else '{0}.{1}'.format(filename, ext))
1893
1894
1895 def replace_extension(filename, ext, expected_real_ext=None):
1896 name, real_ext = os.path.splitext(filename)
1897 return '{0}.{1}'.format(
1898 name if not expected_real_ext or real_ext[1:] == expected_real_ext else filename,
1899 ext)
1900
1901
1902 def check_executable(exe, args=[]):
1903 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
1904 args can be a list of arguments for a short output (like -version) """
1905 try:
1906 subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
1907 except OSError:
1908 return False
1909 return exe
1910
1911
1912 def get_exe_version(exe, args=['--version'],
1913 version_re=None, unrecognized='present'):
1914 """ Returns the version of the specified executable,
1915 or False if the executable is not present """
1916 try:
1917 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
1918 # SIGTTOU if youtube-dl is run in the background.
1919 # See https://github.com/rg3/youtube-dl/issues/955#issuecomment-209789656
1920 out, _ = subprocess.Popen(
1921 [encodeArgument(exe)] + args,
1922 stdin=subprocess.PIPE,
1923 stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
1924 except OSError:
1925 return False
1926 if isinstance(out, bytes): # Python 2.x
1927 out = out.decode('ascii', 'ignore')
1928 return detect_exe_version(out, version_re, unrecognized)
1929
1930
1931 def detect_exe_version(output, version_re=None, unrecognized='present'):
1932 assert isinstance(output, compat_str)
1933 if version_re is None:
1934 version_re = r'version\s+([-0-9._a-zA-Z]+)'
1935 m = re.search(version_re, output)
1936 if m:
1937 return m.group(1)
1938 else:
1939 return unrecognized
1940
1941
1942 class PagedList(object):
1943 def __len__(self):
1944 # This is only useful for tests
1945 return len(self.getslice())
1946
1947
1948 class OnDemandPagedList(PagedList):
1949 def __init__(self, pagefunc, pagesize, use_cache=True):
1950 self._pagefunc = pagefunc
1951 self._pagesize = pagesize
1952 self._use_cache = use_cache
1953 if use_cache:
1954 self._cache = {}
1955
1956 def getslice(self, start=0, end=None):
1957 res = []
1958 for pagenum in itertools.count(start // self._pagesize):
1959 firstid = pagenum * self._pagesize
1960 nextfirstid = pagenum * self._pagesize + self._pagesize
1961 if start >= nextfirstid:
1962 continue
1963
1964 page_results = None
1965 if self._use_cache:
1966 page_results = self._cache.get(pagenum)
1967 if page_results is None:
1968 page_results = list(self._pagefunc(pagenum))
1969 if self._use_cache:
1970 self._cache[pagenum] = page_results
1971
1972 startv = (
1973 start % self._pagesize
1974 if firstid <= start < nextfirstid
1975 else 0)
1976
1977 endv = (
1978 ((end - 1) % self._pagesize) + 1
1979 if (end is not None and firstid <= end <= nextfirstid)
1980 else None)
1981
1982 if startv != 0 or endv is not None:
1983 page_results = page_results[startv:endv]
1984 res.extend(page_results)
1985
1986 # A little optimization - if current page is not "full", ie. does
1987 # not contain page_size videos then we can assume that this page
1988 # is the last one - there are no more ids on further pages -
1989 # i.e. no need to query again.
1990 if len(page_results) + startv < self._pagesize:
1991 break
1992
1993 # If we got the whole page, but the next page is not interesting,
1994 # break out early as well
1995 if end == nextfirstid:
1996 break
1997 return res
1998
1999
2000 class InAdvancePagedList(PagedList):
2001 def __init__(self, pagefunc, pagecount, pagesize):
2002 self._pagefunc = pagefunc
2003 self._pagecount = pagecount
2004 self._pagesize = pagesize
2005
2006 def getslice(self, start=0, end=None):
2007 res = []
2008 start_page = start // self._pagesize
2009 end_page = (
2010 self._pagecount if end is None else (end // self._pagesize + 1))
2011 skip_elems = start - start_page * self._pagesize
2012 only_more = None if end is None else end - start
2013 for pagenum in range(start_page, end_page):
2014 page = list(self._pagefunc(pagenum))
2015 if skip_elems:
2016 page = page[skip_elems:]
2017 skip_elems = None
2018 if only_more is not None:
2019 if len(page) < only_more:
2020 only_more -= len(page)
2021 else:
2022 page = page[:only_more]
2023 res.extend(page)
2024 break
2025 res.extend(page)
2026 return res
2027
2028
2029 def uppercase_escape(s):
2030 unicode_escape = codecs.getdecoder('unicode_escape')
2031 return re.sub(
2032 r'\\U[0-9a-fA-F]{8}',
2033 lambda m: unicode_escape(m.group(0))[0],
2034 s)
2035
2036
2037 def lowercase_escape(s):
2038 unicode_escape = codecs.getdecoder('unicode_escape')
2039 return re.sub(
2040 r'\\u[0-9a-fA-F]{4}',
2041 lambda m: unicode_escape(m.group(0))[0],
2042 s)
2043
2044
2045 def escape_rfc3986(s):
2046 """Escape non-ASCII characters as suggested by RFC 3986"""
2047 if sys.version_info < (3, 0) and isinstance(s, compat_str):
2048 s = s.encode('utf-8')
2049 return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]")
2050
2051
2052 def escape_url(url):
2053 """Escape URL as suggested by RFC 3986"""
2054 url_parsed = compat_urllib_parse_urlparse(url)
2055 return url_parsed._replace(
2056 netloc=url_parsed.netloc.encode('idna').decode('ascii'),
2057 path=escape_rfc3986(url_parsed.path),
2058 params=escape_rfc3986(url_parsed.params),
2059 query=escape_rfc3986(url_parsed.query),
2060 fragment=escape_rfc3986(url_parsed.fragment)
2061 ).geturl()
2062
2063
2064 def read_batch_urls(batch_fd):
2065 def fixup(url):
2066 if not isinstance(url, compat_str):
2067 url = url.decode('utf-8', 'replace')
2068 BOM_UTF8 = '\xef\xbb\xbf'
2069 if url.startswith(BOM_UTF8):
2070 url = url[len(BOM_UTF8):]
2071 url = url.strip()
2072 if url.startswith(('#', ';', ']')):
2073 return False
2074 return url
2075
2076 with contextlib.closing(batch_fd) as fd:
2077 return [url for url in map(fixup, fd) if url]
2078
2079
2080 def urlencode_postdata(*args, **kargs):
2081 return compat_urllib_parse_urlencode(*args, **kargs).encode('ascii')
2082
2083
2084 def update_url_query(url, query):
2085 if not query:
2086 return url
2087 parsed_url = compat_urlparse.urlparse(url)
2088 qs = compat_parse_qs(parsed_url.query)
2089 qs.update(query)
2090 return compat_urlparse.urlunparse(parsed_url._replace(
2091 query=compat_urllib_parse_urlencode(qs, True)))
2092
2093
2094 def update_Request(req, url=None, data=None, headers={}, query={}):
2095 req_headers = req.headers.copy()
2096 req_headers.update(headers)
2097 req_data = data or req.data
2098 req_url = update_url_query(url or req.get_full_url(), query)
2099 req_get_method = req.get_method()
2100 if req_get_method == 'HEAD':
2101 req_type = HEADRequest
2102 elif req_get_method == 'PUT':
2103 req_type = PUTRequest
2104 else:
2105 req_type = compat_urllib_request.Request
2106 new_req = req_type(
2107 req_url, data=req_data, headers=req_headers,
2108 origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
2109 if hasattr(req, 'timeout'):
2110 new_req.timeout = req.timeout
2111 return new_req
2112
2113
2114 def _multipart_encode_impl(data, boundary):
2115 content_type = 'multipart/form-data; boundary=%s' % boundary
2116
2117 out = b''
2118 for k, v in data.items():
2119 out += b'--' + boundary.encode('ascii') + b'\r\n'
2120 if isinstance(k, compat_str):
2121 k = k.encode('utf-8')
2122 if isinstance(v, compat_str):
2123 v = v.encode('utf-8')
2124 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2125 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2126 content = b'Content-Disposition: form-data; name="' + k + b'"\r\n\r\n' + v + b'\r\n'
2127 if boundary.encode('ascii') in content:
2128 raise ValueError('Boundary overlaps with data')
2129 out += content
2130
2131 out += b'--' + boundary.encode('ascii') + b'--\r\n'
2132
2133 return out, content_type
2134
2135
2136 def multipart_encode(data, boundary=None):
2137 '''
2138 Encode a dict to RFC 7578-compliant form-data
2139
2140 data:
2141 A dict where keys and values can be either Unicode or bytes-like
2142 objects.
2143 boundary:
2144 If specified a Unicode object, it's used as the boundary. Otherwise
2145 a random boundary is generated.
2146
2147 Reference: https://tools.ietf.org/html/rfc7578
2148 '''
2149 has_specified_boundary = boundary is not None
2150
2151 while True:
2152 if boundary is None:
2153 boundary = '---------------' + str(random.randrange(0x0fffffff, 0xffffffff))
2154
2155 try:
2156 out, content_type = _multipart_encode_impl(data, boundary)
2157 break
2158 except ValueError:
2159 if has_specified_boundary:
2160 raise
2161 boundary = None
2162
2163 return out, content_type
2164
2165
2166 def dict_get(d, key_or_keys, default=None, skip_false_values=True):
2167 if isinstance(key_or_keys, (list, tuple)):
2168 for key in key_or_keys:
2169 if key not in d or d[key] is None or skip_false_values and not d[key]:
2170 continue
2171 return d[key]
2172 return default
2173 return d.get(key_or_keys, default)
2174
2175
2176 def try_get(src, getter, expected_type=None):
2177 if not isinstance(getter, (list, tuple)):
2178 getter = [getter]
2179 for get in getter:
2180 try:
2181 v = get(src)
2182 except (AttributeError, KeyError, TypeError, IndexError):
2183 pass
2184 else:
2185 if expected_type is None or isinstance(v, expected_type):
2186 return v
2187
2188
2189 def encode_compat_str(string, encoding=preferredencoding(), errors='strict'):
2190 return string if isinstance(string, compat_str) else compat_str(string, encoding, errors)
2191
2192
2193 US_RATINGS = {
2194 'G': 0,
2195 'PG': 10,
2196 'PG-13': 13,
2197 'R': 16,
2198 'NC': 18,
2199 }
2200
2201
2202 TV_PARENTAL_GUIDELINES = {
2203 'TV-Y': 0,
2204 'TV-Y7': 7,
2205 'TV-G': 0,
2206 'TV-PG': 0,
2207 'TV-14': 14,
2208 'TV-MA': 17,
2209 }
2210
2211
2212 def parse_age_limit(s):
2213 if type(s) == int:
2214 return s if 0 <= s <= 21 else None
2215 if not isinstance(s, compat_basestring):
2216 return None
2217 m = re.match(r'^(?P<age>\d{1,2})\+?$', s)
2218 if m:
2219 return int(m.group('age'))
2220 if s in US_RATINGS:
2221 return US_RATINGS[s]
2222 return TV_PARENTAL_GUIDELINES.get(s)
2223
2224
2225 def strip_jsonp(code):
2226 return re.sub(
2227 r'''(?sx)^
2228 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]+)
2229 (?:\s*&&\s*(?P=func_name))?
2230 \s*\(\s*(?P<callback_data>.*)\);?
2231 \s*?(?://[^\n]*)*$''',
2232 r'\g<callback_data>', code)
2233
2234
2235 def js_to_json(code):
2236 COMMENT_RE = r'/\*(?:(?!\*/).)*?\*/|//[^\n]*'
2237 SKIP_RE = r'\s*(?:{comment})?\s*'.format(comment=COMMENT_RE)
2238 INTEGER_TABLE = (
2239 (r'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip=SKIP_RE), 16),
2240 (r'(?s)^(0+[0-7]+){skip}:?$'.format(skip=SKIP_RE), 8),
2241 )
2242
2243 def fix_kv(m):
2244 v = m.group(0)
2245 if v in ('true', 'false', 'null'):
2246 return v
2247 elif v.startswith('/*') or v.startswith('//') or v == ',':
2248 return ""
2249
2250 if v[0] in ("'", '"'):
2251 v = re.sub(r'(?s)\\.|"', lambda m: {
2252 '"': '\\"',
2253 "\\'": "'",
2254 '\\\n': '',
2255 '\\x': '\\u00',
2256 }.get(m.group(0), m.group(0)), v[1:-1])
2257
2258 for regex, base in INTEGER_TABLE:
2259 im = re.match(regex, v)
2260 if im:
2261 i = int(im.group(1), base)
2262 return '"%d":' % i if v.endswith(':') else '%d' % i
2263
2264 return '"%s"' % v
2265
2266 return re.sub(r'''(?sx)
2267 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
2268 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
2269 {comment}|,(?={skip}[\]}}])|
2270 (?:(?<![0-9])[eE]|[a-df-zA-DF-Z_])[.a-zA-Z_0-9]*|
2271 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
2272 [0-9]+(?={skip}:)
2273 '''.format(comment=COMMENT_RE, skip=SKIP_RE), fix_kv, code)
2274
2275
2276 def qualities(quality_ids):
2277 """ Get a numeric quality value out of a list of possible values """
2278 def q(qid):
2279 try:
2280 return quality_ids.index(qid)
2281 except ValueError:
2282 return -1
2283 return q
2284
2285
2286 DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s'
2287
2288
2289 def limit_length(s, length):
2290 """ Add ellipses to overly long strings """
2291 if s is None:
2292 return None
2293 ELLIPSES = '...'
2294 if len(s) > length:
2295 return s[:length - len(ELLIPSES)] + ELLIPSES
2296 return s
2297
2298
2299 def version_tuple(v):
2300 return tuple(int(e) for e in re.split(r'[-.]', v))
2301
2302
2303 def is_outdated_version(version, limit, assume_new=True):
2304 if not version:
2305 return not assume_new
2306 try:
2307 return version_tuple(version) < version_tuple(limit)
2308 except ValueError:
2309 return not assume_new
2310
2311
2312 def ytdl_is_updateable():
2313 """ Returns if youtube-dl can be updated with -U """
2314 from zipimport import zipimporter
2315
2316 return isinstance(globals().get('__loader__'), zipimporter) or hasattr(sys, 'frozen')
2317
2318
2319 def args_to_str(args):
2320 # Get a short string representation for a subprocess command
2321 return ' '.join(compat_shlex_quote(a) for a in args)
2322
2323
2324 def error_to_compat_str(err):
2325 err_str = str(err)
2326 # On python 2 error byte string must be decoded with proper
2327 # encoding rather than ascii
2328 if sys.version_info[0] < 3:
2329 err_str = err_str.decode(preferredencoding())
2330 return err_str
2331
2332
2333 def mimetype2ext(mt):
2334 if mt is None:
2335 return None
2336
2337 ext = {
2338 'audio/mp4': 'm4a',
2339 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
2340 # it's the most popular one
2341 'audio/mpeg': 'mp3',
2342 }.get(mt)
2343 if ext is not None:
2344 return ext
2345
2346 _, _, res = mt.rpartition('/')
2347 res = res.split(';')[0].strip().lower()
2348
2349 return {
2350 '3gpp': '3gp',
2351 'smptett+xml': 'tt',
2352 'ttaf+xml': 'dfxp',
2353 'ttml+xml': 'ttml',
2354 'x-flv': 'flv',
2355 'x-mp4-fragmented': 'mp4',
2356 'x-ms-sami': 'sami',
2357 'x-ms-wmv': 'wmv',
2358 'mpegurl': 'm3u8',
2359 'x-mpegurl': 'm3u8',
2360 'vnd.apple.mpegurl': 'm3u8',
2361 'dash+xml': 'mpd',
2362 'f4m+xml': 'f4m',
2363 'hds+xml': 'f4m',
2364 'vnd.ms-sstr+xml': 'ism',
2365 'quicktime': 'mov',
2366 'mp2t': 'ts',
2367 }.get(res, res)
2368
2369
2370 def parse_codecs(codecs_str):
2371 # http://tools.ietf.org/html/rfc6381
2372 if not codecs_str:
2373 return {}
2374 splited_codecs = list(filter(None, map(
2375 lambda str: str.strip(), codecs_str.strip().strip(',').split(','))))
2376 vcodec, acodec = None, None
2377 for full_codec in splited_codecs:
2378 codec = full_codec.split('.')[0]
2379 if codec in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v', 'hvc1'):
2380 if not vcodec:
2381 vcodec = full_codec
2382 elif codec in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
2383 if not acodec:
2384 acodec = full_codec
2385 else:
2386 write_string('WARNING: Unknown codec %s\n' % full_codec, sys.stderr)
2387 if not vcodec and not acodec:
2388 if len(splited_codecs) == 2:
2389 return {
2390 'vcodec': vcodec,
2391 'acodec': acodec,
2392 }
2393 elif len(splited_codecs) == 1:
2394 return {
2395 'vcodec': 'none',
2396 'acodec': vcodec,
2397 }
2398 else:
2399 return {
2400 'vcodec': vcodec or 'none',
2401 'acodec': acodec or 'none',
2402 }
2403 return {}
2404
2405
2406 def urlhandle_detect_ext(url_handle):
2407 getheader = url_handle.headers.get
2408
2409 cd = getheader('Content-Disposition')
2410 if cd:
2411 m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd)
2412 if m:
2413 e = determine_ext(m.group('filename'), default_ext=None)
2414 if e:
2415 return e
2416
2417 return mimetype2ext(getheader('Content-Type'))
2418
2419
2420 def encode_data_uri(data, mime_type):
2421 return 'data:%s;base64,%s' % (mime_type, base64.b64encode(data).decode('ascii'))
2422
2423
2424 def age_restricted(content_limit, age_limit):
2425 """ Returns True iff the content should be blocked """
2426
2427 if age_limit is None: # No limit set
2428 return False
2429 if content_limit is None:
2430 return False # Content available for everyone
2431 return age_limit < content_limit
2432
2433
2434 def is_html(first_bytes):
2435 """ Detect whether a file contains HTML by examining its first bytes. """
2436
2437 BOMS = [
2438 (b'\xef\xbb\xbf', 'utf-8'),
2439 (b'\x00\x00\xfe\xff', 'utf-32-be'),
2440 (b'\xff\xfe\x00\x00', 'utf-32-le'),
2441 (b'\xff\xfe', 'utf-16-le'),
2442 (b'\xfe\xff', 'utf-16-be'),
2443 ]
2444 for bom, enc in BOMS:
2445 if first_bytes.startswith(bom):
2446 s = first_bytes[len(bom):].decode(enc, 'replace')
2447 break
2448 else:
2449 s = first_bytes.decode('utf-8', 'replace')
2450
2451 return re.match(r'^\s*<', s)
2452
2453
2454 def determine_protocol(info_dict):
2455 protocol = info_dict.get('protocol')
2456 if protocol is not None:
2457 return protocol
2458
2459 url = info_dict['url']
2460 if url.startswith('rtmp'):
2461 return 'rtmp'
2462 elif url.startswith('mms'):
2463 return 'mms'
2464 elif url.startswith('rtsp'):
2465 return 'rtsp'
2466
2467 ext = determine_ext(url)
2468 if ext == 'm3u8':
2469 return 'm3u8'
2470 elif ext == 'f4m':
2471 return 'f4m'
2472
2473 return compat_urllib_parse_urlparse(url).scheme
2474
2475
2476 def render_table(header_row, data):
2477 """ Render a list of rows, each as a list of values """
2478 table = [header_row] + data
2479 max_lens = [max(len(compat_str(v)) for v in col) for col in zip(*table)]
2480 format_str = ' '.join('%-' + compat_str(ml + 1) + 's' for ml in max_lens[:-1]) + '%s'
2481 return '\n'.join(format_str % tuple(row) for row in table)
2482
2483
2484 def _match_one(filter_part, dct):
2485 COMPARISON_OPERATORS = {
2486 '<': operator.lt,
2487 '<=': operator.le,
2488 '>': operator.gt,
2489 '>=': operator.ge,
2490 '=': operator.eq,
2491 '!=': operator.ne,
2492 }
2493 operator_rex = re.compile(r'''(?x)\s*
2494 (?P<key>[a-z_]+)
2495 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
2496 (?:
2497 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
2498 (?P<quote>["\'])(?P<quotedstrval>(?:\\.|(?!(?P=quote)|\\).)+?)(?P=quote)|
2499 (?P<strval>(?![0-9.])[a-z0-9A-Z]*)
2500 )
2501 \s*$
2502 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
2503 m = operator_rex.search(filter_part)
2504 if m:
2505 op = COMPARISON_OPERATORS[m.group('op')]
2506 actual_value = dct.get(m.group('key'))
2507 if (m.group('quotedstrval') is not None or
2508 m.group('strval') is not None or
2509 # If the original field is a string and matching comparisonvalue is
2510 # a number we should respect the origin of the original field
2511 # and process comparison value as a string (see
2512 # https://github.com/rg3/youtube-dl/issues/11082).
2513 actual_value is not None and m.group('intval') is not None and
2514 isinstance(actual_value, compat_str)):
2515 if m.group('op') not in ('=', '!='):
2516 raise ValueError(
2517 'Operator %s does not support string values!' % m.group('op'))
2518 comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval')
2519 quote = m.group('quote')
2520 if quote is not None:
2521 comparison_value = comparison_value.replace(r'\%s' % quote, quote)
2522 else:
2523 try:
2524 comparison_value = int(m.group('intval'))
2525 except ValueError:
2526 comparison_value = parse_filesize(m.group('intval'))
2527 if comparison_value is None:
2528 comparison_value = parse_filesize(m.group('intval') + 'B')
2529 if comparison_value is None:
2530 raise ValueError(
2531 'Invalid integer value %r in filter part %r' % (
2532 m.group('intval'), filter_part))
2533 if actual_value is None:
2534 return m.group('none_inclusive')
2535 return op(actual_value, comparison_value)
2536
2537 UNARY_OPERATORS = {
2538 '': lambda v: v is not None,
2539 '!': lambda v: v is None,
2540 }
2541 operator_rex = re.compile(r'''(?x)\s*
2542 (?P<op>%s)\s*(?P<key>[a-z_]+)
2543 \s*$
2544 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
2545 m = operator_rex.search(filter_part)
2546 if m:
2547 op = UNARY_OPERATORS[m.group('op')]
2548 actual_value = dct.get(m.group('key'))
2549 return op(actual_value)
2550
2551 raise ValueError('Invalid filter part %r' % filter_part)
2552
2553
2554 def match_str(filter_str, dct):
2555 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """
2556
2557 return all(
2558 _match_one(filter_part, dct) for filter_part in filter_str.split('&'))
2559
2560
2561 def match_filter_func(filter_str):
2562 def _match_func(info_dict):
2563 if match_str(filter_str, info_dict):
2564 return None
2565 else:
2566 video_title = info_dict.get('title', info_dict.get('id', 'video'))
2567 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
2568 return _match_func
2569
2570
2571 def parse_dfxp_time_expr(time_expr):
2572 if not time_expr:
2573 return
2574
2575 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
2576 if mobj:
2577 return float(mobj.group('time_offset'))
2578
2579 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
2580 if mobj:
2581 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
2582
2583
2584 def srt_subtitles_timecode(seconds):
2585 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
2586
2587
2588 def dfxp2srt(dfxp_data):
2589 '''
2590 @param dfxp_data A bytes-like object containing DFXP data
2591 @returns A unicode object containing converted SRT data
2592 '''
2593 LEGACY_NAMESPACES = (
2594 (b'http://www.w3.org/ns/ttml', [
2595 b'http://www.w3.org/2004/11/ttaf1',
2596 b'http://www.w3.org/2006/04/ttaf1',
2597 b'http://www.w3.org/2006/10/ttaf1',
2598 ]),
2599 (b'http://www.w3.org/ns/ttml#styling', [
2600 b'http://www.w3.org/ns/ttml#style',
2601 ]),
2602 )
2603
2604 SUPPORTED_STYLING = [
2605 'color',
2606 'fontFamily',
2607 'fontSize',
2608 'fontStyle',
2609 'fontWeight',
2610 'textDecoration'
2611 ]
2612
2613 _x = functools.partial(xpath_with_ns, ns_map={
2614 'ttml': 'http://www.w3.org/ns/ttml',
2615 'tts': 'http://www.w3.org/ns/ttml#styling',
2616 })
2617
2618 styles = {}
2619 default_style = {}
2620
2621 class TTMLPElementParser(object):
2622 _out = ''
2623 _unclosed_elements = []
2624 _applied_styles = []
2625
2626 def start(self, tag, attrib):
2627 if tag in (_x('ttml:br'), 'br'):
2628 self._out += '\n'
2629 else:
2630 unclosed_elements = []
2631 style = {}
2632 element_style_id = attrib.get('style')
2633 if default_style:
2634 style.update(default_style)
2635 if element_style_id:
2636 style.update(styles.get(element_style_id, {}))
2637 for prop in SUPPORTED_STYLING:
2638 prop_val = attrib.get(_x('tts:' + prop))
2639 if prop_val:
2640 style[prop] = prop_val
2641 if style:
2642 font = ''
2643 for k, v in sorted(style.items()):
2644 if self._applied_styles and self._applied_styles[-1].get(k) == v:
2645 continue
2646 if k == 'color':
2647 font += ' color="%s"' % v
2648 elif k == 'fontSize':
2649 font += ' size="%s"' % v
2650 elif k == 'fontFamily':
2651 font += ' face="%s"' % v
2652 elif k == 'fontWeight' and v == 'bold':
2653 self._out += '<b>'
2654 unclosed_elements.append('b')
2655 elif k == 'fontStyle' and v == 'italic':
2656 self._out += '<i>'
2657 unclosed_elements.append('i')
2658 elif k == 'textDecoration' and v == 'underline':
2659 self._out += '<u>'
2660 unclosed_elements.append('u')
2661 if font:
2662 self._out += '<font' + font + '>'
2663 unclosed_elements.append('font')
2664 applied_style = {}
2665 if self._applied_styles:
2666 applied_style.update(self._applied_styles[-1])
2667 applied_style.update(style)
2668 self._applied_styles.append(applied_style)
2669 self._unclosed_elements.append(unclosed_elements)
2670
2671 def end(self, tag):
2672 if tag not in (_x('ttml:br'), 'br'):
2673 unclosed_elements = self._unclosed_elements.pop()
2674 for element in reversed(unclosed_elements):
2675 self._out += '</%s>' % element
2676 if unclosed_elements and self._applied_styles:
2677 self._applied_styles.pop()
2678
2679 def data(self, data):
2680 self._out += data
2681
2682 def close(self):
2683 return self._out.strip()
2684
2685 def parse_node(node):
2686 target = TTMLPElementParser()
2687 parser = xml.etree.ElementTree.XMLParser(target=target)
2688 parser.feed(xml.etree.ElementTree.tostring(node))
2689 return parser.close()
2690
2691 for k, v in LEGACY_NAMESPACES:
2692 for ns in v:
2693 dfxp_data = dfxp_data.replace(ns, k)
2694
2695 dfxp = compat_etree_fromstring(dfxp_data)
2696 out = []
2697 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
2698
2699 if not paras:
2700 raise ValueError('Invalid dfxp/TTML subtitle')
2701
2702 repeat = False
2703 while True:
2704 for style in dfxp.findall(_x('.//ttml:style')):
2705 style_id = style.get('id')
2706 parent_style_id = style.get('style')
2707 if parent_style_id:
2708 if parent_style_id not in styles:
2709 repeat = True
2710 continue
2711 styles[style_id] = styles[parent_style_id].copy()
2712 for prop in SUPPORTED_STYLING:
2713 prop_val = style.get(_x('tts:' + prop))
2714 if prop_val:
2715 styles.setdefault(style_id, {})[prop] = prop_val
2716 if repeat:
2717 repeat = False
2718 else:
2719 break
2720
2721 for p in ('body', 'div'):
2722 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
2723 if ele is None:
2724 continue
2725 style = styles.get(ele.get('style'))
2726 if not style:
2727 continue
2728 default_style.update(style)
2729
2730 for para, index in zip(paras, itertools.count(1)):
2731 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
2732 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
2733 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
2734 if begin_time is None:
2735 continue
2736 if not end_time:
2737 if not dur:
2738 continue
2739 end_time = begin_time + dur
2740 out.append('%d\n%s --> %s\n%s\n\n' % (
2741 index,
2742 srt_subtitles_timecode(begin_time),
2743 srt_subtitles_timecode(end_time),
2744 parse_node(para)))
2745
2746 return ''.join(out)
2747
2748
2749 def cli_option(params, command_option, param):
2750 param = params.get(param)
2751 if param:
2752 param = compat_str(param)
2753 return [command_option, param] if param is not None else []
2754
2755
2756 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
2757 param = params.get(param)
2758 if param is None:
2759 return []
2760 assert isinstance(param, bool)
2761 if separator:
2762 return [command_option + separator + (true_value if param else false_value)]
2763 return [command_option, true_value if param else false_value]
2764
2765
2766 def cli_valueless_option(params, command_option, param, expected_value=True):
2767 param = params.get(param)
2768 return [command_option] if param == expected_value else []
2769
2770
2771 def cli_configuration_args(params, param, default=[]):
2772 ex_args = params.get(param)
2773 if ex_args is None:
2774 return default
2775 assert isinstance(ex_args, list)
2776 return ex_args
2777
2778
2779 class ISO639Utils(object):
2780 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
2781 _lang_map = {
2782 'aa': 'aar',
2783 'ab': 'abk',
2784 'ae': 'ave',
2785 'af': 'afr',
2786 'ak': 'aka',
2787 'am': 'amh',
2788 'an': 'arg',
2789 'ar': 'ara',
2790 'as': 'asm',
2791 'av': 'ava',
2792 'ay': 'aym',
2793 'az': 'aze',
2794 'ba': 'bak',
2795 'be': 'bel',
2796 'bg': 'bul',
2797 'bh': 'bih',
2798 'bi': 'bis',
2799 'bm': 'bam',
2800 'bn': 'ben',
2801 'bo': 'bod',
2802 'br': 'bre',
2803 'bs': 'bos',
2804 'ca': 'cat',
2805 'ce': 'che',
2806 'ch': 'cha',
2807 'co': 'cos',
2808 'cr': 'cre',
2809 'cs': 'ces',
2810 'cu': 'chu',
2811 'cv': 'chv',
2812 'cy': 'cym',
2813 'da': 'dan',
2814 'de': 'deu',
2815 'dv': 'div',
2816 'dz': 'dzo',
2817 'ee': 'ewe',
2818 'el': 'ell',
2819 'en': 'eng',
2820 'eo': 'epo',
2821 'es': 'spa',
2822 'et': 'est',
2823 'eu': 'eus',
2824 'fa': 'fas',
2825 'ff': 'ful',
2826 'fi': 'fin',
2827 'fj': 'fij',
2828 'fo': 'fao',
2829 'fr': 'fra',
2830 'fy': 'fry',
2831 'ga': 'gle',
2832 'gd': 'gla',
2833 'gl': 'glg',
2834 'gn': 'grn',
2835 'gu': 'guj',
2836 'gv': 'glv',
2837 'ha': 'hau',
2838 'he': 'heb',
2839 'hi': 'hin',
2840 'ho': 'hmo',
2841 'hr': 'hrv',
2842 'ht': 'hat',
2843 'hu': 'hun',
2844 'hy': 'hye',
2845 'hz': 'her',
2846 'ia': 'ina',
2847 'id': 'ind',
2848 'ie': 'ile',
2849 'ig': 'ibo',
2850 'ii': 'iii',
2851 'ik': 'ipk',
2852 'io': 'ido',
2853 'is': 'isl',
2854 'it': 'ita',
2855 'iu': 'iku',
2856 'ja': 'jpn',
2857 'jv': 'jav',
2858 'ka': 'kat',
2859 'kg': 'kon',
2860 'ki': 'kik',
2861 'kj': 'kua',
2862 'kk': 'kaz',
2863 'kl': 'kal',
2864 'km': 'khm',
2865 'kn': 'kan',
2866 'ko': 'kor',
2867 'kr': 'kau',
2868 'ks': 'kas',
2869 'ku': 'kur',
2870 'kv': 'kom',
2871 'kw': 'cor',
2872 'ky': 'kir',
2873 'la': 'lat',
2874 'lb': 'ltz',
2875 'lg': 'lug',
2876 'li': 'lim',
2877 'ln': 'lin',
2878 'lo': 'lao',
2879 'lt': 'lit',
2880 'lu': 'lub',
2881 'lv': 'lav',
2882 'mg': 'mlg',
2883 'mh': 'mah',
2884 'mi': 'mri',
2885 'mk': 'mkd',
2886 'ml': 'mal',
2887 'mn': 'mon',
2888 'mr': 'mar',
2889 'ms': 'msa',
2890 'mt': 'mlt',
2891 'my': 'mya',
2892 'na': 'nau',
2893 'nb': 'nob',
2894 'nd': 'nde',
2895 'ne': 'nep',
2896 'ng': 'ndo',
2897 'nl': 'nld',
2898 'nn': 'nno',
2899 'no': 'nor',
2900 'nr': 'nbl',
2901 'nv': 'nav',
2902 'ny': 'nya',
2903 'oc': 'oci',
2904 'oj': 'oji',
2905 'om': 'orm',
2906 'or': 'ori',
2907 'os': 'oss',
2908 'pa': 'pan',
2909 'pi': 'pli',
2910 'pl': 'pol',
2911 'ps': 'pus',
2912 'pt': 'por',
2913 'qu': 'que',
2914 'rm': 'roh',
2915 'rn': 'run',
2916 'ro': 'ron',
2917 'ru': 'rus',
2918 'rw': 'kin',
2919 'sa': 'san',
2920 'sc': 'srd',
2921 'sd': 'snd',
2922 'se': 'sme',
2923 'sg': 'sag',
2924 'si': 'sin',
2925 'sk': 'slk',
2926 'sl': 'slv',
2927 'sm': 'smo',
2928 'sn': 'sna',
2929 'so': 'som',
2930 'sq': 'sqi',
2931 'sr': 'srp',
2932 'ss': 'ssw',
2933 'st': 'sot',
2934 'su': 'sun',
2935 'sv': 'swe',
2936 'sw': 'swa',
2937 'ta': 'tam',
2938 'te': 'tel',
2939 'tg': 'tgk',
2940 'th': 'tha',
2941 'ti': 'tir',
2942 'tk': 'tuk',
2943 'tl': 'tgl',
2944 'tn': 'tsn',
2945 'to': 'ton',
2946 'tr': 'tur',
2947 'ts': 'tso',
2948 'tt': 'tat',
2949 'tw': 'twi',
2950 'ty': 'tah',
2951 'ug': 'uig',
2952 'uk': 'ukr',
2953 'ur': 'urd',
2954 'uz': 'uzb',
2955 've': 'ven',
2956 'vi': 'vie',
2957 'vo': 'vol',
2958 'wa': 'wln',
2959 'wo': 'wol',
2960 'xh': 'xho',
2961 'yi': 'yid',
2962 'yo': 'yor',
2963 'za': 'zha',
2964 'zh': 'zho',
2965 'zu': 'zul',
2966 }
2967
2968 @classmethod
2969 def short2long(cls, code):
2970 """Convert language code from ISO 639-1 to ISO 639-2/T"""
2971 return cls._lang_map.get(code[:2])
2972
2973 @classmethod
2974 def long2short(cls, code):
2975 """Convert language code from ISO 639-2/T to ISO 639-1"""
2976 for short_name, long_name in cls._lang_map.items():
2977 if long_name == code:
2978 return short_name
2979
2980
2981 class ISO3166Utils(object):
2982 # From http://data.okfn.org/data/core/country-list
2983 _country_map = {
2984 'AF': 'Afghanistan',
2985 'AX': 'Åland Islands',
2986 'AL': 'Albania',
2987 'DZ': 'Algeria',
2988 'AS': 'American Samoa',
2989 'AD': 'Andorra',
2990 'AO': 'Angola',
2991 'AI': 'Anguilla',
2992 'AQ': 'Antarctica',
2993 'AG': 'Antigua and Barbuda',
2994 'AR': 'Argentina',
2995 'AM': 'Armenia',
2996 'AW': 'Aruba',
2997 'AU': 'Australia',
2998 'AT': 'Austria',
2999 'AZ': 'Azerbaijan',
3000 'BS': 'Bahamas',
3001 'BH': 'Bahrain',
3002 'BD': 'Bangladesh',
3003 'BB': 'Barbados',
3004 'BY': 'Belarus',
3005 'BE': 'Belgium',
3006 'BZ': 'Belize',
3007 'BJ': 'Benin',
3008 'BM': 'Bermuda',
3009 'BT': 'Bhutan',
3010 'BO': 'Bolivia, Plurinational State of',
3011 'BQ': 'Bonaire, Sint Eustatius and Saba',
3012 'BA': 'Bosnia and Herzegovina',
3013 'BW': 'Botswana',
3014 'BV': 'Bouvet Island',
3015 'BR': 'Brazil',
3016 'IO': 'British Indian Ocean Territory',
3017 'BN': 'Brunei Darussalam',
3018 'BG': 'Bulgaria',
3019 'BF': 'Burkina Faso',
3020 'BI': 'Burundi',
3021 'KH': 'Cambodia',
3022 'CM': 'Cameroon',
3023 'CA': 'Canada',
3024 'CV': 'Cape Verde',
3025 'KY': 'Cayman Islands',
3026 'CF': 'Central African Republic',
3027 'TD': 'Chad',
3028 'CL': 'Chile',
3029 'CN': 'China',
3030 'CX': 'Christmas Island',
3031 'CC': 'Cocos (Keeling) Islands',
3032 'CO': 'Colombia',
3033 'KM': 'Comoros',
3034 'CG': 'Congo',
3035 'CD': 'Congo, the Democratic Republic of the',
3036 'CK': 'Cook Islands',
3037 'CR': 'Costa Rica',
3038 'CI': 'Côte d\'Ivoire',
3039 'HR': 'Croatia',
3040 'CU': 'Cuba',
3041 'CW': 'Curaçao',
3042 'CY': 'Cyprus',
3043 'CZ': 'Czech Republic',
3044 'DK': 'Denmark',
3045 'DJ': 'Djibouti',
3046 'DM': 'Dominica',
3047 'DO': 'Dominican Republic',
3048 'EC': 'Ecuador',
3049 'EG': 'Egypt',
3050 'SV': 'El Salvador',
3051 'GQ': 'Equatorial Guinea',
3052 'ER': 'Eritrea',
3053 'EE': 'Estonia',
3054 'ET': 'Ethiopia',
3055 'FK': 'Falkland Islands (Malvinas)',
3056 'FO': 'Faroe Islands',
3057 'FJ': 'Fiji',
3058 'FI': 'Finland',
3059 'FR': 'France',
3060 'GF': 'French Guiana',
3061 'PF': 'French Polynesia',
3062 'TF': 'French Southern Territories',
3063 'GA': 'Gabon',
3064 'GM': 'Gambia',
3065 'GE': 'Georgia',
3066 'DE': 'Germany',
3067 'GH': 'Ghana',
3068 'GI': 'Gibraltar',
3069 'GR': 'Greece',
3070 'GL': 'Greenland',
3071 'GD': 'Grenada',
3072 'GP': 'Guadeloupe',
3073 'GU': 'Guam',
3074 'GT': 'Guatemala',
3075 'GG': 'Guernsey',
3076 'GN': 'Guinea',
3077 'GW': 'Guinea-Bissau',
3078 'GY': 'Guyana',
3079 'HT': 'Haiti',
3080 'HM': 'Heard Island and McDonald Islands',
3081 'VA': 'Holy See (Vatican City State)',
3082 'HN': 'Honduras',
3083 'HK': 'Hong Kong',
3084 'HU': 'Hungary',
3085 'IS': 'Iceland',
3086 'IN': 'India',
3087 'ID': 'Indonesia',
3088 'IR': 'Iran, Islamic Republic of',
3089 'IQ': 'Iraq',
3090 'IE': 'Ireland',
3091 'IM': 'Isle of Man',
3092 'IL': 'Israel',
3093 'IT': 'Italy',
3094 'JM': 'Jamaica',
3095 'JP': 'Japan',
3096 'JE': 'Jersey',
3097 'JO': 'Jordan',
3098 'KZ': 'Kazakhstan',
3099 'KE': 'Kenya',
3100 'KI': 'Kiribati',
3101 'KP': 'Korea, Democratic People\'s Republic of',
3102 'KR': 'Korea, Republic of',
3103 'KW': 'Kuwait',
3104 'KG': 'Kyrgyzstan',
3105 'LA': 'Lao People\'s Democratic Republic',
3106 'LV': 'Latvia',
3107 'LB': 'Lebanon',
3108 'LS': 'Lesotho',
3109 'LR': 'Liberia',
3110 'LY': 'Libya',
3111 'LI': 'Liechtenstein',
3112 'LT': 'Lithuania',
3113 'LU': 'Luxembourg',
3114 'MO': 'Macao',
3115 'MK': 'Macedonia, the Former Yugoslav Republic of',
3116 'MG': 'Madagascar',
3117 'MW': 'Malawi',
3118 'MY': 'Malaysia',
3119 'MV': 'Maldives',
3120 'ML': 'Mali',
3121 'MT': 'Malta',
3122 'MH': 'Marshall Islands',
3123 'MQ': 'Martinique',
3124 'MR': 'Mauritania',
3125 'MU': 'Mauritius',
3126 'YT': 'Mayotte',
3127 'MX': 'Mexico',
3128 'FM': 'Micronesia, Federated States of',
3129 'MD': 'Moldova, Republic of',
3130 'MC': 'Monaco',
3131 'MN': 'Mongolia',
3132 'ME': 'Montenegro',
3133 'MS': 'Montserrat',
3134 'MA': 'Morocco',
3135 'MZ': 'Mozambique',
3136 'MM': 'Myanmar',
3137 'NA': 'Namibia',
3138 'NR': 'Nauru',
3139 'NP': 'Nepal',
3140 'NL': 'Netherlands',
3141 'NC': 'New Caledonia',
3142 'NZ': 'New Zealand',
3143 'NI': 'Nicaragua',
3144 'NE': 'Niger',
3145 'NG': 'Nigeria',
3146 'NU': 'Niue',
3147 'NF': 'Norfolk Island',
3148 'MP': 'Northern Mariana Islands',
3149 'NO': 'Norway',
3150 'OM': 'Oman',
3151 'PK': 'Pakistan',
3152 'PW': 'Palau',
3153 'PS': 'Palestine, State of',
3154 'PA': 'Panama',
3155 'PG': 'Papua New Guinea',
3156 'PY': 'Paraguay',
3157 'PE': 'Peru',
3158 'PH': 'Philippines',
3159 'PN': 'Pitcairn',
3160 'PL': 'Poland',
3161 'PT': 'Portugal',
3162 'PR': 'Puerto Rico',
3163 'QA': 'Qatar',
3164 'RE': 'Réunion',
3165 'RO': 'Romania',
3166 'RU': 'Russian Federation',
3167 'RW': 'Rwanda',
3168 'BL': 'Saint Barthélemy',
3169 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
3170 'KN': 'Saint Kitts and Nevis',
3171 'LC': 'Saint Lucia',
3172 'MF': 'Saint Martin (French part)',
3173 'PM': 'Saint Pierre and Miquelon',
3174 'VC': 'Saint Vincent and the Grenadines',
3175 'WS': 'Samoa',
3176 'SM': 'San Marino',
3177 'ST': 'Sao Tome and Principe',
3178 'SA': 'Saudi Arabia',
3179 'SN': 'Senegal',
3180 'RS': 'Serbia',
3181 'SC': 'Seychelles',
3182 'SL': 'Sierra Leone',
3183 'SG': 'Singapore',
3184 'SX': 'Sint Maarten (Dutch part)',
3185 'SK': 'Slovakia',
3186 'SI': 'Slovenia',
3187 'SB': 'Solomon Islands',
3188 'SO': 'Somalia',
3189 'ZA': 'South Africa',
3190 'GS': 'South Georgia and the South Sandwich Islands',
3191 'SS': 'South Sudan',
3192 'ES': 'Spain',
3193 'LK': 'Sri Lanka',
3194 'SD': 'Sudan',
3195 'SR': 'Suriname',
3196 'SJ': 'Svalbard and Jan Mayen',
3197 'SZ': 'Swaziland',
3198 'SE': 'Sweden',
3199 'CH': 'Switzerland',
3200 'SY': 'Syrian Arab Republic',
3201 'TW': 'Taiwan, Province of China',
3202 'TJ': 'Tajikistan',
3203 'TZ': 'Tanzania, United Republic of',
3204 'TH': 'Thailand',
3205 'TL': 'Timor-Leste',
3206 'TG': 'Togo',
3207 'TK': 'Tokelau',
3208 'TO': 'Tonga',
3209 'TT': 'Trinidad and Tobago',
3210 'TN': 'Tunisia',
3211 'TR': 'Turkey',
3212 'TM': 'Turkmenistan',
3213 'TC': 'Turks and Caicos Islands',
3214 'TV': 'Tuvalu',
3215 'UG': 'Uganda',
3216 'UA': 'Ukraine',
3217 'AE': 'United Arab Emirates',
3218 'GB': 'United Kingdom',
3219 'US': 'United States',
3220 'UM': 'United States Minor Outlying Islands',
3221 'UY': 'Uruguay',
3222 'UZ': 'Uzbekistan',
3223 'VU': 'Vanuatu',
3224 'VE': 'Venezuela, Bolivarian Republic of',
3225 'VN': 'Viet Nam',
3226 'VG': 'Virgin Islands, British',
3227 'VI': 'Virgin Islands, U.S.',
3228 'WF': 'Wallis and Futuna',
3229 'EH': 'Western Sahara',
3230 'YE': 'Yemen',
3231 'ZM': 'Zambia',
3232 'ZW': 'Zimbabwe',
3233 }
3234
3235 @classmethod
3236 def short2full(cls, code):
3237 """Convert an ISO 3166-2 country code to the corresponding full name"""
3238 return cls._country_map.get(code.upper())
3239
3240
3241 class GeoUtils(object):
3242 # Major IPv4 address blocks per country
3243 _country_ip_map = {
3244 'AD': '85.94.160.0/19',
3245 'AE': '94.200.0.0/13',
3246 'AF': '149.54.0.0/17',
3247 'AG': '209.59.64.0/18',
3248 'AI': '204.14.248.0/21',
3249 'AL': '46.99.0.0/16',
3250 'AM': '46.70.0.0/15',
3251 'AO': '105.168.0.0/13',
3252 'AP': '159.117.192.0/21',
3253 'AR': '181.0.0.0/12',
3254 'AS': '202.70.112.0/20',
3255 'AT': '84.112.0.0/13',
3256 'AU': '1.128.0.0/11',
3257 'AW': '181.41.0.0/18',
3258 'AZ': '5.191.0.0/16',
3259 'BA': '31.176.128.0/17',
3260 'BB': '65.48.128.0/17',
3261 'BD': '114.130.0.0/16',
3262 'BE': '57.0.0.0/8',
3263 'BF': '129.45.128.0/17',
3264 'BG': '95.42.0.0/15',
3265 'BH': '37.131.0.0/17',
3266 'BI': '154.117.192.0/18',
3267 'BJ': '137.255.0.0/16',
3268 'BL': '192.131.134.0/24',
3269 'BM': '196.12.64.0/18',
3270 'BN': '156.31.0.0/16',
3271 'BO': '161.56.0.0/16',
3272 'BQ': '161.0.80.0/20',
3273 'BR': '152.240.0.0/12',
3274 'BS': '24.51.64.0/18',
3275 'BT': '119.2.96.0/19',
3276 'BW': '168.167.0.0/16',
3277 'BY': '178.120.0.0/13',
3278 'BZ': '179.42.192.0/18',
3279 'CA': '99.224.0.0/11',
3280 'CD': '41.243.0.0/16',
3281 'CF': '196.32.200.0/21',
3282 'CG': '197.214.128.0/17',
3283 'CH': '85.0.0.0/13',
3284 'CI': '154.232.0.0/14',
3285 'CK': '202.65.32.0/19',
3286 'CL': '152.172.0.0/14',
3287 'CM': '165.210.0.0/15',
3288 'CN': '36.128.0.0/10',
3289 'CO': '181.240.0.0/12',
3290 'CR': '201.192.0.0/12',
3291 'CU': '152.206.0.0/15',
3292 'CV': '165.90.96.0/19',
3293 'CW': '190.88.128.0/17',
3294 'CY': '46.198.0.0/15',
3295 'CZ': '88.100.0.0/14',
3296 'DE': '53.0.0.0/8',
3297 'DJ': '197.241.0.0/17',
3298 'DK': '87.48.0.0/12',
3299 'DM': '192.243.48.0/20',
3300 'DO': '152.166.0.0/15',
3301 'DZ': '41.96.0.0/12',
3302 'EC': '186.68.0.0/15',
3303 'EE': '90.190.0.0/15',
3304 'EG': '156.160.0.0/11',
3305 'ER': '196.200.96.0/20',
3306 'ES': '88.0.0.0/11',
3307 'ET': '196.188.0.0/14',
3308 'EU': '2.16.0.0/13',
3309 'FI': '91.152.0.0/13',
3310 'FJ': '144.120.0.0/16',
3311 'FM': '119.252.112.0/20',
3312 'FO': '88.85.32.0/19',
3313 'FR': '90.0.0.0/9',
3314 'GA': '41.158.0.0/15',
3315 'GB': '25.0.0.0/8',
3316 'GD': '74.122.88.0/21',
3317 'GE': '31.146.0.0/16',
3318 'GF': '161.22.64.0/18',
3319 'GG': '62.68.160.0/19',
3320 'GH': '45.208.0.0/14',
3321 'GI': '85.115.128.0/19',
3322 'GL': '88.83.0.0/19',
3323 'GM': '160.182.0.0/15',
3324 'GN': '197.149.192.0/18',
3325 'GP': '104.250.0.0/19',
3326 'GQ': '105.235.224.0/20',
3327 'GR': '94.64.0.0/13',
3328 'GT': '168.234.0.0/16',
3329 'GU': '168.123.0.0/16',
3330 'GW': '197.214.80.0/20',
3331 'GY': '181.41.64.0/18',
3332 'HK': '113.252.0.0/14',
3333 'HN': '181.210.0.0/16',
3334 'HR': '93.136.0.0/13',
3335 'HT': '148.102.128.0/17',
3336 'HU': '84.0.0.0/14',
3337 'ID': '39.192.0.0/10',
3338 'IE': '87.32.0.0/12',
3339 'IL': '79.176.0.0/13',
3340 'IM': '5.62.80.0/20',
3341 'IN': '117.192.0.0/10',
3342 'IO': '203.83.48.0/21',
3343 'IQ': '37.236.0.0/14',
3344 'IR': '2.176.0.0/12',
3345 'IS': '82.221.0.0/16',
3346 'IT': '79.0.0.0/10',
3347 'JE': '87.244.64.0/18',
3348 'JM': '72.27.0.0/17',
3349 'JO': '176.29.0.0/16',
3350 'JP': '126.0.0.0/8',
3351 'KE': '105.48.0.0/12',
3352 'KG': '158.181.128.0/17',
3353 'KH': '36.37.128.0/17',
3354 'KI': '103.25.140.0/22',
3355 'KM': '197.255.224.0/20',
3356 'KN': '198.32.32.0/19',
3357 'KP': '175.45.176.0/22',
3358 'KR': '175.192.0.0/10',
3359 'KW': '37.36.0.0/14',
3360 'KY': '64.96.0.0/15',
3361 'KZ': '2.72.0.0/13',
3362 'LA': '115.84.64.0/18',
3363 'LB': '178.135.0.0/16',
3364 'LC': '192.147.231.0/24',
3365 'LI': '82.117.0.0/19',
3366 'LK': '112.134.0.0/15',
3367 'LR': '41.86.0.0/19',
3368 'LS': '129.232.0.0/17',
3369 'LT': '78.56.0.0/13',
3370 'LU': '188.42.0.0/16',
3371 'LV': '46.109.0.0/16',
3372 'LY': '41.252.0.0/14',
3373 'MA': '105.128.0.0/11',
3374 'MC': '88.209.64.0/18',
3375 'MD': '37.246.0.0/16',
3376 'ME': '178.175.0.0/17',
3377 'MF': '74.112.232.0/21',
3378 'MG': '154.126.0.0/17',
3379 'MH': '117.103.88.0/21',
3380 'MK': '77.28.0.0/15',
3381 'ML': '154.118.128.0/18',
3382 'MM': '37.111.0.0/17',
3383 'MN': '49.0.128.0/17',
3384 'MO': '60.246.0.0/16',
3385 'MP': '202.88.64.0/20',
3386 'MQ': '109.203.224.0/19',
3387 'MR': '41.188.64.0/18',
3388 'MS': '208.90.112.0/22',
3389 'MT': '46.11.0.0/16',
3390 'MU': '105.16.0.0/12',
3391 'MV': '27.114.128.0/18',
3392 'MW': '105.234.0.0/16',
3393 'MX': '187.192.0.0/11',
3394 'MY': '175.136.0.0/13',
3395 'MZ': '197.218.0.0/15',
3396 'NA': '41.182.0.0/16',
3397 'NC': '101.101.0.0/18',
3398 'NE': '197.214.0.0/18',
3399 'NF': '203.17.240.0/22',
3400 'NG': '105.112.0.0/12',
3401 'NI': '186.76.0.0/15',
3402 'NL': '145.96.0.0/11',
3403 'NO': '84.208.0.0/13',
3404 'NP': '36.252.0.0/15',
3405 'NR': '203.98.224.0/19',
3406 'NU': '49.156.48.0/22',
3407 'NZ': '49.224.0.0/14',
3408 'OM': '5.36.0.0/15',
3409 'PA': '186.72.0.0/15',
3410 'PE': '186.160.0.0/14',
3411 'PF': '123.50.64.0/18',
3412 'PG': '124.240.192.0/19',
3413 'PH': '49.144.0.0/13',
3414 'PK': '39.32.0.0/11',
3415 'PL': '83.0.0.0/11',
3416 'PM': '70.36.0.0/20',
3417 'PR': '66.50.0.0/16',
3418 'PS': '188.161.0.0/16',
3419 'PT': '85.240.0.0/13',
3420 'PW': '202.124.224.0/20',
3421 'PY': '181.120.0.0/14',
3422 'QA': '37.210.0.0/15',
3423 'RE': '139.26.0.0/16',
3424 'RO': '79.112.0.0/13',
3425 'RS': '178.220.0.0/14',
3426 'RU': '5.136.0.0/13',
3427 'RW': '105.178.0.0/15',
3428 'SA': '188.48.0.0/13',
3429 'SB': '202.1.160.0/19',
3430 'SC': '154.192.0.0/11',
3431 'SD': '154.96.0.0/13',
3432 'SE': '78.64.0.0/12',
3433 'SG': '152.56.0.0/14',
3434 'SI': '188.196.0.0/14',
3435 'SK': '78.98.0.0/15',
3436 'SL': '197.215.0.0/17',
3437 'SM': '89.186.32.0/19',
3438 'SN': '41.82.0.0/15',
3439 'SO': '197.220.64.0/19',
3440 'SR': '186.179.128.0/17',
3441 'SS': '105.235.208.0/21',
3442 'ST': '197.159.160.0/19',
3443 'SV': '168.243.0.0/16',
3444 'SX': '190.102.0.0/20',
3445 'SY': '5.0.0.0/16',
3446 'SZ': '41.84.224.0/19',
3447 'TC': '65.255.48.0/20',
3448 'TD': '154.68.128.0/19',
3449 'TG': '196.168.0.0/14',
3450 'TH': '171.96.0.0/13',
3451 'TJ': '85.9.128.0/18',
3452 'TK': '27.96.24.0/21',
3453 'TL': '180.189.160.0/20',
3454 'TM': '95.85.96.0/19',
3455 'TN': '197.0.0.0/11',
3456 'TO': '175.176.144.0/21',
3457 'TR': '78.160.0.0/11',
3458 'TT': '186.44.0.0/15',
3459 'TV': '202.2.96.0/19',
3460 'TW': '120.96.0.0/11',
3461 'TZ': '156.156.0.0/14',
3462 'UA': '93.72.0.0/13',
3463 'UG': '154.224.0.0/13',
3464 'US': '3.0.0.0/8',
3465 'UY': '167.56.0.0/13',
3466 'UZ': '82.215.64.0/18',
3467 'VA': '212.77.0.0/19',
3468 'VC': '24.92.144.0/20',
3469 'VE': '186.88.0.0/13',
3470 'VG': '172.103.64.0/18',
3471 'VI': '146.226.0.0/16',
3472 'VN': '14.160.0.0/11',
3473 'VU': '202.80.32.0/20',
3474 'WF': '117.20.32.0/21',
3475 'WS': '202.4.32.0/19',
3476 'YE': '134.35.0.0/16',
3477 'YT': '41.242.116.0/22',
3478 'ZA': '41.0.0.0/11',
3479 'ZM': '165.56.0.0/13',
3480 'ZW': '41.85.192.0/19',
3481 }
3482
3483 @classmethod
3484 def random_ipv4(cls, code):
3485 block = cls._country_ip_map.get(code.upper())
3486 if not block:
3487 return None
3488 addr, preflen = block.split('/')
3489 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
3490 addr_max = addr_min | (0xffffffff >> int(preflen))
3491 return compat_str(socket.inet_ntoa(
3492 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
3493
3494
3495 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
3496 def __init__(self, proxies=None):
3497 # Set default handlers
3498 for type in ('http', 'https'):
3499 setattr(self, '%s_open' % type,
3500 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
3501 meth(r, proxy, type))
3502 return compat_urllib_request.ProxyHandler.__init__(self, proxies)
3503
3504 def proxy_open(self, req, proxy, type):
3505 req_proxy = req.headers.get('Ytdl-request-proxy')
3506 if req_proxy is not None:
3507 proxy = req_proxy
3508 del req.headers['Ytdl-request-proxy']
3509
3510 if proxy == '__noproxy__':
3511 return None # No Proxy
3512 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
3513 req.add_header('Ytdl-socks-proxy', proxy)
3514 # youtube-dl's http/https handlers do wrapping the socket with socks
3515 return None
3516 return compat_urllib_request.ProxyHandler.proxy_open(
3517 self, req, proxy, type)
3518
3519
3520 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
3521 # released into Public Domain
3522 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
3523
3524 def long_to_bytes(n, blocksize=0):
3525 """long_to_bytes(n:long, blocksize:int) : string
3526 Convert a long integer to a byte string.
3527
3528 If optional blocksize is given and greater than zero, pad the front of the
3529 byte string with binary zeros so that the length is a multiple of
3530 blocksize.
3531 """
3532 # after much testing, this algorithm was deemed to be the fastest
3533 s = b''
3534 n = int(n)
3535 while n > 0:
3536 s = compat_struct_pack('>I', n & 0xffffffff) + s
3537 n = n >> 32
3538 # strip off leading zeros
3539 for i in range(len(s)):
3540 if s[i] != b'\000'[0]:
3541 break
3542 else:
3543 # only happens when n == 0
3544 s = b'\000'
3545 i = 0
3546 s = s[i:]
3547 # add back some pad bytes. this could be done more efficiently w.r.t. the
3548 # de-padding being done above, but sigh...
3549 if blocksize > 0 and len(s) % blocksize:
3550 s = (blocksize - len(s) % blocksize) * b'\000' + s
3551 return s
3552
3553
3554 def bytes_to_long(s):
3555 """bytes_to_long(string) : long
3556 Convert a byte string to a long integer.
3557
3558 This is (essentially) the inverse of long_to_bytes().
3559 """
3560 acc = 0
3561 length = len(s)
3562 if length % 4:
3563 extra = (4 - length % 4)
3564 s = b'\000' * extra + s
3565 length = length + extra
3566 for i in range(0, length, 4):
3567 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
3568 return acc
3569
3570
3571 def ohdave_rsa_encrypt(data, exponent, modulus):
3572 '''
3573 Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
3574
3575 Input:
3576 data: data to encrypt, bytes-like object
3577 exponent, modulus: parameter e and N of RSA algorithm, both integer
3578 Output: hex string of encrypted data
3579
3580 Limitation: supports one block encryption only
3581 '''
3582
3583 payload = int(binascii.hexlify(data[::-1]), 16)
3584 encrypted = pow(payload, exponent, modulus)
3585 return '%x' % encrypted
3586
3587
3588 def pkcs1pad(data, length):
3589 """
3590 Padding input data with PKCS#1 scheme
3591
3592 @param {int[]} data input data
3593 @param {int} length target length
3594 @returns {int[]} padded data
3595 """
3596 if len(data) > length - 11:
3597 raise ValueError('Input data too long for PKCS#1 padding')
3598
3599 pseudo_random = [random.randint(0, 254) for _ in range(length - len(data) - 3)]
3600 return [0, 2] + pseudo_random + [0] + data
3601
3602
3603 def encode_base_n(num, n, table=None):
3604 FULL_TABLE = '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
3605 if not table:
3606 table = FULL_TABLE[:n]
3607
3608 if n > len(table):
3609 raise ValueError('base %d exceeds table length %d' % (n, len(table)))
3610
3611 if num == 0:
3612 return table[0]
3613
3614 ret = ''
3615 while num:
3616 ret = table[num % n] + ret
3617 num = num // n
3618 return ret
3619
3620
3621 def decode_packed_codes(code):
3622 mobj = re.search(PACKED_CODES_RE, code)
3623 obfucasted_code, base, count, symbols = mobj.groups()
3624 base = int(base)
3625 count = int(count)
3626 symbols = symbols.split('|')
3627 symbol_table = {}
3628
3629 while count:
3630 count -= 1
3631 base_n_count = encode_base_n(count, base)
3632 symbol_table[base_n_count] = symbols[count] or base_n_count
3633
3634 return re.sub(
3635 r'\b(\w+)\b', lambda mobj: symbol_table[mobj.group(0)],
3636 obfucasted_code)
3637
3638
3639 def parse_m3u8_attributes(attrib):
3640 info = {}
3641 for (key, val) in re.findall(r'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib):
3642 if val.startswith('"'):
3643 val = val[1:-1]
3644 info[key] = val
3645 return info
3646
3647
3648 def urshift(val, n):
3649 return val >> n if val >= 0 else (val + 0x100000000) >> n
3650
3651
3652 # Based on png2str() written by @gdkchan and improved by @yokrysty
3653 # Originally posted at https://github.com/rg3/youtube-dl/issues/9706
3654 def decode_png(png_data):
3655 # Reference: https://www.w3.org/TR/PNG/
3656 header = png_data[8:]
3657
3658 if png_data[:8] != b'\x89PNG\x0d\x0a\x1a\x0a' or header[4:8] != b'IHDR':
3659 raise IOError('Not a valid PNG file.')
3660
3661 int_map = {1: '>B', 2: '>H', 4: '>I'}
3662 unpack_integer = lambda x: compat_struct_unpack(int_map[len(x)], x)[0]
3663
3664 chunks = []
3665
3666 while header:
3667 length = unpack_integer(header[:4])
3668 header = header[4:]
3669
3670 chunk_type = header[:4]
3671 header = header[4:]
3672
3673 chunk_data = header[:length]
3674 header = header[length:]
3675
3676 header = header[4:] # Skip CRC
3677
3678 chunks.append({
3679 'type': chunk_type,
3680 'length': length,
3681 'data': chunk_data
3682 })
3683
3684 ihdr = chunks[0]['data']
3685
3686 width = unpack_integer(ihdr[:4])
3687 height = unpack_integer(ihdr[4:8])
3688
3689 idat = b''
3690
3691 for chunk in chunks:
3692 if chunk['type'] == b'IDAT':
3693 idat += chunk['data']
3694
3695 if not idat:
3696 raise IOError('Unable to read PNG data.')
3697
3698 decompressed_data = bytearray(zlib.decompress(idat))
3699
3700 stride = width * 3
3701 pixels = []
3702
3703 def _get_pixel(idx):
3704 x = idx % stride
3705 y = idx // stride
3706 return pixels[y][x]
3707
3708 for y in range(height):
3709 basePos = y * (1 + stride)
3710 filter_type = decompressed_data[basePos]
3711
3712 current_row = []
3713
3714 pixels.append(current_row)
3715
3716 for x in range(stride):
3717 color = decompressed_data[1 + basePos + x]
3718 basex = y * stride + x
3719 left = 0
3720 up = 0
3721
3722 if x > 2:
3723 left = _get_pixel(basex - 3)
3724 if y > 0:
3725 up = _get_pixel(basex - stride)
3726
3727 if filter_type == 1: # Sub
3728 color = (color + left) & 0xff
3729 elif filter_type == 2: # Up
3730 color = (color + up) & 0xff
3731 elif filter_type == 3: # Average
3732 color = (color + ((left + up) >> 1)) & 0xff
3733 elif filter_type == 4: # Paeth
3734 a = left
3735 b = up
3736 c = 0
3737
3738 if x > 2 and y > 0:
3739 c = _get_pixel(basex - stride - 3)
3740
3741 p = a + b - c
3742
3743 pa = abs(p - a)
3744 pb = abs(p - b)
3745 pc = abs(p - c)
3746
3747 if pa <= pb and pa <= pc:
3748 color = (color + a) & 0xff
3749 elif pb <= pc:
3750 color = (color + b) & 0xff
3751 else:
3752 color = (color + c) & 0xff
3753
3754 current_row.append(color)
3755
3756 return width, height, pixels
3757
3758
3759 def write_xattr(path, key, value):
3760 # This mess below finds the best xattr tool for the job
3761 try:
3762 # try the pyxattr module...
3763 import xattr
3764
3765 if hasattr(xattr, 'set'): # pyxattr
3766 # Unicode arguments are not supported in python-pyxattr until
3767 # version 0.5.0
3768 # See https://github.com/rg3/youtube-dl/issues/5498
3769 pyxattr_required_version = '0.5.0'
3770 if version_tuple(xattr.__version__) < version_tuple(pyxattr_required_version):
3771 # TODO: fallback to CLI tools
3772 raise XAttrUnavailableError(
3773 'python-pyxattr is detected but is too old. '
3774 'youtube-dl requires %s or above while your version is %s. '
3775 'Falling back to other xattr implementations' % (
3776 pyxattr_required_version, xattr.__version__))
3777
3778 setxattr = xattr.set
3779 else: # xattr
3780 setxattr = xattr.setxattr
3781
3782 try:
3783 setxattr(path, key, value)
3784 except EnvironmentError as e:
3785 raise XAttrMetadataError(e.errno, e.strerror)
3786
3787 except ImportError:
3788 if compat_os_name == 'nt':
3789 # Write xattrs to NTFS Alternate Data Streams:
3790 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
3791 assert ':' not in key
3792 assert os.path.exists(path)
3793
3794 ads_fn = path + ':' + key
3795 try:
3796 with open(ads_fn, 'wb') as f:
3797 f.write(value)
3798 except EnvironmentError as e:
3799 raise XAttrMetadataError(e.errno, e.strerror)
3800 else:
3801 user_has_setfattr = check_executable('setfattr', ['--version'])
3802 user_has_xattr = check_executable('xattr', ['-h'])
3803
3804 if user_has_setfattr or user_has_xattr:
3805
3806 value = value.decode('utf-8')
3807 if user_has_setfattr:
3808 executable = 'setfattr'
3809 opts = ['-n', key, '-v', value]
3810 elif user_has_xattr:
3811 executable = 'xattr'
3812 opts = ['-w', key, value]
3813
3814 cmd = ([encodeFilename(executable, True)] +
3815 [encodeArgument(o) for o in opts] +
3816 [encodeFilename(path, True)])
3817
3818 try:
3819 p = subprocess.Popen(
3820 cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, stdin=subprocess.PIPE)
3821 except EnvironmentError as e:
3822 raise XAttrMetadataError(e.errno, e.strerror)
3823 stdout, stderr = p.communicate()
3824 stderr = stderr.decode('utf-8', 'replace')
3825 if p.returncode != 0:
3826 raise XAttrMetadataError(p.returncode, stderr)
3827
3828 else:
3829 # On Unix, and can't find pyxattr, setfattr, or xattr.
3830 if sys.platform.startswith('linux'):
3831 raise XAttrUnavailableError(
3832 "Couldn't find a tool to set the xattrs. "
3833 "Install either the python 'pyxattr' or 'xattr' "
3834 "modules, or the GNU 'attr' package "
3835 "(which contains the 'setfattr' tool).")
3836 else:
3837 raise XAttrUnavailableError(
3838 "Couldn't find a tool to set the xattrs. "
3839 "Install either the python 'xattr' module, "
3840 "or the 'xattr' binary.")
3841
3842
3843 def random_birthday(year_field, month_field, day_field):
3844 return {
3845 year_field: str(random.randint(1950, 1995)),
3846 month_field: str(random.randint(1, 12)),
3847 day_field: str(random.randint(1, 31)),
3848 }