]>
Commit | Line | Data |
---|---|---|
1 | #!/usr/bin/env python | |
2 | # -*- coding: utf-8 -*- | |
3 | ||
4 | from __future__ import unicode_literals | |
5 | ||
6 | import calendar | |
7 | import codecs | |
8 | import contextlib | |
9 | import ctypes | |
10 | import datetime | |
11 | import email.utils | |
12 | import errno | |
13 | import functools | |
14 | import gzip | |
15 | import itertools | |
16 | import io | |
17 | import json | |
18 | import locale | |
19 | import math | |
20 | import os | |
21 | import pipes | |
22 | import platform | |
23 | import re | |
24 | import ssl | |
25 | import socket | |
26 | import struct | |
27 | import subprocess | |
28 | import sys | |
29 | import tempfile | |
30 | import traceback | |
31 | import xml.etree.ElementTree | |
32 | import zlib | |
33 | ||
34 | from .compat import ( | |
35 | compat_chr, | |
36 | compat_getenv, | |
37 | compat_html_entities, | |
38 | compat_http_client, | |
39 | compat_parse_qs, | |
40 | compat_socket_create_connection, | |
41 | compat_str, | |
42 | compat_urllib_error, | |
43 | compat_urllib_parse, | |
44 | compat_urllib_parse_urlparse, | |
45 | compat_urllib_request, | |
46 | compat_urlparse, | |
47 | shlex_quote, | |
48 | ) | |
49 | ||
50 | ||
51 | # This is not clearly defined otherwise | |
52 | compiled_regex_type = type(re.compile('')) | |
53 | ||
54 | std_headers = { | |
55 | 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20100101 Firefox/10.0 (Chrome)', | |
56 | 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7', | |
57 | 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8', | |
58 | 'Accept-Encoding': 'gzip, deflate', | |
59 | 'Accept-Language': 'en-us,en;q=0.5', | |
60 | } | |
61 | ||
62 | ||
63 | def preferredencoding(): | |
64 | """Get preferred encoding. | |
65 | ||
66 | Returns the best encoding scheme for the system, based on | |
67 | locale.getpreferredencoding() and some further tweaks. | |
68 | """ | |
69 | try: | |
70 | pref = locale.getpreferredencoding() | |
71 | 'TEST'.encode(pref) | |
72 | except: | |
73 | pref = 'UTF-8' | |
74 | ||
75 | return pref | |
76 | ||
77 | ||
78 | def write_json_file(obj, fn): | |
79 | """ Encode obj as JSON and write it to fn, atomically if possible """ | |
80 | ||
81 | fn = encodeFilename(fn) | |
82 | if sys.version_info < (3, 0) and sys.platform != 'win32': | |
83 | encoding = get_filesystem_encoding() | |
84 | # os.path.basename returns a bytes object, but NamedTemporaryFile | |
85 | # will fail if the filename contains non ascii characters unless we | |
86 | # use a unicode object | |
87 | path_basename = lambda f: os.path.basename(fn).decode(encoding) | |
88 | # the same for os.path.dirname | |
89 | path_dirname = lambda f: os.path.dirname(fn).decode(encoding) | |
90 | else: | |
91 | path_basename = os.path.basename | |
92 | path_dirname = os.path.dirname | |
93 | ||
94 | args = { | |
95 | 'suffix': '.tmp', | |
96 | 'prefix': path_basename(fn) + '.', | |
97 | 'dir': path_dirname(fn), | |
98 | 'delete': False, | |
99 | } | |
100 | ||
101 | # In Python 2.x, json.dump expects a bytestream. | |
102 | # In Python 3.x, it writes to a character stream | |
103 | if sys.version_info < (3, 0): | |
104 | args['mode'] = 'wb' | |
105 | else: | |
106 | args.update({ | |
107 | 'mode': 'w', | |
108 | 'encoding': 'utf-8', | |
109 | }) | |
110 | ||
111 | tf = tempfile.NamedTemporaryFile(**args) | |
112 | ||
113 | try: | |
114 | with tf: | |
115 | json.dump(obj, tf) | |
116 | if sys.platform == 'win32': | |
117 | # Need to remove existing file on Windows, else os.rename raises | |
118 | # WindowsError or FileExistsError. | |
119 | try: | |
120 | os.unlink(fn) | |
121 | except OSError: | |
122 | pass | |
123 | os.rename(tf.name, fn) | |
124 | except: | |
125 | try: | |
126 | os.remove(tf.name) | |
127 | except OSError: | |
128 | pass | |
129 | raise | |
130 | ||
131 | ||
132 | if sys.version_info >= (2, 7): | |
133 | def find_xpath_attr(node, xpath, key, val): | |
134 | """ Find the xpath xpath[@key=val] """ | |
135 | assert re.match(r'^[a-zA-Z-]+$', key) | |
136 | assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val) | |
137 | expr = xpath + "[@%s='%s']" % (key, val) | |
138 | return node.find(expr) | |
139 | else: | |
140 | def find_xpath_attr(node, xpath, key, val): | |
141 | # Here comes the crazy part: In 2.6, if the xpath is a unicode, | |
142 | # .//node does not match if a node is a direct child of . ! | |
143 | if isinstance(xpath, unicode): | |
144 | xpath = xpath.encode('ascii') | |
145 | ||
146 | for f in node.findall(xpath): | |
147 | if f.attrib.get(key) == val: | |
148 | return f | |
149 | return None | |
150 | ||
151 | # On python2.6 the xml.etree.ElementTree.Element methods don't support | |
152 | # the namespace parameter | |
153 | ||
154 | ||
155 | def xpath_with_ns(path, ns_map): | |
156 | components = [c.split(':') for c in path.split('/')] | |
157 | replaced = [] | |
158 | for c in components: | |
159 | if len(c) == 1: | |
160 | replaced.append(c[0]) | |
161 | else: | |
162 | ns, tag = c | |
163 | replaced.append('{%s}%s' % (ns_map[ns], tag)) | |
164 | return '/'.join(replaced) | |
165 | ||
166 | ||
167 | def xpath_text(node, xpath, name=None, fatal=False): | |
168 | if sys.version_info < (2, 7): # Crazy 2.6 | |
169 | xpath = xpath.encode('ascii') | |
170 | ||
171 | n = node.find(xpath) | |
172 | if n is None or n.text is None: | |
173 | if fatal: | |
174 | name = xpath if name is None else name | |
175 | raise ExtractorError('Could not find XML element %s' % name) | |
176 | else: | |
177 | return None | |
178 | return n.text | |
179 | ||
180 | ||
181 | def get_element_by_id(id, html): | |
182 | """Return the content of the tag with the specified ID in the passed HTML document""" | |
183 | return get_element_by_attribute("id", id, html) | |
184 | ||
185 | ||
186 | def get_element_by_attribute(attribute, value, html): | |
187 | """Return the content of the tag with the specified attribute in the passed HTML document""" | |
188 | ||
189 | m = re.search(r'''(?xs) | |
190 | <([a-zA-Z0-9:._-]+) | |
191 | (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*? | |
192 | \s+%s=['"]?%s['"]? | |
193 | (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*? | |
194 | \s*> | |
195 | (?P<content>.*?) | |
196 | </\1> | |
197 | ''' % (re.escape(attribute), re.escape(value)), html) | |
198 | ||
199 | if not m: | |
200 | return None | |
201 | res = m.group('content') | |
202 | ||
203 | if res.startswith('"') or res.startswith("'"): | |
204 | res = res[1:-1] | |
205 | ||
206 | return unescapeHTML(res) | |
207 | ||
208 | ||
209 | def clean_html(html): | |
210 | """Clean an HTML snippet into a readable string""" | |
211 | ||
212 | if html is None: # Convenience for sanitizing descriptions etc. | |
213 | return html | |
214 | ||
215 | # Newline vs <br /> | |
216 | html = html.replace('\n', ' ') | |
217 | html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html) | |
218 | html = re.sub(r'<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html) | |
219 | # Strip html tags | |
220 | html = re.sub('<.*?>', '', html) | |
221 | # Replace html entities | |
222 | html = unescapeHTML(html) | |
223 | return html.strip() | |
224 | ||
225 | ||
226 | def sanitize_open(filename, open_mode): | |
227 | """Try to open the given filename, and slightly tweak it if this fails. | |
228 | ||
229 | Attempts to open the given filename. If this fails, it tries to change | |
230 | the filename slightly, step by step, until it's either able to open it | |
231 | or it fails and raises a final exception, like the standard open() | |
232 | function. | |
233 | ||
234 | It returns the tuple (stream, definitive_file_name). | |
235 | """ | |
236 | try: | |
237 | if filename == '-': | |
238 | if sys.platform == 'win32': | |
239 | import msvcrt | |
240 | msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY) | |
241 | return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename) | |
242 | stream = open(encodeFilename(filename), open_mode) | |
243 | return (stream, filename) | |
244 | except (IOError, OSError) as err: | |
245 | if err.errno in (errno.EACCES,): | |
246 | raise | |
247 | ||
248 | # In case of error, try to remove win32 forbidden chars | |
249 | alt_filename = os.path.join( | |
250 | re.sub('[/<>:"\\|\\\\?\\*]', '#', path_part) | |
251 | for path_part in os.path.split(filename) | |
252 | ) | |
253 | if alt_filename == filename: | |
254 | raise | |
255 | else: | |
256 | # An exception here should be caught in the caller | |
257 | stream = open(encodeFilename(filename), open_mode) | |
258 | return (stream, alt_filename) | |
259 | ||
260 | ||
261 | def timeconvert(timestr): | |
262 | """Convert RFC 2822 defined time string into system timestamp""" | |
263 | timestamp = None | |
264 | timetuple = email.utils.parsedate_tz(timestr) | |
265 | if timetuple is not None: | |
266 | timestamp = email.utils.mktime_tz(timetuple) | |
267 | return timestamp | |
268 | ||
269 | ||
270 | def sanitize_filename(s, restricted=False, is_id=False): | |
271 | """Sanitizes a string so it could be used as part of a filename. | |
272 | If restricted is set, use a stricter subset of allowed characters. | |
273 | Set is_id if this is not an arbitrary string, but an ID that should be kept if possible | |
274 | """ | |
275 | def replace_insane(char): | |
276 | if char == '?' or ord(char) < 32 or ord(char) == 127: | |
277 | return '' | |
278 | elif char == '"': | |
279 | return '' if restricted else '\'' | |
280 | elif char == ':': | |
281 | return '_-' if restricted else ' -' | |
282 | elif char in '\\/|*<>': | |
283 | return '_' | |
284 | if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()): | |
285 | return '_' | |
286 | if restricted and ord(char) > 127: | |
287 | return '_' | |
288 | return char | |
289 | ||
290 | # Handle timestamps | |
291 | s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s) | |
292 | result = ''.join(map(replace_insane, s)) | |
293 | if not is_id: | |
294 | while '__' in result: | |
295 | result = result.replace('__', '_') | |
296 | result = result.strip('_') | |
297 | # Common case of "Foreign band name - English song title" | |
298 | if restricted and result.startswith('-_'): | |
299 | result = result[2:] | |
300 | if not result: | |
301 | result = '_' | |
302 | return result | |
303 | ||
304 | ||
305 | def orderedSet(iterable): | |
306 | """ Remove all duplicates from the input iterable """ | |
307 | res = [] | |
308 | for el in iterable: | |
309 | if el not in res: | |
310 | res.append(el) | |
311 | return res | |
312 | ||
313 | ||
314 | def _htmlentity_transform(entity): | |
315 | """Transforms an HTML entity to a character.""" | |
316 | # Known non-numeric HTML entity | |
317 | if entity in compat_html_entities.name2codepoint: | |
318 | return compat_chr(compat_html_entities.name2codepoint[entity]) | |
319 | ||
320 | mobj = re.match(r'#(x?[0-9]+)', entity) | |
321 | if mobj is not None: | |
322 | numstr = mobj.group(1) | |
323 | if numstr.startswith('x'): | |
324 | base = 16 | |
325 | numstr = '0%s' % numstr | |
326 | else: | |
327 | base = 10 | |
328 | return compat_chr(int(numstr, base)) | |
329 | ||
330 | # Unknown entity in name, return its literal representation | |
331 | return ('&%s;' % entity) | |
332 | ||
333 | ||
334 | def unescapeHTML(s): | |
335 | if s is None: | |
336 | return None | |
337 | assert type(s) == compat_str | |
338 | ||
339 | return re.sub( | |
340 | r'&([^;]+);', lambda m: _htmlentity_transform(m.group(1)), s) | |
341 | ||
342 | ||
343 | def encodeFilename(s, for_subprocess=False): | |
344 | """ | |
345 | @param s The name of the file | |
346 | """ | |
347 | ||
348 | assert type(s) == compat_str | |
349 | ||
350 | # Python 3 has a Unicode API | |
351 | if sys.version_info >= (3, 0): | |
352 | return s | |
353 | ||
354 | if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5: | |
355 | # Pass '' directly to use Unicode APIs on Windows 2000 and up | |
356 | # (Detecting Windows NT 4 is tricky because 'major >= 4' would | |
357 | # match Windows 9x series as well. Besides, NT 4 is obsolete.) | |
358 | if not for_subprocess: | |
359 | return s | |
360 | else: | |
361 | # For subprocess calls, encode with locale encoding | |
362 | # Refer to http://stackoverflow.com/a/9951851/35070 | |
363 | encoding = preferredencoding() | |
364 | else: | |
365 | encoding = sys.getfilesystemencoding() | |
366 | if encoding is None: | |
367 | encoding = 'utf-8' | |
368 | return s.encode(encoding, 'ignore') | |
369 | ||
370 | ||
371 | def encodeArgument(s): | |
372 | if not isinstance(s, compat_str): | |
373 | # Legacy code that uses byte strings | |
374 | # Uncomment the following line after fixing all post processors | |
375 | # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s)) | |
376 | s = s.decode('ascii') | |
377 | return encodeFilename(s, True) | |
378 | ||
379 | ||
380 | def decodeOption(optval): | |
381 | if optval is None: | |
382 | return optval | |
383 | if isinstance(optval, bytes): | |
384 | optval = optval.decode(preferredencoding()) | |
385 | ||
386 | assert isinstance(optval, compat_str) | |
387 | return optval | |
388 | ||
389 | ||
390 | def formatSeconds(secs): | |
391 | if secs > 3600: | |
392 | return '%d:%02d:%02d' % (secs // 3600, (secs % 3600) // 60, secs % 60) | |
393 | elif secs > 60: | |
394 | return '%d:%02d' % (secs // 60, secs % 60) | |
395 | else: | |
396 | return '%d' % secs | |
397 | ||
398 | ||
399 | def make_HTTPS_handler(params, **kwargs): | |
400 | opts_no_check_certificate = params.get('nocheckcertificate', False) | |
401 | if hasattr(ssl, 'create_default_context'): # Python >= 3.4 or 2.7.9 | |
402 | context = ssl.create_default_context(ssl.Purpose.SERVER_AUTH) | |
403 | if opts_no_check_certificate: | |
404 | context.check_hostname = False | |
405 | context.verify_mode = ssl.CERT_NONE | |
406 | try: | |
407 | return YoutubeDLHTTPSHandler(params, context=context, **kwargs) | |
408 | except TypeError: | |
409 | # Python 2.7.8 | |
410 | # (create_default_context present but HTTPSHandler has no context=) | |
411 | pass | |
412 | ||
413 | if sys.version_info < (3, 2): | |
414 | return YoutubeDLHTTPSHandler(params, **kwargs) | |
415 | else: # Python < 3.4 | |
416 | context = ssl.SSLContext(ssl.PROTOCOL_TLSv1) | |
417 | context.verify_mode = (ssl.CERT_NONE | |
418 | if opts_no_check_certificate | |
419 | else ssl.CERT_REQUIRED) | |
420 | context.set_default_verify_paths() | |
421 | return YoutubeDLHTTPSHandler(params, context=context, **kwargs) | |
422 | ||
423 | ||
424 | class ExtractorError(Exception): | |
425 | """Error during info extraction.""" | |
426 | ||
427 | def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None): | |
428 | """ tb, if given, is the original traceback (so that it can be printed out). | |
429 | If expected is set, this is a normal error message and most likely not a bug in youtube-dl. | |
430 | """ | |
431 | ||
432 | if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError): | |
433 | expected = True | |
434 | if video_id is not None: | |
435 | msg = video_id + ': ' + msg | |
436 | if cause: | |
437 | msg += ' (caused by %r)' % cause | |
438 | if not expected: | |
439 | if ytdl_is_updateable(): | |
440 | update_cmd = 'type youtube-dl -U to update' | |
441 | else: | |
442 | update_cmd = 'see https://yt-dl.org/update on how to update' | |
443 | msg += '; please report this issue on https://yt-dl.org/bug .' | |
444 | msg += ' Make sure you are using the latest version; %s.' % update_cmd | |
445 | msg += ' Be sure to call youtube-dl with the --verbose flag and include its complete output.' | |
446 | super(ExtractorError, self).__init__(msg) | |
447 | ||
448 | self.traceback = tb | |
449 | self.exc_info = sys.exc_info() # preserve original exception | |
450 | self.cause = cause | |
451 | self.video_id = video_id | |
452 | ||
453 | def format_traceback(self): | |
454 | if self.traceback is None: | |
455 | return None | |
456 | return ''.join(traceback.format_tb(self.traceback)) | |
457 | ||
458 | ||
459 | class UnsupportedError(ExtractorError): | |
460 | def __init__(self, url): | |
461 | super(UnsupportedError, self).__init__( | |
462 | 'Unsupported URL: %s' % url, expected=True) | |
463 | self.url = url | |
464 | ||
465 | ||
466 | class RegexNotFoundError(ExtractorError): | |
467 | """Error when a regex didn't match""" | |
468 | pass | |
469 | ||
470 | ||
471 | class DownloadError(Exception): | |
472 | """Download Error exception. | |
473 | ||
474 | This exception may be thrown by FileDownloader objects if they are not | |
475 | configured to continue on errors. They will contain the appropriate | |
476 | error message. | |
477 | """ | |
478 | ||
479 | def __init__(self, msg, exc_info=None): | |
480 | """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """ | |
481 | super(DownloadError, self).__init__(msg) | |
482 | self.exc_info = exc_info | |
483 | ||
484 | ||
485 | class SameFileError(Exception): | |
486 | """Same File exception. | |
487 | ||
488 | This exception will be thrown by FileDownloader objects if they detect | |
489 | multiple files would have to be downloaded to the same file on disk. | |
490 | """ | |
491 | pass | |
492 | ||
493 | ||
494 | class PostProcessingError(Exception): | |
495 | """Post Processing exception. | |
496 | ||
497 | This exception may be raised by PostProcessor's .run() method to | |
498 | indicate an error in the postprocessing task. | |
499 | """ | |
500 | ||
501 | def __init__(self, msg): | |
502 | self.msg = msg | |
503 | ||
504 | ||
505 | class MaxDownloadsReached(Exception): | |
506 | """ --max-downloads limit has been reached. """ | |
507 | pass | |
508 | ||
509 | ||
510 | class UnavailableVideoError(Exception): | |
511 | """Unavailable Format exception. | |
512 | ||
513 | This exception will be thrown when a video is requested | |
514 | in a format that is not available for that video. | |
515 | """ | |
516 | pass | |
517 | ||
518 | ||
519 | class ContentTooShortError(Exception): | |
520 | """Content Too Short exception. | |
521 | ||
522 | This exception may be raised by FileDownloader objects when a file they | |
523 | download is too small for what the server announced first, indicating | |
524 | the connection was probably interrupted. | |
525 | """ | |
526 | # Both in bytes | |
527 | downloaded = None | |
528 | expected = None | |
529 | ||
530 | def __init__(self, downloaded, expected): | |
531 | self.downloaded = downloaded | |
532 | self.expected = expected | |
533 | ||
534 | ||
535 | def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs): | |
536 | hc = http_class(*args, **kwargs) | |
537 | source_address = ydl_handler._params.get('source_address') | |
538 | if source_address is not None: | |
539 | sa = (source_address, 0) | |
540 | if hasattr(hc, 'source_address'): # Python 2.7+ | |
541 | hc.source_address = sa | |
542 | else: # Python 2.6 | |
543 | def _hc_connect(self, *args, **kwargs): | |
544 | sock = compat_socket_create_connection( | |
545 | (self.host, self.port), self.timeout, sa) | |
546 | if is_https: | |
547 | self.sock = ssl.wrap_socket( | |
548 | sock, self.key_file, self.cert_file, | |
549 | ssl_version=ssl.PROTOCOL_TLSv1) | |
550 | else: | |
551 | self.sock = sock | |
552 | hc.connect = functools.partial(_hc_connect, hc) | |
553 | ||
554 | return hc | |
555 | ||
556 | ||
557 | class YoutubeDLHandler(compat_urllib_request.HTTPHandler): | |
558 | """Handler for HTTP requests and responses. | |
559 | ||
560 | This class, when installed with an OpenerDirector, automatically adds | |
561 | the standard headers to every HTTP request and handles gzipped and | |
562 | deflated responses from web servers. If compression is to be avoided in | |
563 | a particular request, the original request in the program code only has | |
564 | to include the HTTP header "Youtubedl-No-Compression", which will be | |
565 | removed before making the real request. | |
566 | ||
567 | Part of this code was copied from: | |
568 | ||
569 | http://techknack.net/python-urllib2-handlers/ | |
570 | ||
571 | Andrew Rowls, the author of that code, agreed to release it to the | |
572 | public domain. | |
573 | """ | |
574 | ||
575 | def __init__(self, params, *args, **kwargs): | |
576 | compat_urllib_request.HTTPHandler.__init__(self, *args, **kwargs) | |
577 | self._params = params | |
578 | ||
579 | def http_open(self, req): | |
580 | return self.do_open(functools.partial( | |
581 | _create_http_connection, self, compat_http_client.HTTPConnection, False), | |
582 | req) | |
583 | ||
584 | @staticmethod | |
585 | def deflate(data): | |
586 | try: | |
587 | return zlib.decompress(data, -zlib.MAX_WBITS) | |
588 | except zlib.error: | |
589 | return zlib.decompress(data) | |
590 | ||
591 | @staticmethod | |
592 | def addinfourl_wrapper(stream, headers, url, code): | |
593 | if hasattr(compat_urllib_request.addinfourl, 'getcode'): | |
594 | return compat_urllib_request.addinfourl(stream, headers, url, code) | |
595 | ret = compat_urllib_request.addinfourl(stream, headers, url) | |
596 | ret.code = code | |
597 | return ret | |
598 | ||
599 | def http_request(self, req): | |
600 | for h, v in std_headers.items(): | |
601 | # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275 | |
602 | # The dict keys are capitalized because of this bug by urllib | |
603 | if h.capitalize() not in req.headers: | |
604 | req.add_header(h, v) | |
605 | if 'Youtubedl-no-compression' in req.headers: | |
606 | if 'Accept-encoding' in req.headers: | |
607 | del req.headers['Accept-encoding'] | |
608 | del req.headers['Youtubedl-no-compression'] | |
609 | if 'Youtubedl-user-agent' in req.headers: | |
610 | if 'User-agent' in req.headers: | |
611 | del req.headers['User-agent'] | |
612 | req.headers['User-agent'] = req.headers['Youtubedl-user-agent'] | |
613 | del req.headers['Youtubedl-user-agent'] | |
614 | ||
615 | if sys.version_info < (2, 7) and '#' in req.get_full_url(): | |
616 | # Python 2.6 is brain-dead when it comes to fragments | |
617 | req._Request__original = req._Request__original.partition('#')[0] | |
618 | req._Request__r_type = req._Request__r_type.partition('#')[0] | |
619 | ||
620 | return req | |
621 | ||
622 | def http_response(self, req, resp): | |
623 | old_resp = resp | |
624 | # gzip | |
625 | if resp.headers.get('Content-encoding', '') == 'gzip': | |
626 | content = resp.read() | |
627 | gz = gzip.GzipFile(fileobj=io.BytesIO(content), mode='rb') | |
628 | try: | |
629 | uncompressed = io.BytesIO(gz.read()) | |
630 | except IOError as original_ioerror: | |
631 | # There may be junk add the end of the file | |
632 | # See http://stackoverflow.com/q/4928560/35070 for details | |
633 | for i in range(1, 1024): | |
634 | try: | |
635 | gz = gzip.GzipFile(fileobj=io.BytesIO(content[:-i]), mode='rb') | |
636 | uncompressed = io.BytesIO(gz.read()) | |
637 | except IOError: | |
638 | continue | |
639 | break | |
640 | else: | |
641 | raise original_ioerror | |
642 | resp = self.addinfourl_wrapper(uncompressed, old_resp.headers, old_resp.url, old_resp.code) | |
643 | resp.msg = old_resp.msg | |
644 | # deflate | |
645 | if resp.headers.get('Content-encoding', '') == 'deflate': | |
646 | gz = io.BytesIO(self.deflate(resp.read())) | |
647 | resp = self.addinfourl_wrapper(gz, old_resp.headers, old_resp.url, old_resp.code) | |
648 | resp.msg = old_resp.msg | |
649 | return resp | |
650 | ||
651 | https_request = http_request | |
652 | https_response = http_response | |
653 | ||
654 | ||
655 | class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler): | |
656 | def __init__(self, params, https_conn_class=None, *args, **kwargs): | |
657 | compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs) | |
658 | self._https_conn_class = https_conn_class or compat_http_client.HTTPSConnection | |
659 | self._params = params | |
660 | ||
661 | def https_open(self, req): | |
662 | return self.do_open(functools.partial( | |
663 | _create_http_connection, self, self._https_conn_class, True), | |
664 | req) | |
665 | ||
666 | ||
667 | def parse_iso8601(date_str, delimiter='T'): | |
668 | """ Return a UNIX timestamp from the given date """ | |
669 | ||
670 | if date_str is None: | |
671 | return None | |
672 | ||
673 | m = re.search( | |
674 | r'(\.[0-9]+)?(?:Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)', | |
675 | date_str) | |
676 | if not m: | |
677 | timezone = datetime.timedelta() | |
678 | else: | |
679 | date_str = date_str[:-len(m.group(0))] | |
680 | if not m.group('sign'): | |
681 | timezone = datetime.timedelta() | |
682 | else: | |
683 | sign = 1 if m.group('sign') == '+' else -1 | |
684 | timezone = datetime.timedelta( | |
685 | hours=sign * int(m.group('hours')), | |
686 | minutes=sign * int(m.group('minutes'))) | |
687 | date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter) | |
688 | dt = datetime.datetime.strptime(date_str, date_format) - timezone | |
689 | return calendar.timegm(dt.timetuple()) | |
690 | ||
691 | ||
692 | def unified_strdate(date_str, day_first=True): | |
693 | """Return a string with the date in the format YYYYMMDD""" | |
694 | ||
695 | if date_str is None: | |
696 | return None | |
697 | upload_date = None | |
698 | # Replace commas | |
699 | date_str = date_str.replace(',', ' ') | |
700 | # %z (UTC offset) is only supported in python>=3.2 | |
701 | date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str) | |
702 | # Remove AM/PM + timezone | |
703 | date_str = re.sub(r'(?i)\s*(?:AM|PM)\s+[A-Z]+', '', date_str) | |
704 | ||
705 | format_expressions = [ | |
706 | '%d %B %Y', | |
707 | '%d %b %Y', | |
708 | '%B %d %Y', | |
709 | '%b %d %Y', | |
710 | '%b %dst %Y %I:%M%p', | |
711 | '%b %dnd %Y %I:%M%p', | |
712 | '%b %dth %Y %I:%M%p', | |
713 | '%Y %m %d', | |
714 | '%Y-%m-%d', | |
715 | '%Y/%m/%d', | |
716 | '%Y/%m/%d %H:%M:%S', | |
717 | '%Y-%m-%d %H:%M:%S', | |
718 | '%Y-%m-%d %H:%M:%S.%f', | |
719 | '%d.%m.%Y %H:%M', | |
720 | '%d.%m.%Y %H.%M', | |
721 | '%Y-%m-%dT%H:%M:%SZ', | |
722 | '%Y-%m-%dT%H:%M:%S.%fZ', | |
723 | '%Y-%m-%dT%H:%M:%S.%f0Z', | |
724 | '%Y-%m-%dT%H:%M:%S', | |
725 | '%Y-%m-%dT%H:%M:%S.%f', | |
726 | '%Y-%m-%dT%H:%M', | |
727 | ] | |
728 | if day_first: | |
729 | format_expressions.extend([ | |
730 | '%d.%m.%Y', | |
731 | '%d/%m/%Y', | |
732 | '%d/%m/%y', | |
733 | '%d/%m/%Y %H:%M:%S', | |
734 | ]) | |
735 | else: | |
736 | format_expressions.extend([ | |
737 | '%m.%d.%Y', | |
738 | '%m/%d/%Y', | |
739 | '%m/%d/%y', | |
740 | '%m/%d/%Y %H:%M:%S', | |
741 | ]) | |
742 | for expression in format_expressions: | |
743 | try: | |
744 | upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d') | |
745 | except ValueError: | |
746 | pass | |
747 | if upload_date is None: | |
748 | timetuple = email.utils.parsedate_tz(date_str) | |
749 | if timetuple: | |
750 | upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d') | |
751 | return upload_date | |
752 | ||
753 | ||
754 | def determine_ext(url, default_ext='unknown_video'): | |
755 | if url is None: | |
756 | return default_ext | |
757 | guess = url.partition('?')[0].rpartition('.')[2] | |
758 | if re.match(r'^[A-Za-z0-9]+$', guess): | |
759 | return guess | |
760 | else: | |
761 | return default_ext | |
762 | ||
763 | ||
764 | def subtitles_filename(filename, sub_lang, sub_format): | |
765 | return filename.rsplit('.', 1)[0] + '.' + sub_lang + '.' + sub_format | |
766 | ||
767 | ||
768 | def date_from_str(date_str): | |
769 | """ | |
770 | Return a datetime object from a string in the format YYYYMMDD or | |
771 | (now|today)[+-][0-9](day|week|month|year)(s)?""" | |
772 | today = datetime.date.today() | |
773 | if date_str in ('now', 'today'): | |
774 | return today | |
775 | if date_str == 'yesterday': | |
776 | return today - datetime.timedelta(days=1) | |
777 | match = re.match('(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str) | |
778 | if match is not None: | |
779 | sign = match.group('sign') | |
780 | time = int(match.group('time')) | |
781 | if sign == '-': | |
782 | time = -time | |
783 | unit = match.group('unit') | |
784 | # A bad aproximation? | |
785 | if unit == 'month': | |
786 | unit = 'day' | |
787 | time *= 30 | |
788 | elif unit == 'year': | |
789 | unit = 'day' | |
790 | time *= 365 | |
791 | unit += 's' | |
792 | delta = datetime.timedelta(**{unit: time}) | |
793 | return today + delta | |
794 | return datetime.datetime.strptime(date_str, "%Y%m%d").date() | |
795 | ||
796 | ||
797 | def hyphenate_date(date_str): | |
798 | """ | |
799 | Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format""" | |
800 | match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str) | |
801 | if match is not None: | |
802 | return '-'.join(match.groups()) | |
803 | else: | |
804 | return date_str | |
805 | ||
806 | ||
807 | class DateRange(object): | |
808 | """Represents a time interval between two dates""" | |
809 | ||
810 | def __init__(self, start=None, end=None): | |
811 | """start and end must be strings in the format accepted by date""" | |
812 | if start is not None: | |
813 | self.start = date_from_str(start) | |
814 | else: | |
815 | self.start = datetime.datetime.min.date() | |
816 | if end is not None: | |
817 | self.end = date_from_str(end) | |
818 | else: | |
819 | self.end = datetime.datetime.max.date() | |
820 | if self.start > self.end: | |
821 | raise ValueError('Date range: "%s" , the start date must be before the end date' % self) | |
822 | ||
823 | @classmethod | |
824 | def day(cls, day): | |
825 | """Returns a range that only contains the given day""" | |
826 | return cls(day, day) | |
827 | ||
828 | def __contains__(self, date): | |
829 | """Check if the date is in the range""" | |
830 | if not isinstance(date, datetime.date): | |
831 | date = date_from_str(date) | |
832 | return self.start <= date <= self.end | |
833 | ||
834 | def __str__(self): | |
835 | return '%s - %s' % (self.start.isoformat(), self.end.isoformat()) | |
836 | ||
837 | ||
838 | def platform_name(): | |
839 | """ Returns the platform name as a compat_str """ | |
840 | res = platform.platform() | |
841 | if isinstance(res, bytes): | |
842 | res = res.decode(preferredencoding()) | |
843 | ||
844 | assert isinstance(res, compat_str) | |
845 | return res | |
846 | ||
847 | ||
848 | def _windows_write_string(s, out): | |
849 | """ Returns True if the string was written using special methods, | |
850 | False if it has yet to be written out.""" | |
851 | # Adapted from http://stackoverflow.com/a/3259271/35070 | |
852 | ||
853 | import ctypes | |
854 | import ctypes.wintypes | |
855 | ||
856 | WIN_OUTPUT_IDS = { | |
857 | 1: -11, | |
858 | 2: -12, | |
859 | } | |
860 | ||
861 | try: | |
862 | fileno = out.fileno() | |
863 | except AttributeError: | |
864 | # If the output stream doesn't have a fileno, it's virtual | |
865 | return False | |
866 | except io.UnsupportedOperation: | |
867 | # Some strange Windows pseudo files? | |
868 | return False | |
869 | if fileno not in WIN_OUTPUT_IDS: | |
870 | return False | |
871 | ||
872 | GetStdHandle = ctypes.WINFUNCTYPE( | |
873 | ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)( | |
874 | (b"GetStdHandle", ctypes.windll.kernel32)) | |
875 | h = GetStdHandle(WIN_OUTPUT_IDS[fileno]) | |
876 | ||
877 | WriteConsoleW = ctypes.WINFUNCTYPE( | |
878 | ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR, | |
879 | ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD), | |
880 | ctypes.wintypes.LPVOID)((b"WriteConsoleW", ctypes.windll.kernel32)) | |
881 | written = ctypes.wintypes.DWORD(0) | |
882 | ||
883 | GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b"GetFileType", ctypes.windll.kernel32)) | |
884 | FILE_TYPE_CHAR = 0x0002 | |
885 | FILE_TYPE_REMOTE = 0x8000 | |
886 | GetConsoleMode = ctypes.WINFUNCTYPE( | |
887 | ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, | |
888 | ctypes.POINTER(ctypes.wintypes.DWORD))( | |
889 | (b"GetConsoleMode", ctypes.windll.kernel32)) | |
890 | INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value | |
891 | ||
892 | def not_a_console(handle): | |
893 | if handle == INVALID_HANDLE_VALUE or handle is None: | |
894 | return True | |
895 | return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR | |
896 | or GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0) | |
897 | ||
898 | if not_a_console(h): | |
899 | return False | |
900 | ||
901 | def next_nonbmp_pos(s): | |
902 | try: | |
903 | return next(i for i, c in enumerate(s) if ord(c) > 0xffff) | |
904 | except StopIteration: | |
905 | return len(s) | |
906 | ||
907 | while s: | |
908 | count = min(next_nonbmp_pos(s), 1024) | |
909 | ||
910 | ret = WriteConsoleW( | |
911 | h, s, count if count else 2, ctypes.byref(written), None) | |
912 | if ret == 0: | |
913 | raise OSError('Failed to write string') | |
914 | if not count: # We just wrote a non-BMP character | |
915 | assert written.value == 2 | |
916 | s = s[1:] | |
917 | else: | |
918 | assert written.value > 0 | |
919 | s = s[written.value:] | |
920 | return True | |
921 | ||
922 | ||
923 | def write_string(s, out=None, encoding=None): | |
924 | if out is None: | |
925 | out = sys.stderr | |
926 | assert type(s) == compat_str | |
927 | ||
928 | if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'): | |
929 | if _windows_write_string(s, out): | |
930 | return | |
931 | ||
932 | if ('b' in getattr(out, 'mode', '') or | |
933 | sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr | |
934 | byt = s.encode(encoding or preferredencoding(), 'ignore') | |
935 | out.write(byt) | |
936 | elif hasattr(out, 'buffer'): | |
937 | enc = encoding or getattr(out, 'encoding', None) or preferredencoding() | |
938 | byt = s.encode(enc, 'ignore') | |
939 | out.buffer.write(byt) | |
940 | else: | |
941 | out.write(s) | |
942 | out.flush() | |
943 | ||
944 | ||
945 | def bytes_to_intlist(bs): | |
946 | if not bs: | |
947 | return [] | |
948 | if isinstance(bs[0], int): # Python 3 | |
949 | return list(bs) | |
950 | else: | |
951 | return [ord(c) for c in bs] | |
952 | ||
953 | ||
954 | def intlist_to_bytes(xs): | |
955 | if not xs: | |
956 | return b'' | |
957 | return struct_pack('%dB' % len(xs), *xs) | |
958 | ||
959 | ||
960 | # Cross-platform file locking | |
961 | if sys.platform == 'win32': | |
962 | import ctypes.wintypes | |
963 | import msvcrt | |
964 | ||
965 | class OVERLAPPED(ctypes.Structure): | |
966 | _fields_ = [ | |
967 | ('Internal', ctypes.wintypes.LPVOID), | |
968 | ('InternalHigh', ctypes.wintypes.LPVOID), | |
969 | ('Offset', ctypes.wintypes.DWORD), | |
970 | ('OffsetHigh', ctypes.wintypes.DWORD), | |
971 | ('hEvent', ctypes.wintypes.HANDLE), | |
972 | ] | |
973 | ||
974 | kernel32 = ctypes.windll.kernel32 | |
975 | LockFileEx = kernel32.LockFileEx | |
976 | LockFileEx.argtypes = [ | |
977 | ctypes.wintypes.HANDLE, # hFile | |
978 | ctypes.wintypes.DWORD, # dwFlags | |
979 | ctypes.wintypes.DWORD, # dwReserved | |
980 | ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow | |
981 | ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh | |
982 | ctypes.POINTER(OVERLAPPED) # Overlapped | |
983 | ] | |
984 | LockFileEx.restype = ctypes.wintypes.BOOL | |
985 | UnlockFileEx = kernel32.UnlockFileEx | |
986 | UnlockFileEx.argtypes = [ | |
987 | ctypes.wintypes.HANDLE, # hFile | |
988 | ctypes.wintypes.DWORD, # dwReserved | |
989 | ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow | |
990 | ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh | |
991 | ctypes.POINTER(OVERLAPPED) # Overlapped | |
992 | ] | |
993 | UnlockFileEx.restype = ctypes.wintypes.BOOL | |
994 | whole_low = 0xffffffff | |
995 | whole_high = 0x7fffffff | |
996 | ||
997 | def _lock_file(f, exclusive): | |
998 | overlapped = OVERLAPPED() | |
999 | overlapped.Offset = 0 | |
1000 | overlapped.OffsetHigh = 0 | |
1001 | overlapped.hEvent = 0 | |
1002 | f._lock_file_overlapped_p = ctypes.pointer(overlapped) | |
1003 | handle = msvcrt.get_osfhandle(f.fileno()) | |
1004 | if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0, | |
1005 | whole_low, whole_high, f._lock_file_overlapped_p): | |
1006 | raise OSError('Locking file failed: %r' % ctypes.FormatError()) | |
1007 | ||
1008 | def _unlock_file(f): | |
1009 | assert f._lock_file_overlapped_p | |
1010 | handle = msvcrt.get_osfhandle(f.fileno()) | |
1011 | if not UnlockFileEx(handle, 0, | |
1012 | whole_low, whole_high, f._lock_file_overlapped_p): | |
1013 | raise OSError('Unlocking file failed: %r' % ctypes.FormatError()) | |
1014 | ||
1015 | else: | |
1016 | import fcntl | |
1017 | ||
1018 | def _lock_file(f, exclusive): | |
1019 | fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH) | |
1020 | ||
1021 | def _unlock_file(f): | |
1022 | fcntl.flock(f, fcntl.LOCK_UN) | |
1023 | ||
1024 | ||
1025 | class locked_file(object): | |
1026 | def __init__(self, filename, mode, encoding=None): | |
1027 | assert mode in ['r', 'a', 'w'] | |
1028 | self.f = io.open(filename, mode, encoding=encoding) | |
1029 | self.mode = mode | |
1030 | ||
1031 | def __enter__(self): | |
1032 | exclusive = self.mode != 'r' | |
1033 | try: | |
1034 | _lock_file(self.f, exclusive) | |
1035 | except IOError: | |
1036 | self.f.close() | |
1037 | raise | |
1038 | return self | |
1039 | ||
1040 | def __exit__(self, etype, value, traceback): | |
1041 | try: | |
1042 | _unlock_file(self.f) | |
1043 | finally: | |
1044 | self.f.close() | |
1045 | ||
1046 | def __iter__(self): | |
1047 | return iter(self.f) | |
1048 | ||
1049 | def write(self, *args): | |
1050 | return self.f.write(*args) | |
1051 | ||
1052 | def read(self, *args): | |
1053 | return self.f.read(*args) | |
1054 | ||
1055 | ||
1056 | def get_filesystem_encoding(): | |
1057 | encoding = sys.getfilesystemencoding() | |
1058 | return encoding if encoding is not None else 'utf-8' | |
1059 | ||
1060 | ||
1061 | def shell_quote(args): | |
1062 | quoted_args = [] | |
1063 | encoding = get_filesystem_encoding() | |
1064 | for a in args: | |
1065 | if isinstance(a, bytes): | |
1066 | # We may get a filename encoded with 'encodeFilename' | |
1067 | a = a.decode(encoding) | |
1068 | quoted_args.append(pipes.quote(a)) | |
1069 | return ' '.join(quoted_args) | |
1070 | ||
1071 | ||
1072 | def takewhile_inclusive(pred, seq): | |
1073 | """ Like itertools.takewhile, but include the latest evaluated element | |
1074 | (the first element so that Not pred(e)) """ | |
1075 | for e in seq: | |
1076 | yield e | |
1077 | if not pred(e): | |
1078 | return | |
1079 | ||
1080 | ||
1081 | def smuggle_url(url, data): | |
1082 | """ Pass additional data in a URL for internal use. """ | |
1083 | ||
1084 | sdata = compat_urllib_parse.urlencode( | |
1085 | {'__youtubedl_smuggle': json.dumps(data)}) | |
1086 | return url + '#' + sdata | |
1087 | ||
1088 | ||
1089 | def unsmuggle_url(smug_url, default=None): | |
1090 | if '#__youtubedl_smuggle' not in smug_url: | |
1091 | return smug_url, default | |
1092 | url, _, sdata = smug_url.rpartition('#') | |
1093 | jsond = compat_parse_qs(sdata)['__youtubedl_smuggle'][0] | |
1094 | data = json.loads(jsond) | |
1095 | return url, data | |
1096 | ||
1097 | ||
1098 | def format_bytes(bytes): | |
1099 | if bytes is None: | |
1100 | return 'N/A' | |
1101 | if type(bytes) is str: | |
1102 | bytes = float(bytes) | |
1103 | if bytes == 0.0: | |
1104 | exponent = 0 | |
1105 | else: | |
1106 | exponent = int(math.log(bytes, 1024.0)) | |
1107 | suffix = ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent] | |
1108 | converted = float(bytes) / float(1024 ** exponent) | |
1109 | return '%.2f%s' % (converted, suffix) | |
1110 | ||
1111 | ||
1112 | def parse_filesize(s): | |
1113 | if s is None: | |
1114 | return None | |
1115 | ||
1116 | # The lower-case forms are of course incorrect and inofficial, | |
1117 | # but we support those too | |
1118 | _UNIT_TABLE = { | |
1119 | 'B': 1, | |
1120 | 'b': 1, | |
1121 | 'KiB': 1024, | |
1122 | 'KB': 1000, | |
1123 | 'kB': 1024, | |
1124 | 'Kb': 1000, | |
1125 | 'MiB': 1024 ** 2, | |
1126 | 'MB': 1000 ** 2, | |
1127 | 'mB': 1024 ** 2, | |
1128 | 'Mb': 1000 ** 2, | |
1129 | 'GiB': 1024 ** 3, | |
1130 | 'GB': 1000 ** 3, | |
1131 | 'gB': 1024 ** 3, | |
1132 | 'Gb': 1000 ** 3, | |
1133 | 'TiB': 1024 ** 4, | |
1134 | 'TB': 1000 ** 4, | |
1135 | 'tB': 1024 ** 4, | |
1136 | 'Tb': 1000 ** 4, | |
1137 | 'PiB': 1024 ** 5, | |
1138 | 'PB': 1000 ** 5, | |
1139 | 'pB': 1024 ** 5, | |
1140 | 'Pb': 1000 ** 5, | |
1141 | 'EiB': 1024 ** 6, | |
1142 | 'EB': 1000 ** 6, | |
1143 | 'eB': 1024 ** 6, | |
1144 | 'Eb': 1000 ** 6, | |
1145 | 'ZiB': 1024 ** 7, | |
1146 | 'ZB': 1000 ** 7, | |
1147 | 'zB': 1024 ** 7, | |
1148 | 'Zb': 1000 ** 7, | |
1149 | 'YiB': 1024 ** 8, | |
1150 | 'YB': 1000 ** 8, | |
1151 | 'yB': 1024 ** 8, | |
1152 | 'Yb': 1000 ** 8, | |
1153 | } | |
1154 | ||
1155 | units_re = '|'.join(re.escape(u) for u in _UNIT_TABLE) | |
1156 | m = re.match( | |
1157 | r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)' % units_re, s) | |
1158 | if not m: | |
1159 | return None | |
1160 | ||
1161 | num_str = m.group('num').replace(',', '.') | |
1162 | mult = _UNIT_TABLE[m.group('unit')] | |
1163 | return int(float(num_str) * mult) | |
1164 | ||
1165 | ||
1166 | def get_term_width(): | |
1167 | columns = compat_getenv('COLUMNS', None) | |
1168 | if columns: | |
1169 | return int(columns) | |
1170 | ||
1171 | try: | |
1172 | sp = subprocess.Popen( | |
1173 | ['stty', 'size'], | |
1174 | stdout=subprocess.PIPE, stderr=subprocess.PIPE) | |
1175 | out, err = sp.communicate() | |
1176 | return int(out.split()[1]) | |
1177 | except: | |
1178 | pass | |
1179 | return None | |
1180 | ||
1181 | ||
1182 | def month_by_name(name): | |
1183 | """ Return the number of a month by (locale-independently) English name """ | |
1184 | ||
1185 | ENGLISH_NAMES = [ | |
1186 | 'January', 'February', 'March', 'April', 'May', 'June', | |
1187 | 'July', 'August', 'September', 'October', 'November', 'December'] | |
1188 | try: | |
1189 | return ENGLISH_NAMES.index(name) + 1 | |
1190 | except ValueError: | |
1191 | return None | |
1192 | ||
1193 | ||
1194 | def fix_xml_ampersands(xml_str): | |
1195 | """Replace all the '&' by '&' in XML""" | |
1196 | return re.sub( | |
1197 | r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)', | |
1198 | '&', | |
1199 | xml_str) | |
1200 | ||
1201 | ||
1202 | def setproctitle(title): | |
1203 | assert isinstance(title, compat_str) | |
1204 | try: | |
1205 | libc = ctypes.cdll.LoadLibrary("libc.so.6") | |
1206 | except OSError: | |
1207 | return | |
1208 | title_bytes = title.encode('utf-8') | |
1209 | buf = ctypes.create_string_buffer(len(title_bytes)) | |
1210 | buf.value = title_bytes | |
1211 | try: | |
1212 | libc.prctl(15, buf, 0, 0, 0) | |
1213 | except AttributeError: | |
1214 | return # Strange libc, just skip this | |
1215 | ||
1216 | ||
1217 | def remove_start(s, start): | |
1218 | if s.startswith(start): | |
1219 | return s[len(start):] | |
1220 | return s | |
1221 | ||
1222 | ||
1223 | def remove_end(s, end): | |
1224 | if s.endswith(end): | |
1225 | return s[:-len(end)] | |
1226 | return s | |
1227 | ||
1228 | ||
1229 | def url_basename(url): | |
1230 | path = compat_urlparse.urlparse(url).path | |
1231 | return path.strip('/').split('/')[-1] | |
1232 | ||
1233 | ||
1234 | class HEADRequest(compat_urllib_request.Request): | |
1235 | def get_method(self): | |
1236 | return "HEAD" | |
1237 | ||
1238 | ||
1239 | def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1): | |
1240 | if get_attr: | |
1241 | if v is not None: | |
1242 | v = getattr(v, get_attr, None) | |
1243 | if v == '': | |
1244 | v = None | |
1245 | return default if v is None else (int(v) * invscale // scale) | |
1246 | ||
1247 | ||
1248 | def str_or_none(v, default=None): | |
1249 | return default if v is None else compat_str(v) | |
1250 | ||
1251 | ||
1252 | def str_to_int(int_str): | |
1253 | """ A more relaxed version of int_or_none """ | |
1254 | if int_str is None: | |
1255 | return None | |
1256 | int_str = re.sub(r'[,\.\+]', '', int_str) | |
1257 | return int(int_str) | |
1258 | ||
1259 | ||
1260 | def float_or_none(v, scale=1, invscale=1, default=None): | |
1261 | return default if v is None else (float(v) * invscale / scale) | |
1262 | ||
1263 | ||
1264 | def parse_duration(s): | |
1265 | if not isinstance(s, basestring if sys.version_info < (3, 0) else compat_str): | |
1266 | return None | |
1267 | ||
1268 | s = s.strip() | |
1269 | ||
1270 | m = re.match( | |
1271 | r'''(?ix)(?:P?T)? | |
1272 | (?: | |
1273 | (?P<only_mins>[0-9.]+)\s*(?:mins?|minutes?)\s*| | |
1274 | (?P<only_hours>[0-9.]+)\s*(?:hours?)| | |
1275 | ||
1276 | (?: | |
1277 | (?:(?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*)? | |
1278 | (?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s* | |
1279 | )? | |
1280 | (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)? | |
1281 | )$''', s) | |
1282 | if not m: | |
1283 | return None | |
1284 | res = 0 | |
1285 | if m.group('only_mins'): | |
1286 | return float_or_none(m.group('only_mins'), invscale=60) | |
1287 | if m.group('only_hours'): | |
1288 | return float_or_none(m.group('only_hours'), invscale=60 * 60) | |
1289 | if m.group('secs'): | |
1290 | res += int(m.group('secs')) | |
1291 | if m.group('mins'): | |
1292 | res += int(m.group('mins')) * 60 | |
1293 | if m.group('hours'): | |
1294 | res += int(m.group('hours')) * 60 * 60 | |
1295 | if m.group('ms'): | |
1296 | res += float(m.group('ms')) | |
1297 | return res | |
1298 | ||
1299 | ||
1300 | def prepend_extension(filename, ext): | |
1301 | name, real_ext = os.path.splitext(filename) | |
1302 | return '{0}.{1}{2}'.format(name, ext, real_ext) | |
1303 | ||
1304 | ||
1305 | def check_executable(exe, args=[]): | |
1306 | """ Checks if the given binary is installed somewhere in PATH, and returns its name. | |
1307 | args can be a list of arguments for a short output (like -version) """ | |
1308 | try: | |
1309 | subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate() | |
1310 | except OSError: | |
1311 | return False | |
1312 | return exe | |
1313 | ||
1314 | ||
1315 | def get_exe_version(exe, args=['--version'], | |
1316 | version_re=None, unrecognized='present'): | |
1317 | """ Returns the version of the specified executable, | |
1318 | or False if the executable is not present """ | |
1319 | try: | |
1320 | out, _ = subprocess.Popen( | |
1321 | [exe] + args, | |
1322 | stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate() | |
1323 | except OSError: | |
1324 | return False | |
1325 | if isinstance(out, bytes): # Python 2.x | |
1326 | out = out.decode('ascii', 'ignore') | |
1327 | return detect_exe_version(out, version_re, unrecognized) | |
1328 | ||
1329 | ||
1330 | def detect_exe_version(output, version_re=None, unrecognized='present'): | |
1331 | assert isinstance(output, compat_str) | |
1332 | if version_re is None: | |
1333 | version_re = r'version\s+([-0-9._a-zA-Z]+)' | |
1334 | m = re.search(version_re, output) | |
1335 | if m: | |
1336 | return m.group(1) | |
1337 | else: | |
1338 | return unrecognized | |
1339 | ||
1340 | ||
1341 | class PagedList(object): | |
1342 | def __len__(self): | |
1343 | # This is only useful for tests | |
1344 | return len(self.getslice()) | |
1345 | ||
1346 | ||
1347 | class OnDemandPagedList(PagedList): | |
1348 | def __init__(self, pagefunc, pagesize): | |
1349 | self._pagefunc = pagefunc | |
1350 | self._pagesize = pagesize | |
1351 | ||
1352 | def getslice(self, start=0, end=None): | |
1353 | res = [] | |
1354 | for pagenum in itertools.count(start // self._pagesize): | |
1355 | firstid = pagenum * self._pagesize | |
1356 | nextfirstid = pagenum * self._pagesize + self._pagesize | |
1357 | if start >= nextfirstid: | |
1358 | continue | |
1359 | ||
1360 | page_results = list(self._pagefunc(pagenum)) | |
1361 | ||
1362 | startv = ( | |
1363 | start % self._pagesize | |
1364 | if firstid <= start < nextfirstid | |
1365 | else 0) | |
1366 | ||
1367 | endv = ( | |
1368 | ((end - 1) % self._pagesize) + 1 | |
1369 | if (end is not None and firstid <= end <= nextfirstid) | |
1370 | else None) | |
1371 | ||
1372 | if startv != 0 or endv is not None: | |
1373 | page_results = page_results[startv:endv] | |
1374 | res.extend(page_results) | |
1375 | ||
1376 | # A little optimization - if current page is not "full", ie. does | |
1377 | # not contain page_size videos then we can assume that this page | |
1378 | # is the last one - there are no more ids on further pages - | |
1379 | # i.e. no need to query again. | |
1380 | if len(page_results) + startv < self._pagesize: | |
1381 | break | |
1382 | ||
1383 | # If we got the whole page, but the next page is not interesting, | |
1384 | # break out early as well | |
1385 | if end == nextfirstid: | |
1386 | break | |
1387 | return res | |
1388 | ||
1389 | ||
1390 | class InAdvancePagedList(PagedList): | |
1391 | def __init__(self, pagefunc, pagecount, pagesize): | |
1392 | self._pagefunc = pagefunc | |
1393 | self._pagecount = pagecount | |
1394 | self._pagesize = pagesize | |
1395 | ||
1396 | def getslice(self, start=0, end=None): | |
1397 | res = [] | |
1398 | start_page = start // self._pagesize | |
1399 | end_page = ( | |
1400 | self._pagecount if end is None else (end // self._pagesize + 1)) | |
1401 | skip_elems = start - start_page * self._pagesize | |
1402 | only_more = None if end is None else end - start | |
1403 | for pagenum in range(start_page, end_page): | |
1404 | page = list(self._pagefunc(pagenum)) | |
1405 | if skip_elems: | |
1406 | page = page[skip_elems:] | |
1407 | skip_elems = None | |
1408 | if only_more is not None: | |
1409 | if len(page) < only_more: | |
1410 | only_more -= len(page) | |
1411 | else: | |
1412 | page = page[:only_more] | |
1413 | res.extend(page) | |
1414 | break | |
1415 | res.extend(page) | |
1416 | return res | |
1417 | ||
1418 | ||
1419 | def uppercase_escape(s): | |
1420 | unicode_escape = codecs.getdecoder('unicode_escape') | |
1421 | return re.sub( | |
1422 | r'\\U[0-9a-fA-F]{8}', | |
1423 | lambda m: unicode_escape(m.group(0))[0], | |
1424 | s) | |
1425 | ||
1426 | ||
1427 | def escape_rfc3986(s): | |
1428 | """Escape non-ASCII characters as suggested by RFC 3986""" | |
1429 | if sys.version_info < (3, 0) and isinstance(s, unicode): | |
1430 | s = s.encode('utf-8') | |
1431 | return compat_urllib_parse.quote(s, b"%/;:@&=+$,!~*'()?#[]") | |
1432 | ||
1433 | ||
1434 | def escape_url(url): | |
1435 | """Escape URL as suggested by RFC 3986""" | |
1436 | url_parsed = compat_urllib_parse_urlparse(url) | |
1437 | return url_parsed._replace( | |
1438 | path=escape_rfc3986(url_parsed.path), | |
1439 | params=escape_rfc3986(url_parsed.params), | |
1440 | query=escape_rfc3986(url_parsed.query), | |
1441 | fragment=escape_rfc3986(url_parsed.fragment) | |
1442 | ).geturl() | |
1443 | ||
1444 | try: | |
1445 | struct.pack('!I', 0) | |
1446 | except TypeError: | |
1447 | # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument | |
1448 | def struct_pack(spec, *args): | |
1449 | if isinstance(spec, compat_str): | |
1450 | spec = spec.encode('ascii') | |
1451 | return struct.pack(spec, *args) | |
1452 | ||
1453 | def struct_unpack(spec, *args): | |
1454 | if isinstance(spec, compat_str): | |
1455 | spec = spec.encode('ascii') | |
1456 | return struct.unpack(spec, *args) | |
1457 | else: | |
1458 | struct_pack = struct.pack | |
1459 | struct_unpack = struct.unpack | |
1460 | ||
1461 | ||
1462 | def read_batch_urls(batch_fd): | |
1463 | def fixup(url): | |
1464 | if not isinstance(url, compat_str): | |
1465 | url = url.decode('utf-8', 'replace') | |
1466 | BOM_UTF8 = '\xef\xbb\xbf' | |
1467 | if url.startswith(BOM_UTF8): | |
1468 | url = url[len(BOM_UTF8):] | |
1469 | url = url.strip() | |
1470 | if url.startswith(('#', ';', ']')): | |
1471 | return False | |
1472 | return url | |
1473 | ||
1474 | with contextlib.closing(batch_fd) as fd: | |
1475 | return [url for url in map(fixup, fd) if url] | |
1476 | ||
1477 | ||
1478 | def urlencode_postdata(*args, **kargs): | |
1479 | return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii') | |
1480 | ||
1481 | ||
1482 | try: | |
1483 | etree_iter = xml.etree.ElementTree.Element.iter | |
1484 | except AttributeError: # Python <=2.6 | |
1485 | etree_iter = lambda n: n.findall('.//*') | |
1486 | ||
1487 | ||
1488 | def parse_xml(s): | |
1489 | class TreeBuilder(xml.etree.ElementTree.TreeBuilder): | |
1490 | def doctype(self, name, pubid, system): | |
1491 | pass # Ignore doctypes | |
1492 | ||
1493 | parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder()) | |
1494 | kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {} | |
1495 | tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs) | |
1496 | # Fix up XML parser in Python 2.x | |
1497 | if sys.version_info < (3, 0): | |
1498 | for n in etree_iter(tree): | |
1499 | if n.text is not None: | |
1500 | if not isinstance(n.text, compat_str): | |
1501 | n.text = n.text.decode('utf-8') | |
1502 | return tree | |
1503 | ||
1504 | ||
1505 | US_RATINGS = { | |
1506 | 'G': 0, | |
1507 | 'PG': 10, | |
1508 | 'PG-13': 13, | |
1509 | 'R': 16, | |
1510 | 'NC': 18, | |
1511 | } | |
1512 | ||
1513 | ||
1514 | def parse_age_limit(s): | |
1515 | if s is None: | |
1516 | return None | |
1517 | m = re.match(r'^(?P<age>\d{1,2})\+?$', s) | |
1518 | return int(m.group('age')) if m else US_RATINGS.get(s, None) | |
1519 | ||
1520 | ||
1521 | def strip_jsonp(code): | |
1522 | return re.sub( | |
1523 | r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code) | |
1524 | ||
1525 | ||
1526 | def js_to_json(code): | |
1527 | def fix_kv(m): | |
1528 | v = m.group(0) | |
1529 | if v in ('true', 'false', 'null'): | |
1530 | return v | |
1531 | if v.startswith('"'): | |
1532 | return v | |
1533 | if v.startswith("'"): | |
1534 | v = v[1:-1] | |
1535 | v = re.sub(r"\\\\|\\'|\"", lambda m: { | |
1536 | '\\\\': '\\\\', | |
1537 | "\\'": "'", | |
1538 | '"': '\\"', | |
1539 | }[m.group(0)], v) | |
1540 | return '"%s"' % v | |
1541 | ||
1542 | res = re.sub(r'''(?x) | |
1543 | "(?:[^"\\]*(?:\\\\|\\")?)*"| | |
1544 | '(?:[^'\\]*(?:\\\\|\\')?)*'| | |
1545 | [a-zA-Z_][a-zA-Z_0-9]* | |
1546 | ''', fix_kv, code) | |
1547 | res = re.sub(r',(\s*\])', lambda m: m.group(1), res) | |
1548 | return res | |
1549 | ||
1550 | ||
1551 | def qualities(quality_ids): | |
1552 | """ Get a numeric quality value out of a list of possible values """ | |
1553 | def q(qid): | |
1554 | try: | |
1555 | return quality_ids.index(qid) | |
1556 | except ValueError: | |
1557 | return -1 | |
1558 | return q | |
1559 | ||
1560 | ||
1561 | DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s' | |
1562 | ||
1563 | ||
1564 | def limit_length(s, length): | |
1565 | """ Add ellipses to overly long strings """ | |
1566 | if s is None: | |
1567 | return None | |
1568 | ELLIPSES = '...' | |
1569 | if len(s) > length: | |
1570 | return s[:length - len(ELLIPSES)] + ELLIPSES | |
1571 | return s | |
1572 | ||
1573 | ||
1574 | def version_tuple(v): | |
1575 | return tuple(int(e) for e in re.split(r'[-.]', v)) | |
1576 | ||
1577 | ||
1578 | def is_outdated_version(version, limit, assume_new=True): | |
1579 | if not version: | |
1580 | return not assume_new | |
1581 | try: | |
1582 | return version_tuple(version) < version_tuple(limit) | |
1583 | except ValueError: | |
1584 | return not assume_new | |
1585 | ||
1586 | ||
1587 | def ytdl_is_updateable(): | |
1588 | """ Returns if youtube-dl can be updated with -U """ | |
1589 | from zipimport import zipimporter | |
1590 | ||
1591 | return isinstance(globals().get('__loader__'), zipimporter) or hasattr(sys, 'frozen') | |
1592 | ||
1593 | ||
1594 | def args_to_str(args): | |
1595 | # Get a short string representation for a subprocess command | |
1596 | return ' '.join(shlex_quote(a) for a in args) | |
1597 | ||
1598 | ||
1599 | def urlhandle_detect_ext(url_handle): | |
1600 | try: | |
1601 | url_handle.headers | |
1602 | getheader = lambda h: url_handle.headers[h] | |
1603 | except AttributeError: # Python < 3 | |
1604 | getheader = url_handle.info().getheader | |
1605 | ||
1606 | cd = getheader('Content-Disposition') | |
1607 | if cd: | |
1608 | m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd) | |
1609 | if m: | |
1610 | e = determine_ext(m.group('filename'), default_ext=None) | |
1611 | if e: | |
1612 | return e | |
1613 | ||
1614 | return getheader('Content-Type').split("/")[1] | |
1615 | ||
1616 | ||
1617 | def age_restricted(content_limit, age_limit): | |
1618 | """ Returns True iff the content should be blocked """ | |
1619 | ||
1620 | if age_limit is None: # No limit set | |
1621 | return False | |
1622 | if content_limit is None: | |
1623 | return False # Content available for everyone | |
1624 | return age_limit < content_limit | |
1625 | ||
1626 | ||
1627 | def is_html(first_bytes): | |
1628 | """ Detect whether a file contains HTML by examining its first bytes. """ | |
1629 | ||
1630 | BOMS = [ | |
1631 | (b'\xef\xbb\xbf', 'utf-8'), | |
1632 | (b'\x00\x00\xfe\xff', 'utf-32-be'), | |
1633 | (b'\xff\xfe\x00\x00', 'utf-32-le'), | |
1634 | (b'\xff\xfe', 'utf-16-le'), | |
1635 | (b'\xfe\xff', 'utf-16-be'), | |
1636 | ] | |
1637 | for bom, enc in BOMS: | |
1638 | if first_bytes.startswith(bom): | |
1639 | s = first_bytes[len(bom):].decode(enc, 'replace') | |
1640 | break | |
1641 | else: | |
1642 | s = first_bytes.decode('utf-8', 'replace') | |
1643 | ||
1644 | return re.match(r'^\s*<', s) |