Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/utils.py

   1 #!/usr/bin/env python
   2 # -*- coding: utf-8 -*-
   3
   4 import calendar
   5 import codecs
   6 import contextlib
   7 import ctypes
   8 import datetime
   9 import email.utils
  10 import errno
  11 import getpass
  12 import gzip
  13 import itertools
  14 import io
  15 import json
  16 import locale
  17 import math
  18 import os
  19 import pipes
  20 import platform
  21 import re
  22 import ssl
  23 import socket
  24 import struct
  25 import subprocess
  26 import sys
  27 import traceback
  28 import xml.etree.ElementTree
  29 import zlib
  30
  31 try:
  32     import urllib.request as compat_urllib_request
  33 except ImportError: # Python 2
  34     import urllib2 as compat_urllib_request
  35
  36 try:
  37     import urllib.error as compat_urllib_error
  38 except ImportError: # Python 2
  39     import urllib2 as compat_urllib_error
  40
  41 try:
  42     import urllib.parse as compat_urllib_parse
  43 except ImportError: # Python 2
  44     import urllib as compat_urllib_parse
  45
  46 try:
  47     from urllib.parse import urlparse as compat_urllib_parse_urlparse
  48 except ImportError: # Python 2
  49     from urlparse import urlparse as compat_urllib_parse_urlparse
  50
  51 try:
  52     import urllib.parse as compat_urlparse
  53 except ImportError: # Python 2
  54     import urlparse as compat_urlparse
  55
  56 try:
  57     import http.cookiejar as compat_cookiejar
  58 except ImportError: # Python 2
  59     import cookielib as compat_cookiejar
  60
  61 try:
  62     import html.entities as compat_html_entities
  63 except ImportError: # Python 2
  64     import htmlentitydefs as compat_html_entities
  65
  66 try:
  67     import html.parser as compat_html_parser
  68 except ImportError: # Python 2
  69     import HTMLParser as compat_html_parser
  70
  71 try:
  72     import http.client as compat_http_client
  73 except ImportError: # Python 2
  74     import httplib as compat_http_client
  75
  76 try:
  77     from urllib.error import HTTPError as compat_HTTPError
  78 except ImportError:  # Python 2
  79     from urllib2 import HTTPError as compat_HTTPError
  80
  81 try:
  82     from urllib.request import urlretrieve as compat_urlretrieve
  83 except ImportError:  # Python 2
  84     from urllib import urlretrieve as compat_urlretrieve
  85
  86
  87 try:
  88     from subprocess import DEVNULL
  89     compat_subprocess_get_DEVNULL = lambda: DEVNULL
  90 except ImportError:
  91     compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w')
  92
  93 try:
  94     from urllib.parse import unquote as compat_urllib_parse_unquote
  95 except ImportError:
  96     def compat_urllib_parse_unquote(string, encoding='utf-8', errors='replace'):
  97         if string == '':
  98             return string
  99         res = string.split('%')
 100         if len(res) == 1:
 101             return string
 102         if encoding is None:
 103             encoding = 'utf-8'
 104         if errors is None:
 105             errors = 'replace'
 106         # pct_sequence: contiguous sequence of percent-encoded bytes, decoded
 107         pct_sequence = b''
 108         string = res[0]
 109         for item in res[1:]:
 110             try:
 111                 if not item:
 112                     raise ValueError
 113                 pct_sequence += item[:2].decode('hex')
 114                 rest = item[2:]
 115                 if not rest:
 116                     # This segment was just a single percent-encoded character.
 117                     # May be part of a sequence of code units, so delay decoding.
 118                     # (Stored in pct_sequence).
 119                     continue
 120             except ValueError:
 121                 rest = '%' + item
 122             # Encountered non-percent-encoded characters. Flush the current
 123             # pct_sequence.
 124             string += pct_sequence.decode(encoding, errors) + rest
 125             pct_sequence = b''
 126         if pct_sequence:
 127             # Flush the final pct_sequence
 128             string += pct_sequence.decode(encoding, errors)
 129         return string
 130
 131
 132 try:
 133     from urllib.parse import parse_qs as compat_parse_qs
 134 except ImportError: # Python 2
 135     # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib.
 136     # Python 2's version is apparently totally broken
 137
 138     def _parse_qsl(qs, keep_blank_values=False, strict_parsing=False,
 139                 encoding='utf-8', errors='replace'):
 140         qs, _coerce_result = qs, unicode
 141         pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]
 142         r = []
 143         for name_value in pairs:
 144             if not name_value and not strict_parsing:
 145                 continue
 146             nv = name_value.split('=', 1)
 147             if len(nv) != 2:
 148                 if strict_parsing:
 149                     raise ValueError("bad query field: %r" % (name_value,))
 150                 # Handle case of a control-name with no equal sign
 151                 if keep_blank_values:
 152                     nv.append('')
 153                 else:
 154                     continue
 155             if len(nv[1]) or keep_blank_values:
 156                 name = nv[0].replace('+', ' ')
 157                 name = compat_urllib_parse_unquote(
 158                     name, encoding=encoding, errors=errors)
 159                 name = _coerce_result(name)
 160                 value = nv[1].replace('+', ' ')
 161                 value = compat_urllib_parse_unquote(
 162                     value, encoding=encoding, errors=errors)
 163                 value = _coerce_result(value)
 164                 r.append((name, value))
 165         return r
 166
 167     def compat_parse_qs(qs, keep_blank_values=False, strict_parsing=False,
 168                 encoding='utf-8', errors='replace'):
 169         parsed_result = {}
 170         pairs = _parse_qsl(qs, keep_blank_values, strict_parsing,
 171                         encoding=encoding, errors=errors)
 172         for name, value in pairs:
 173             if name in parsed_result:
 174                 parsed_result[name].append(value)
 175             else:
 176                 parsed_result[name] = [value]
 177         return parsed_result
 178
 179 try:
 180     compat_str = unicode # Python 2
 181 except NameError:
 182     compat_str = str
 183
 184 try:
 185     compat_chr = unichr # Python 2
 186 except NameError:
 187     compat_chr = chr
 188
 189 try:
 190     from xml.etree.ElementTree import ParseError as compat_xml_parse_error
 191 except ImportError:  # Python 2.6
 192     from xml.parsers.expat import ExpatError as compat_xml_parse_error
 193
 194 def compat_ord(c):
 195     if type(c) is int: return c
 196     else: return ord(c)
 197
 198 # This is not clearly defined otherwise
 199 compiled_regex_type = type(re.compile(''))
 200
 201 std_headers = {
 202     'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20100101 Firefox/10.0 (Chrome)',
 203     'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
 204     'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
 205     'Accept-Encoding': 'gzip, deflate',
 206     'Accept-Language': 'en-us,en;q=0.5',
 207 }
 208
 209 def preferredencoding():
 210     """Get preferred encoding.
 211
 212     Returns the best encoding scheme for the system, based on
 213     locale.getpreferredencoding() and some further tweaks.
 214     """
 215     try:
 216         pref = locale.getpreferredencoding()
 217         u'TEST'.encode(pref)
 218     except:
 219         pref = 'UTF-8'
 220
 221     return pref
 222
 223 if sys.version_info < (3,0):
 224     def compat_print(s):
 225         print(s.encode(preferredencoding(), 'xmlcharrefreplace'))
 226 else:
 227     def compat_print(s):
 228         assert type(s) == type(u'')
 229         print(s)
 230
 231 # In Python 2.x, json.dump expects a bytestream.
 232 # In Python 3.x, it writes to a character stream
 233 if sys.version_info < (3,0):
 234     def write_json_file(obj, fn):
 235         with open(fn, 'wb') as f:
 236             json.dump(obj, f)
 237 else:
 238     def write_json_file(obj, fn):
 239         with open(fn, 'w', encoding='utf-8') as f:
 240             json.dump(obj, f)
 241
 242 if sys.version_info >= (2,7):
 243     def find_xpath_attr(node, xpath, key, val):
 244         """ Find the xpath xpath[@key=val] """
 245         assert re.match(r'^[a-zA-Z-]+$', key)
 246         assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val)
 247         expr = xpath + u"[@%s='%s']" % (key, val)
 248         return node.find(expr)
 249 else:
 250     def find_xpath_attr(node, xpath, key, val):
 251         for f in node.findall(xpath):
 252             if f.attrib.get(key) == val:
 253                 return f
 254         return None
 255
 256 # On python2.6 the xml.etree.ElementTree.Element methods don't support
 257 # the namespace parameter
 258 def xpath_with_ns(path, ns_map):
 259     components = [c.split(':') for c in path.split('/')]
 260     replaced = []
 261     for c in components:
 262         if len(c) == 1:
 263             replaced.append(c[0])
 264         else:
 265             ns, tag = c
 266             replaced.append('{%s}%s' % (ns_map[ns], tag))
 267     return '/'.join(replaced)
 268
 269 def htmlentity_transform(matchobj):
 270     """Transforms an HTML entity to a character.
 271
 272     This function receives a match object and is intended to be used with
 273     the re.sub() function.
 274     """
 275     entity = matchobj.group(1)
 276
 277     # Known non-numeric HTML entity
 278     if entity in compat_html_entities.name2codepoint:
 279         return compat_chr(compat_html_entities.name2codepoint[entity])
 280
 281     mobj = re.match(u'(?u)#(x?\\d+)', entity)
 282     if mobj is not None:
 283         numstr = mobj.group(1)
 284         if numstr.startswith(u'x'):
 285             base = 16
 286             numstr = u'0%s' % numstr
 287         else:
 288             base = 10
 289         return compat_chr(int(numstr, base))
 290
 291     # Unknown entity in name, return its literal representation
 292     return (u'&%s;' % entity)
 293
 294 compat_html_parser.locatestarttagend = re.compile(r"""<[a-zA-Z][-.a-zA-Z0-9:_]*(?:\s+(?:(?<=['"\s])[^\s/>][^\s/=>]*(?:\s*=+\s*(?:'[^']*'|"[^"]*"|(?!['"])[^>\s]*))?\s*)*)?\s*""", re.VERBOSE) # backport bugfix
 295 class BaseHTMLParser(compat_html_parser.HTMLParser):
 296     def __init(self):
 297         compat_html_parser.HTMLParser.__init__(self)
 298         self.html = None
 299
 300     def loads(self, html):
 301         self.html = html
 302         self.feed(html)
 303         self.close()
 304
 305 class AttrParser(BaseHTMLParser):
 306     """Modified HTMLParser that isolates a tag with the specified attribute"""
 307     def __init__(self, attribute, value):
 308         self.attribute = attribute
 309         self.value = value
 310         self.result = None
 311         self.started = False
 312         self.depth = {}
 313         self.watch_startpos = False
 314         self.error_count = 0
 315         BaseHTMLParser.__init__(self)
 316
 317     def error(self, message):
 318         if self.error_count > 10 or self.started:
 319             raise compat_html_parser.HTMLParseError(message, self.getpos())
 320         self.rawdata = '\n'.join(self.html.split('\n')[self.getpos()[0]:]) # skip one line
 321         self.error_count += 1
 322         self.goahead(1)
 323
 324     def handle_starttag(self, tag, attrs):
 325         attrs = dict(attrs)
 326         if self.started:
 327             self.find_startpos(None)
 328         if self.attribute in attrs and attrs[self.attribute] == self.value:
 329             self.result = [tag]
 330             self.started = True
 331             self.watch_startpos = True
 332         if self.started:
 333             if not tag in self.depth: self.depth[tag] = 0
 334             self.depth[tag] += 1
 335
 336     def handle_endtag(self, tag):
 337         if self.started:
 338             if tag in self.depth: self.depth[tag] -= 1
 339             if self.depth[self.result[0]] == 0:
 340                 self.started = False
 341                 self.result.append(self.getpos())
 342
 343     def find_startpos(self, x):
 344         """Needed to put the start position of the result (self.result[1])
 345         after the opening tag with the requested id"""
 346         if self.watch_startpos:
 347             self.watch_startpos = False
 348             self.result.append(self.getpos())
 349     handle_entityref = handle_charref = handle_data = handle_comment = \
 350     handle_decl = handle_pi = unknown_decl = find_startpos
 351
 352     def get_result(self):
 353         if self.result is None:
 354             return None
 355         if len(self.result) != 3:
 356             return None
 357         lines = self.html.split('\n')
 358         lines = lines[self.result[1][0]-1:self.result[2][0]]
 359         lines[0] = lines[0][self.result[1][1]:]
 360         if len(lines) == 1:
 361             lines[-1] = lines[-1][:self.result[2][1]-self.result[1][1]]
 362         lines[-1] = lines[-1][:self.result[2][1]]
 363         return '\n'.join(lines).strip()
 364 # Hack for https://github.com/rg3/youtube-dl/issues/662
 365 if sys.version_info < (2, 7, 3):
 366     AttrParser.parse_endtag = (lambda self, i:
 367         i + len("</scr'+'ipt>")
 368         if self.rawdata[i:].startswith("</scr'+'ipt>")
 369         else compat_html_parser.HTMLParser.parse_endtag(self, i))
 370
 371 def get_element_by_id(id, html):
 372     """Return the content of the tag with the specified ID in the passed HTML document"""
 373     return get_element_by_attribute("id", id, html)
 374
 375 def get_element_by_attribute(attribute, value, html):
 376     """Return the content of the tag with the specified attribute in the passed HTML document"""
 377     parser = AttrParser(attribute, value)
 378     try:
 379         parser.loads(html)
 380     except compat_html_parser.HTMLParseError:
 381         pass
 382     return parser.get_result()
 383
 384 class MetaParser(BaseHTMLParser):
 385     """
 386     Modified HTMLParser that isolates a meta tag with the specified name
 387     attribute.
 388     """
 389     def __init__(self, name):
 390         BaseHTMLParser.__init__(self)
 391         self.name = name
 392         self.content = None
 393         self.result = None
 394
 395     def handle_starttag(self, tag, attrs):
 396         if tag != 'meta':
 397             return
 398         attrs = dict(attrs)
 399         if attrs.get('name') == self.name:
 400             self.result = attrs.get('content')
 401
 402     def get_result(self):
 403         return self.result
 404
 405 def get_meta_content(name, html):
 406     """
 407     Return the content attribute from the meta tag with the given name attribute.
 408     """
 409     parser = MetaParser(name)
 410     try:
 411         parser.loads(html)
 412     except compat_html_parser.HTMLParseError:
 413         pass
 414     return parser.get_result()
 415
 416
 417 def clean_html(html):
 418     """Clean an HTML snippet into a readable string"""
 419     # Newline vs <br />
 420     html = html.replace('\n', ' ')
 421     html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html)
 422     html = re.sub(r'<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
 423     # Strip html tags
 424     html = re.sub('<.*?>', '', html)
 425     # Replace html entities
 426     html = unescapeHTML(html)
 427     return html.strip()
 428
 429
 430 def sanitize_open(filename, open_mode):
 431     """Try to open the given filename, and slightly tweak it if this fails.
 432
 433     Attempts to open the given filename. If this fails, it tries to change
 434     the filename slightly, step by step, until it's either able to open it
 435     or it fails and raises a final exception, like the standard open()
 436     function.
 437
 438     It returns the tuple (stream, definitive_file_name).
 439     """
 440     try:
 441         if filename == u'-':
 442             if sys.platform == 'win32':
 443                 import msvcrt
 444                 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
 445             return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
 446         stream = open(encodeFilename(filename), open_mode)
 447         return (stream, filename)
 448     except (IOError, OSError) as err:
 449         if err.errno in (errno.EACCES,):
 450             raise
 451
 452         # In case of error, try to remove win32 forbidden chars
 453         alt_filename = os.path.join(
 454                         re.sub(u'[/<>:"\\|\\\\?\\*]', u'#', path_part)
 455                         for path_part in os.path.split(filename)
 456                        )
 457         if alt_filename == filename:
 458             raise
 459         else:
 460             # An exception here should be caught in the caller
 461             stream = open(encodeFilename(filename), open_mode)
 462             return (stream, alt_filename)
 463
 464
 465 def timeconvert(timestr):
 466     """Convert RFC 2822 defined time string into system timestamp"""
 467     timestamp = None
 468     timetuple = email.utils.parsedate_tz(timestr)
 469     if timetuple is not None:
 470         timestamp = email.utils.mktime_tz(timetuple)
 471     return timestamp
 472
 473 def sanitize_filename(s, restricted=False, is_id=False):
 474     """Sanitizes a string so it could be used as part of a filename.
 475     If restricted is set, use a stricter subset of allowed characters.
 476     Set is_id if this is not an arbitrary string, but an ID that should be kept if possible
 477     """
 478     def replace_insane(char):
 479         if char == '?' or ord(char) < 32 or ord(char) == 127:
 480             return ''
 481         elif char == '"':
 482             return '' if restricted else '\''
 483         elif char == ':':
 484             return '_-' if restricted else ' -'
 485         elif char in '\\/|*<>':
 486             return '_'
 487         if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
 488             return '_'
 489         if restricted and ord(char) > 127:
 490             return '_'
 491         return char
 492
 493     result = u''.join(map(replace_insane, s))
 494     if not is_id:
 495         while '__' in result:
 496             result = result.replace('__', '_')
 497         result = result.strip('_')
 498         # Common case of "Foreign band name - English song title"
 499         if restricted and result.startswith('-_'):
 500             result = result[2:]
 501         if not result:
 502             result = '_'
 503     return result
 504
 505 def orderedSet(iterable):
 506     """ Remove all duplicates from the input iterable """
 507     res = []
 508     for el in iterable:
 509         if el not in res:
 510             res.append(el)
 511     return res
 512
 513
 514 def unescapeHTML(s):
 515     if s is None:
 516         return None
 517     assert type(s) == compat_str
 518
 519     result = re.sub(r'(?u)&(.+?);', htmlentity_transform, s)
 520     return result
 521
 522
 523 def encodeFilename(s, for_subprocess=False):
 524     """
 525     @param s The name of the file
 526     """
 527
 528     assert type(s) == compat_str
 529
 530     # Python 3 has a Unicode API
 531     if sys.version_info >= (3, 0):
 532         return s
 533
 534     if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
 535         # Pass u'' directly to use Unicode APIs on Windows 2000 and up
 536         # (Detecting Windows NT 4 is tricky because 'major >= 4' would
 537         # match Windows 9x series as well. Besides, NT 4 is obsolete.)
 538         if not for_subprocess:
 539             return s
 540         else:
 541             # For subprocess calls, encode with locale encoding
 542             # Refer to http://stackoverflow.com/a/9951851/35070
 543             encoding = preferredencoding()
 544     else:
 545         encoding = sys.getfilesystemencoding()
 546     if encoding is None:
 547         encoding = 'utf-8'
 548     return s.encode(encoding, 'ignore')
 549
 550
 551 def encodeArgument(s):
 552     if not isinstance(s, compat_str):
 553         # Legacy code that uses byte strings
 554         # Uncomment the following line after fixing all post processors
 555         #assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
 556         s = s.decode('ascii')
 557     return encodeFilename(s, True)
 558
 559
 560 def decodeOption(optval):
 561     if optval is None:
 562         return optval
 563     if isinstance(optval, bytes):
 564         optval = optval.decode(preferredencoding())
 565
 566     assert isinstance(optval, compat_str)
 567     return optval
 568
 569 def formatSeconds(secs):
 570     if secs > 3600:
 571         return '%d:%02d:%02d' % (secs // 3600, (secs % 3600) // 60, secs % 60)
 572     elif secs > 60:
 573         return '%d:%02d' % (secs // 60, secs % 60)
 574     else:
 575         return '%d' % secs
 576
 577
 578 def make_HTTPS_handler(opts_no_check_certificate, **kwargs):
 579     if sys.version_info < (3, 2):
 580         import httplib
 581
 582         class HTTPSConnectionV3(httplib.HTTPSConnection):
 583             def __init__(self, *args, **kwargs):
 584                 httplib.HTTPSConnection.__init__(self, *args, **kwargs)
 585
 586             def connect(self):
 587                 sock = socket.create_connection((self.host, self.port), self.timeout)
 588                 if getattr(self, '_tunnel_host', False):
 589                     self.sock = sock
 590                     self._tunnel()
 591                 try:
 592                     self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_SSLv3)
 593                 except ssl.SSLError:
 594                     self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_SSLv23)
 595
 596         class HTTPSHandlerV3(compat_urllib_request.HTTPSHandler):
 597             def https_open(self, req):
 598                 return self.do_open(HTTPSConnectionV3, req)
 599         return HTTPSHandlerV3(**kwargs)
 600     else:
 601         context = ssl.SSLContext(ssl.PROTOCOL_SSLv3)
 602         context.verify_mode = (ssl.CERT_NONE
 603                                if opts_no_check_certificate
 604                                else ssl.CERT_REQUIRED)
 605         context.set_default_verify_paths()
 606         try:
 607             context.load_default_certs()
 608         except AttributeError:
 609             pass  # Python < 3.4
 610         return compat_urllib_request.HTTPSHandler(context=context, **kwargs)
 611
 612 class ExtractorError(Exception):
 613     """Error during info extraction."""
 614     def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):
 615         """ tb, if given, is the original traceback (so that it can be printed out).
 616         If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
 617         """
 618
 619         if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError):
 620             expected = True
 621         if video_id is not None:
 622             msg = video_id + ': ' + msg
 623         if not expected:
 624             msg = msg + u'; please report this issue on https://yt-dl.org/bug . Be sure to call youtube-dl with the --verbose flag and include its complete output. Make sure you are using the latest version; type  youtube-dl -U  to update.'
 625         super(ExtractorError, self).__init__(msg)
 626
 627         self.traceback = tb
 628         self.exc_info = sys.exc_info()  # preserve original exception
 629         self.cause = cause
 630         self.video_id = video_id
 631
 632     def format_traceback(self):
 633         if self.traceback is None:
 634             return None
 635         return u''.join(traceback.format_tb(self.traceback))
 636
 637
 638 class RegexNotFoundError(ExtractorError):
 639     """Error when a regex didn't match"""
 640     pass
 641
 642
 643 class DownloadError(Exception):
 644     """Download Error exception.
 645
 646     This exception may be thrown by FileDownloader objects if they are not
 647     configured to continue on errors. They will contain the appropriate
 648     error message.
 649     """
 650     def __init__(self, msg, exc_info=None):
 651         """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
 652         super(DownloadError, self).__init__(msg)
 653         self.exc_info = exc_info
 654
 655
 656 class SameFileError(Exception):
 657     """Same File exception.
 658
 659     This exception will be thrown by FileDownloader objects if they detect
 660     multiple files would have to be downloaded to the same file on disk.
 661     """
 662     pass
 663
 664
 665 class PostProcessingError(Exception):
 666     """Post Processing exception.
 667
 668     This exception may be raised by PostProcessor's .run() method to
 669     indicate an error in the postprocessing task.
 670     """
 671     def __init__(self, msg):
 672         self.msg = msg
 673
 674 class MaxDownloadsReached(Exception):
 675     """ --max-downloads limit has been reached. """
 676     pass
 677
 678
 679 class UnavailableVideoError(Exception):
 680     """Unavailable Format exception.
 681
 682     This exception will be thrown when a video is requested
 683     in a format that is not available for that video.
 684     """
 685     pass
 686
 687
 688 class ContentTooShortError(Exception):
 689     """Content Too Short exception.
 690
 691     This exception may be raised by FileDownloader objects when a file they
 692     download is too small for what the server announced first, indicating
 693     the connection was probably interrupted.
 694     """
 695     # Both in bytes
 696     downloaded = None
 697     expected = None
 698
 699     def __init__(self, downloaded, expected):
 700         self.downloaded = downloaded
 701         self.expected = expected
 702
 703 class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
 704     """Handler for HTTP requests and responses.
 705
 706     This class, when installed with an OpenerDirector, automatically adds
 707     the standard headers to every HTTP request and handles gzipped and
 708     deflated responses from web servers. If compression is to be avoided in
 709     a particular request, the original request in the program code only has
 710     to include the HTTP header "Youtubedl-No-Compression", which will be
 711     removed before making the real request.
 712
 713     Part of this code was copied from:
 714
 715     http://techknack.net/python-urllib2-handlers/
 716
 717     Andrew Rowls, the author of that code, agreed to release it to the
 718     public domain.
 719     """
 720
 721     @staticmethod
 722     def deflate(data):
 723         try:
 724             return zlib.decompress(data, -zlib.MAX_WBITS)
 725         except zlib.error:
 726             return zlib.decompress(data)
 727
 728     @staticmethod
 729     def addinfourl_wrapper(stream, headers, url, code):
 730         if hasattr(compat_urllib_request.addinfourl, 'getcode'):
 731             return compat_urllib_request.addinfourl(stream, headers, url, code)
 732         ret = compat_urllib_request.addinfourl(stream, headers, url)
 733         ret.code = code
 734         return ret
 735
 736     def http_request(self, req):
 737         for h,v in std_headers.items():
 738             if h in req.headers:
 739                 del req.headers[h]
 740             req.add_header(h, v)
 741         if 'Youtubedl-no-compression' in req.headers:
 742             if 'Accept-encoding' in req.headers:
 743                 del req.headers['Accept-encoding']
 744             del req.headers['Youtubedl-no-compression']
 745         if 'Youtubedl-user-agent' in req.headers:
 746             if 'User-agent' in req.headers:
 747                 del req.headers['User-agent']
 748             req.headers['User-agent'] = req.headers['Youtubedl-user-agent']
 749             del req.headers['Youtubedl-user-agent']
 750         return req
 751
 752     def http_response(self, req, resp):
 753         old_resp = resp
 754         # gzip
 755         if resp.headers.get('Content-encoding', '') == 'gzip':
 756             content = resp.read()
 757             gz = gzip.GzipFile(fileobj=io.BytesIO(content), mode='rb')
 758             try:
 759                 uncompressed = io.BytesIO(gz.read())
 760             except IOError as original_ioerror:
 761                 # There may be junk add the end of the file
 762                 # See http://stackoverflow.com/q/4928560/35070 for details
 763                 for i in range(1, 1024):
 764                     try:
 765                         gz = gzip.GzipFile(fileobj=io.BytesIO(content[:-i]), mode='rb')
 766                         uncompressed = io.BytesIO(gz.read())
 767                     except IOError:
 768                         continue
 769                     break
 770                 else:
 771                     raise original_ioerror
 772             resp = self.addinfourl_wrapper(uncompressed, old_resp.headers, old_resp.url, old_resp.code)
 773             resp.msg = old_resp.msg
 774         # deflate
 775         if resp.headers.get('Content-encoding', '') == 'deflate':
 776             gz = io.BytesIO(self.deflate(resp.read()))
 777             resp = self.addinfourl_wrapper(gz, old_resp.headers, old_resp.url, old_resp.code)
 778             resp.msg = old_resp.msg
 779         return resp
 780
 781     https_request = http_request
 782     https_response = http_response
 783
 784
 785 def parse_iso8601(date_str, delimiter='T'):
 786     """ Return a UNIX timestamp from the given date """
 787
 788     if date_str is None:
 789         return None
 790
 791     m = re.search(
 792         r'Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$',
 793         date_str)
 794     if not m:
 795         timezone = datetime.timedelta()
 796     else:
 797         date_str = date_str[:-len(m.group(0))]
 798         if not m.group('sign'):
 799             timezone = datetime.timedelta()
 800         else:
 801             sign = 1 if m.group('sign') == '+' else -1
 802             timezone = datetime.timedelta(
 803                 hours=sign * int(m.group('hours')),
 804                 minutes=sign * int(m.group('minutes')))
 805     date_format =  '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
 806     dt = datetime.datetime.strptime(date_str, date_format) - timezone
 807     return calendar.timegm(dt.timetuple())
 808
 809
 810 def unified_strdate(date_str):
 811     """Return a string with the date in the format YYYYMMDD"""
 812
 813     if date_str is None:
 814         return None
 815
 816     upload_date = None
 817     #Replace commas
 818     date_str = date_str.replace(',', ' ')
 819     # %z (UTC offset) is only supported in python>=3.2
 820     date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
 821     format_expressions = [
 822         '%d %B %Y',
 823         '%d %b %Y',
 824         '%B %d %Y',
 825         '%b %d %Y',
 826         '%b %dst %Y %I:%M%p',
 827         '%b %dnd %Y %I:%M%p',
 828         '%b %dth %Y %I:%M%p',
 829         '%Y-%m-%d',
 830         '%d.%m.%Y',
 831         '%d/%m/%Y',
 832         '%Y/%m/%d %H:%M:%S',
 833         '%Y-%m-%d %H:%M:%S',
 834         '%d.%m.%Y %H:%M',
 835         '%d.%m.%Y %H.%M',
 836         '%Y-%m-%dT%H:%M:%SZ',
 837         '%Y-%m-%dT%H:%M:%S.%fZ',
 838         '%Y-%m-%dT%H:%M:%S.%f0Z',
 839         '%Y-%m-%dT%H:%M:%S',
 840         '%Y-%m-%dT%H:%M:%S.%f',
 841         '%Y-%m-%dT%H:%M',
 842     ]
 843     for expression in format_expressions:
 844         try:
 845             upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
 846         except ValueError:
 847             pass
 848     if upload_date is None:
 849         timetuple = email.utils.parsedate_tz(date_str)
 850         if timetuple:
 851             upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
 852     return upload_date
 853
 854 def determine_ext(url, default_ext=u'unknown_video'):
 855     if url is None:
 856         return default_ext
 857     guess = url.partition(u'?')[0].rpartition(u'.')[2]
 858     if re.match(r'^[A-Za-z0-9]+$', guess):
 859         return guess
 860     else:
 861         return default_ext
 862
 863 def subtitles_filename(filename, sub_lang, sub_format):
 864     return filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
 865
 866 def date_from_str(date_str):
 867     """
 868     Return a datetime object from a string in the format YYYYMMDD or
 869     (now|today)[+-][0-9](day|week|month|year)(s)?"""
 870     today = datetime.date.today()
 871     if date_str == 'now'or date_str == 'today':
 872         return today
 873     match = re.match('(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
 874     if match is not None:
 875         sign = match.group('sign')
 876         time = int(match.group('time'))
 877         if sign == '-':
 878             time = -time
 879         unit = match.group('unit')
 880         #A bad aproximation?
 881         if unit == 'month':
 882             unit = 'day'
 883             time *= 30
 884         elif unit == 'year':
 885             unit = 'day'
 886             time *= 365
 887         unit += 's'
 888         delta = datetime.timedelta(**{unit: time})
 889         return today + delta
 890     return datetime.datetime.strptime(date_str, "%Y%m%d").date()
 891
 892 def hyphenate_date(date_str):
 893     """
 894     Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
 895     match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str)
 896     if match is not None:
 897         return '-'.join(match.groups())
 898     else:
 899         return date_str
 900
 901 class DateRange(object):
 902     """Represents a time interval between two dates"""
 903     def __init__(self, start=None, end=None):
 904         """start and end must be strings in the format accepted by date"""
 905         if start is not None:
 906             self.start = date_from_str(start)
 907         else:
 908             self.start = datetime.datetime.min.date()
 909         if end is not None:
 910             self.end = date_from_str(end)
 911         else:
 912             self.end = datetime.datetime.max.date()
 913         if self.start > self.end:
 914             raise ValueError('Date range: "%s" , the start date must be before the end date' % self)
 915     @classmethod
 916     def day(cls, day):
 917         """Returns a range that only contains the given day"""
 918         return cls(day,day)
 919     def __contains__(self, date):
 920         """Check if the date is in the range"""
 921         if not isinstance(date, datetime.date):
 922             date = date_from_str(date)
 923         return self.start <= date <= self.end
 924     def __str__(self):
 925         return '%s - %s' % ( self.start.isoformat(), self.end.isoformat())
 926
 927
 928 def platform_name():
 929     """ Returns the platform name as a compat_str """
 930     res = platform.platform()
 931     if isinstance(res, bytes):
 932         res = res.decode(preferredencoding())
 933
 934     assert isinstance(res, compat_str)
 935     return res
 936
 937
 938 def _windows_write_string(s, out):
 939     """ Returns True if the string was written using special methods,
 940     False if it has yet to be written out."""
 941     # Adapted from http://stackoverflow.com/a/3259271/35070
 942
 943     import ctypes
 944     import ctypes.wintypes
 945
 946     WIN_OUTPUT_IDS = {
 947         1: -11,
 948         2: -12,
 949     }
 950
 951     try:
 952         fileno = out.fileno()
 953     except AttributeError:
 954         # If the output stream doesn't have a fileno, it's virtual
 955         return False
 956     if fileno not in WIN_OUTPUT_IDS:
 957         return False
 958
 959     GetStdHandle = ctypes.WINFUNCTYPE(
 960         ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
 961         ("GetStdHandle", ctypes.windll.kernel32))
 962     h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
 963
 964     WriteConsoleW = ctypes.WINFUNCTYPE(
 965         ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
 966         ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
 967         ctypes.wintypes.LPVOID)(("WriteConsoleW", ctypes.windll.kernel32))
 968     written = ctypes.wintypes.DWORD(0)
 969
 970     GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)(("GetFileType", ctypes.windll.kernel32))
 971     FILE_TYPE_CHAR = 0x0002
 972     FILE_TYPE_REMOTE = 0x8000
 973     GetConsoleMode = ctypes.WINFUNCTYPE(
 974         ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
 975         ctypes.POINTER(ctypes.wintypes.DWORD))(
 976         ("GetConsoleMode", ctypes.windll.kernel32))
 977     INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
 978
 979     def not_a_console(handle):
 980         if handle == INVALID_HANDLE_VALUE or handle is None:
 981             return True
 982         return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR
 983                 or GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0)
 984
 985     if not_a_console(h):
 986         return False
 987
 988     def next_nonbmp_pos(s):
 989         try:
 990             return next(i for i, c in enumerate(s) if ord(c) > 0xffff)
 991         except StopIteration:
 992             return len(s)
 993
 994     while s:
 995         count = min(next_nonbmp_pos(s), 1024)
 996
 997         ret = WriteConsoleW(
 998             h, s, count if count else 2, ctypes.byref(written), None)
 999         if ret == 0:
1000             raise OSError('Failed to write string')
1001         if not count:  # We just wrote a non-BMP character
1002             assert written.value == 2
1003             s = s[1:]
1004         else:
1005             assert written.value > 0
1006             s = s[written.value:]
1007     return True
1008
1009
1010 def write_string(s, out=None, encoding=None):
1011     if out is None:
1012         out = sys.stderr
1013     assert type(s) == compat_str
1014
1015     if sys.platform == 'win32' and encoding is None and hasattr(out, 'fileno'):
1016         if _windows_write_string(s, out):
1017             return
1018
1019     if ('b' in getattr(out, 'mode', '') or
1020             sys.version_info[0] < 3):  # Python 2 lies about mode of sys.stderr
1021         byt = s.encode(encoding or preferredencoding(), 'ignore')
1022         out.write(byt)
1023     elif hasattr(out, 'buffer'):
1024         enc = encoding or getattr(out, 'encoding', None) or preferredencoding()
1025         byt = s.encode(enc, 'ignore')
1026         out.buffer.write(byt)
1027     else:
1028         out.write(s)
1029     out.flush()
1030
1031
1032 def bytes_to_intlist(bs):
1033     if not bs:
1034         return []
1035     if isinstance(bs[0], int):  # Python 3
1036         return list(bs)
1037     else:
1038         return [ord(c) for c in bs]
1039
1040
1041 def intlist_to_bytes(xs):
1042     if not xs:
1043         return b''
1044     if isinstance(chr(0), bytes):  # Python 2
1045         return ''.join([chr(x) for x in xs])
1046     else:
1047         return bytes(xs)
1048
1049
1050 def get_cachedir(params={}):
1051     cache_root = os.environ.get('XDG_CACHE_HOME',
1052                                 os.path.expanduser('~/.cache'))
1053     return params.get('cachedir', os.path.join(cache_root, 'youtube-dl'))
1054
1055
1056 # Cross-platform file locking
1057 if sys.platform == 'win32':
1058     import ctypes.wintypes
1059     import msvcrt
1060
1061     class OVERLAPPED(ctypes.Structure):
1062         _fields_ = [
1063             ('Internal', ctypes.wintypes.LPVOID),
1064             ('InternalHigh', ctypes.wintypes.LPVOID),
1065             ('Offset', ctypes.wintypes.DWORD),
1066             ('OffsetHigh', ctypes.wintypes.DWORD),
1067             ('hEvent', ctypes.wintypes.HANDLE),
1068         ]
1069
1070     kernel32 = ctypes.windll.kernel32
1071     LockFileEx = kernel32.LockFileEx
1072     LockFileEx.argtypes = [
1073         ctypes.wintypes.HANDLE,     # hFile
1074         ctypes.wintypes.DWORD,      # dwFlags
1075         ctypes.wintypes.DWORD,      # dwReserved
1076         ctypes.wintypes.DWORD,      # nNumberOfBytesToLockLow
1077         ctypes.wintypes.DWORD,      # nNumberOfBytesToLockHigh
1078         ctypes.POINTER(OVERLAPPED)  # Overlapped
1079     ]
1080     LockFileEx.restype = ctypes.wintypes.BOOL
1081     UnlockFileEx = kernel32.UnlockFileEx
1082     UnlockFileEx.argtypes = [
1083         ctypes.wintypes.HANDLE,     # hFile
1084         ctypes.wintypes.DWORD,      # dwReserved
1085         ctypes.wintypes.DWORD,      # nNumberOfBytesToLockLow
1086         ctypes.wintypes.DWORD,      # nNumberOfBytesToLockHigh
1087         ctypes.POINTER(OVERLAPPED)  # Overlapped
1088     ]
1089     UnlockFileEx.restype = ctypes.wintypes.BOOL
1090     whole_low = 0xffffffff
1091     whole_high = 0x7fffffff
1092
1093     def _lock_file(f, exclusive):
1094         overlapped = OVERLAPPED()
1095         overlapped.Offset = 0
1096         overlapped.OffsetHigh = 0
1097         overlapped.hEvent = 0
1098         f._lock_file_overlapped_p = ctypes.pointer(overlapped)
1099         handle = msvcrt.get_osfhandle(f.fileno())
1100         if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0,
1101                           whole_low, whole_high, f._lock_file_overlapped_p):
1102             raise OSError('Locking file failed: %r' % ctypes.FormatError())
1103
1104     def _unlock_file(f):
1105         assert f._lock_file_overlapped_p
1106         handle = msvcrt.get_osfhandle(f.fileno())
1107         if not UnlockFileEx(handle, 0,
1108                             whole_low, whole_high, f._lock_file_overlapped_p):
1109             raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
1110
1111 else:
1112     import fcntl
1113
1114     def _lock_file(f, exclusive):
1115         fcntl.lockf(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
1116
1117     def _unlock_file(f):
1118         fcntl.lockf(f, fcntl.LOCK_UN)
1119
1120
1121 class locked_file(object):
1122     def __init__(self, filename, mode, encoding=None):
1123         assert mode in ['r', 'a', 'w']
1124         self.f = io.open(filename, mode, encoding=encoding)
1125         self.mode = mode
1126
1127     def __enter__(self):
1128         exclusive = self.mode != 'r'
1129         try:
1130             _lock_file(self.f, exclusive)
1131         except IOError:
1132             self.f.close()
1133             raise
1134         return self
1135
1136     def __exit__(self, etype, value, traceback):
1137         try:
1138             _unlock_file(self.f)
1139         finally:
1140             self.f.close()
1141
1142     def __iter__(self):
1143         return iter(self.f)
1144
1145     def write(self, *args):
1146         return self.f.write(*args)
1147
1148     def read(self, *args):
1149         return self.f.read(*args)
1150
1151
1152 def shell_quote(args):
1153     quoted_args = []
1154     encoding = sys.getfilesystemencoding()
1155     if encoding is None:
1156         encoding = 'utf-8'
1157     for a in args:
1158         if isinstance(a, bytes):
1159             # We may get a filename encoded with 'encodeFilename'
1160             a = a.decode(encoding)
1161         quoted_args.append(pipes.quote(a))
1162     return u' '.join(quoted_args)
1163
1164
1165 def takewhile_inclusive(pred, seq):
1166     """ Like itertools.takewhile, but include the latest evaluated element
1167         (the first element so that Not pred(e)) """
1168     for e in seq:
1169         yield e
1170         if not pred(e):
1171             return
1172
1173
1174 def smuggle_url(url, data):
1175     """ Pass additional data in a URL for internal use. """
1176
1177     sdata = compat_urllib_parse.urlencode(
1178         {u'__youtubedl_smuggle': json.dumps(data)})
1179     return url + u'#' + sdata
1180
1181
1182 def unsmuggle_url(smug_url, default=None):
1183     if not '#__youtubedl_smuggle' in smug_url:
1184         return smug_url, default
1185     url, _, sdata = smug_url.rpartition(u'#')
1186     jsond = compat_parse_qs(sdata)[u'__youtubedl_smuggle'][0]
1187     data = json.loads(jsond)
1188     return url, data
1189
1190
1191 def format_bytes(bytes):
1192     if bytes is None:
1193         return u'N/A'
1194     if type(bytes) is str:
1195         bytes = float(bytes)
1196     if bytes == 0.0:
1197         exponent = 0
1198     else:
1199         exponent = int(math.log(bytes, 1024.0))
1200     suffix = [u'B', u'KiB', u'MiB', u'GiB', u'TiB', u'PiB', u'EiB', u'ZiB', u'YiB'][exponent]
1201     converted = float(bytes) / float(1024 ** exponent)
1202     return u'%.2f%s' % (converted, suffix)
1203
1204
1205 def get_term_width():
1206     columns = os.environ.get('COLUMNS', None)
1207     if columns:
1208         return int(columns)
1209
1210     try:
1211         sp = subprocess.Popen(
1212             ['stty', 'size'],
1213             stdout=subprocess.PIPE, stderr=subprocess.PIPE)
1214         out, err = sp.communicate()
1215         return int(out.split()[1])
1216     except:
1217         pass
1218     return None
1219
1220
1221 def month_by_name(name):
1222     """ Return the number of a month by (locale-independently) English name """
1223
1224     ENGLISH_NAMES = [
1225         u'January', u'February', u'March', u'April', u'May', u'June',
1226         u'July', u'August', u'September', u'October', u'November', u'December']
1227     try:
1228         return ENGLISH_NAMES.index(name) + 1
1229     except ValueError:
1230         return None
1231
1232
1233 def fix_xml_ampersands(xml_str):
1234     """Replace all the '&' by '&amp;' in XML"""
1235     return re.sub(
1236         r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1237         u'&amp;',
1238         xml_str)
1239
1240
1241 def setproctitle(title):
1242     assert isinstance(title, compat_str)
1243     try:
1244         libc = ctypes.cdll.LoadLibrary("libc.so.6")
1245     except OSError:
1246         return
1247     title_bytes = title.encode('utf-8')
1248     buf = ctypes.create_string_buffer(len(title_bytes))
1249     buf.value = title_bytes
1250     try:
1251         libc.prctl(15, buf, 0, 0, 0)
1252     except AttributeError:
1253         return  # Strange libc, just skip this
1254
1255
1256 def remove_start(s, start):
1257     if s.startswith(start):
1258         return s[len(start):]
1259     return s
1260
1261
1262 def url_basename(url):
1263     path = compat_urlparse.urlparse(url).path
1264     return path.strip(u'/').split(u'/')[-1]
1265
1266
1267 class HEADRequest(compat_urllib_request.Request):
1268     def get_method(self):
1269         return "HEAD"
1270
1271
1272 def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
1273     if get_attr:
1274         if v is not None:
1275             v = getattr(v, get_attr, None)
1276     return default if v is None else (int(v) * invscale // scale)
1277
1278
1279 def str_to_int(int_str):
1280     if int_str is None:
1281         return None
1282     int_str = re.sub(r'[,\.]', u'', int_str)
1283     return int(int_str)
1284
1285
1286 def float_or_none(v, scale=1, invscale=1, default=None):
1287     return default if v is None else (float(v) * invscale / scale)
1288
1289
1290 def parse_duration(s):
1291     if s is None:
1292         return None
1293
1294     m = re.match(
1295         r'(?:(?:(?P<hours>[0-9]+)[:h])?(?P<mins>[0-9]+)[:m])?(?P<secs>[0-9]+)s?(?::[0-9]+)?$', s)
1296     if not m:
1297         return None
1298     res = int(m.group('secs'))
1299     if m.group('mins'):
1300         res += int(m.group('mins')) * 60
1301         if m.group('hours'):
1302             res += int(m.group('hours')) * 60 * 60
1303     return res
1304
1305
1306 def prepend_extension(filename, ext):
1307     name, real_ext = os.path.splitext(filename)
1308     return u'{0}.{1}{2}'.format(name, ext, real_ext)
1309
1310
1311 def check_executable(exe, args=[]):
1312     """ Checks if the given binary is installed somewhere in PATH, and returns its name.
1313     args can be a list of arguments for a short output (like -version) """
1314     try:
1315         subprocess.Popen([exe] + args, stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
1316     except OSError:
1317         return False
1318     return exe
1319
1320
1321 class PagedList(object):
1322     def __init__(self, pagefunc, pagesize):
1323         self._pagefunc = pagefunc
1324         self._pagesize = pagesize
1325
1326     def __len__(self):
1327         # This is only useful for tests
1328         return len(self.getslice())
1329
1330     def getslice(self, start=0, end=None):
1331         res = []
1332         for pagenum in itertools.count(start // self._pagesize):
1333             firstid = pagenum * self._pagesize
1334             nextfirstid = pagenum * self._pagesize + self._pagesize
1335             if start >= nextfirstid:
1336                 continue
1337
1338             page_results = list(self._pagefunc(pagenum))
1339
1340             startv = (
1341                 start % self._pagesize
1342                 if firstid <= start < nextfirstid
1343                 else 0)
1344
1345             endv = (
1346                 ((end - 1) % self._pagesize) + 1
1347                 if (end is not None and firstid <= end <= nextfirstid)
1348                 else None)
1349
1350             if startv != 0 or endv is not None:
1351                 page_results = page_results[startv:endv]
1352             res.extend(page_results)
1353
1354             # A little optimization - if current page is not "full", ie. does
1355             # not contain page_size videos then we can assume that this page
1356             # is the last one - there are no more ids on further pages -
1357             # i.e. no need to query again.
1358             if len(page_results) + startv < self._pagesize:
1359                 break
1360
1361             # If we got the whole page, but the next page is not interesting,
1362             # break out early as well
1363             if end == nextfirstid:
1364                 break
1365         return res
1366
1367
1368 def uppercase_escape(s):
1369     unicode_escape = codecs.getdecoder('unicode_escape')
1370     return re.sub(
1371         r'\\U[0-9a-fA-F]{8}',
1372         lambda m: unicode_escape(m.group(0))[0],
1373         s)
1374
1375 try:
1376     struct.pack(u'!I', 0)
1377 except TypeError:
1378     # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
1379     def struct_pack(spec, *args):
1380         if isinstance(spec, compat_str):
1381             spec = spec.encode('ascii')
1382         return struct.pack(spec, *args)
1383
1384     def struct_unpack(spec, *args):
1385         if isinstance(spec, compat_str):
1386             spec = spec.encode('ascii')
1387         return struct.unpack(spec, *args)
1388 else:
1389     struct_pack = struct.pack
1390     struct_unpack = struct.unpack
1391
1392
1393 def read_batch_urls(batch_fd):
1394     def fixup(url):
1395         if not isinstance(url, compat_str):
1396             url = url.decode('utf-8', 'replace')
1397         BOM_UTF8 = u'\xef\xbb\xbf'
1398         if url.startswith(BOM_UTF8):
1399             url = url[len(BOM_UTF8):]
1400         url = url.strip()
1401         if url.startswith(('#', ';', ']')):
1402             return False
1403         return url
1404
1405     with contextlib.closing(batch_fd) as fd:
1406         return [url for url in map(fixup, fd) if url]
1407
1408
1409 def urlencode_postdata(*args, **kargs):
1410     return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
1411
1412
1413 def parse_xml(s):
1414     class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
1415         def doctype(self, name, pubid, system):
1416             pass  # Ignore doctypes
1417
1418     parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
1419     kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
1420     return xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
1421
1422
1423 if sys.version_info < (3, 0) and sys.platform == 'win32':
1424     def compat_getpass(prompt, *args, **kwargs):
1425         if isinstance(prompt, compat_str):
1426             prompt = prompt.encode(preferredencoding())
1427         return getpass.getpass(prompt, *args, **kwargs)
1428 else:
1429     compat_getpass = getpass.getpass
1430
1431
1432 US_RATINGS = {
1433     'G': 0,
1434     'PG': 10,
1435     'PG-13': 13,
1436     'R': 16,
1437     'NC': 18,
1438 }
1439
1440
1441 def strip_jsonp(code):
1442     return re.sub(r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?\s*$', r'\1', code)
1443
1444
1445 def qualities(quality_ids):
1446     """ Get a numeric quality value out of a list of possible values """
1447     def q(qid):
1448         try:
1449             return quality_ids.index(qid)
1450         except ValueError:
1451             return -1
1452     return q
1453
1454
1455 DEFAULT_OUTTMPL = '%(title)s-%(id)s.%(ext)s'
1456
1457 try:
1458     subprocess_check_output = subprocess.check_output
1459 except AttributeError:
1460     def subprocess_check_output(*args, **kwargs):
1461         assert 'input' not in kwargs
1462         p = subprocess.Popen(*args, stdout=subprocess.PIPE, **kwargs)
1463         output, _ = p.communicate()
1464         ret = p.poll()
1465         if ret:
1466             raise subprocess.CalledProcessError(ret, p.args, output=output)
1467         return output