]> git.bitcoin.ninja Git - youtube-dl/blob - youtube-dl
4f767cf3761b73a855280466a2e9e086af1e70dd
[youtube-dl] / youtube-dl
1 #!/usr/bin/env python
2 # -*- coding: utf-8 -*-
3 # Author: Ricardo Garcia Gonzalez
4 # Author: Danny Colligan
5 # License: Public domain code
6 import htmlentitydefs
7 import httplib
8 import locale
9 import math
10 import netrc
11 import os
12 import os.path
13 import re
14 import socket
15 import string
16 import sys
17 import time
18 import urllib
19 import urllib2
20
21 std_headers = {
22         'User-Agent': 'Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.0.5) Gecko/2008120122 Firefox/3.0.5',
23         'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
24         'Accept': 'text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5',
25         'Accept-Language': 'en-us,en;q=0.5',
26 }
27
28 simple_title_chars = string.ascii_letters.decode('ascii') + string.digits.decode('ascii')
29
30 class DownloadError(Exception):
31         """Download Error exception.
32         
33         This exception may be thrown by FileDownloader objects if they are not
34         configured to continue on errors. They will contain the appropriate
35         error message.
36         """
37         pass
38
39 class SameFileError(Exception):
40         """Same File exception.
41
42         This exception will be thrown by FileDownloader objects if they detect
43         multiple files would have to be downloaded to the same file on disk.
44         """
45         pass
46
47 class PostProcessingError(Exception):
48         """Post Processing exception.
49
50         This exception may be raised by PostProcessor's .run() method to
51         indicate an error in the postprocessing task.
52         """
53         pass
54
55 class FileDownloader(object):
56         """File Downloader class.
57
58         File downloader objects are the ones responsible of downloading the
59         actual video file and writing it to disk if the user has requested
60         it, among some other tasks. In most cases there should be one per
61         program. As, given a video URL, the downloader doesn't know how to
62         extract all the needed information, task that InfoExtractors do, it
63         has to pass the URL to one of them.
64
65         For this, file downloader objects have a method that allows
66         InfoExtractors to be registered in a given order. When it is passed
67         a URL, the file downloader handles it to the first InfoExtractor it
68         finds that reports being able to handle it. The InfoExtractor returns
69         all the information to the FileDownloader and the latter downloads the
70         file or does whatever it's instructed to do.
71
72         File downloaders accept a lot of parameters. In order not to saturate
73         the object constructor with arguments, it receives a dictionary of
74         options instead. These options are available through the params
75         attribute for the InfoExtractors to use. The FileDownloader also
76         registers itself as the downloader in charge for the InfoExtractors
77         that are added to it, so this is a "mutual registration".
78
79         Available options:
80
81         username:       Username for authentication purposes.
82         password:       Password for authentication purposes.
83         usenetrc:       Use netrc for authentication instead.
84         quiet:          Do not print messages to stdout.
85         forceurl:       Force printing final URL.
86         forcetitle:     Force printing title.
87         simulate:       Do not download the video files.
88         format:         Video format code.
89         outtmpl:        Template for output names.
90         ignoreerrors:   Do not stop on download errors.
91         ratelimit:      Download speed limit, in bytes/sec.
92         nooverwrites:   Prevent overwriting files.
93         """
94
95         params = None
96         _ies = []
97         _pps = []
98
99         def __init__(self, params):
100                 """Create a FileDownloader object with the given options."""
101                 self._ies = []
102                 self._pps = []
103                 self.params = params
104         
105         @staticmethod
106         def pmkdir(filename):
107                 """Create directory components in filename. Similar to Unix "mkdir -p"."""
108                 components = filename.split(os.sep)
109                 aggregate = [os.sep.join(components[0:x]) for x in xrange(1, len(components))]
110                 aggregate = ['%s%s' % (x, os.sep) for x in aggregate] # Finish names with separator
111                 for dir in aggregate:
112                         if not os.path.exists(dir):
113                                 os.mkdir(dir)
114         
115         @staticmethod
116         def format_bytes(bytes):
117                 if bytes is None:
118                         return 'N/A'
119                 if bytes == 0:
120                         exponent = 0
121                 else:
122                         exponent = long(math.log(float(bytes), 1024.0))
123                 suffix = 'bkMGTPEZY'[exponent]
124                 converted = float(bytes) / float(1024**exponent)
125                 return '%.2f%s' % (converted, suffix)
126
127         @staticmethod
128         def calc_percent(byte_counter, data_len):
129                 if data_len is None:
130                         return '---.-%'
131                 return '%6s' % ('%3.1f%%' % (float(byte_counter) / float(data_len) * 100.0))
132
133         @staticmethod
134         def calc_eta(start, now, total, current):
135                 if total is None:
136                         return '--:--'
137                 dif = now - start
138                 if current == 0 or dif < 0.001: # One millisecond
139                         return '--:--'
140                 rate = float(current) / dif
141                 eta = long((float(total) - float(current)) / rate)
142                 (eta_mins, eta_secs) = divmod(eta, 60)
143                 if eta_mins > 99:
144                         return '--:--'
145                 return '%02d:%02d' % (eta_mins, eta_secs)
146
147         @staticmethod
148         def calc_speed(start, now, bytes):
149                 dif = now - start
150                 if bytes == 0 or dif < 0.001: # One millisecond
151                         return '%10s' % '---b/s'
152                 return '%10s' % ('%s/s' % FileDownloader.format_bytes(float(bytes) / dif))
153
154         @staticmethod
155         def best_block_size(elapsed_time, bytes):
156                 new_min = max(bytes / 2.0, 1.0)
157                 new_max = min(max(bytes * 2.0, 1.0), 4194304) # Do not surpass 4 MB
158                 if elapsed_time < 0.001:
159                         return int(new_max)
160                 rate = bytes / elapsed_time
161                 if rate > new_max:
162                         return int(new_max)
163                 if rate < new_min:
164                         return int(new_min)
165                 return int(rate)
166
167         @staticmethod
168         def parse_bytes(bytestr):
169                 """Parse a string indicating a byte quantity into a long integer."""
170                 matchobj = re.match(r'(?i)^(\d+(?:\.\d+)?)([kMGTPEZY]?)$', bytestr)
171                 if matchobj is None:
172                         return None
173                 number = float(matchobj.group(1))
174                 multiplier = 1024.0 ** 'bkmgtpezy'.index(matchobj.group(2).lower())
175                 return long(round(number * multiplier))
176
177         def add_info_extractor(self, ie):
178                 """Add an InfoExtractor object to the end of the list."""
179                 self._ies.append(ie)
180                 ie.set_downloader(self)
181         
182         def add_post_processor(self, pp):
183                 """Add a PostProcessor object to the end of the chain."""
184                 self._pps.append(pp)
185                 pp.set_downloader(self)
186         
187         def to_stdout(self, message, skip_eol=False):
188                 """Print message to stdout if not in quiet mode."""
189                 if not self.params.get('quiet', False):
190                         print u'%s%s' % (message, [u'\n', u''][skip_eol]),
191                         sys.stdout.flush()
192         
193         def to_stderr(self, message):
194                 """Print message to stderr."""
195                 print >>sys.stderr, message
196         
197         def fixed_template(self):
198                 """Checks if the output template is fixed."""
199                 return (re.search(ur'(?u)%\(.+?\)s', self.params['outtmpl']) is None)
200
201         def trouble(self, message=None):
202                 """Determine action to take when a download problem appears.
203
204                 Depending on if the downloader has been configured to ignore
205                 download errors or not, this method may throw an exception or
206                 not when errors are found, after printing the message. If it
207                 doesn't raise, it returns an error code suitable to be returned
208                 later as a program exit code to indicate error.
209                 """
210                 if message is not None:
211                         self.to_stderr(message)
212                 if not self.params.get('ignoreerrors', False):
213                         raise DownloadError(message)
214                 return 1
215
216         def slow_down(self, start_time, byte_counter):
217                 """Sleep if the download speed is over the rate limit."""
218                 rate_limit = self.params.get('ratelimit', None)
219                 if rate_limit is None or byte_counter == 0:
220                         return
221                 now = time.time()
222                 elapsed = now - start_time
223                 if elapsed <= 0.0:
224                         return
225                 speed = float(byte_counter) / elapsed
226                 if speed > rate_limit:
227                         time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
228
229         def report_destination(self, filename):
230                 """Report destination filename."""
231                 self.to_stdout(u'[download] Destination: %s' % filename)
232         
233         def report_progress(self, percent_str, data_len_str, speed_str, eta_str):
234                 """Report download progress."""
235                 self.to_stdout(u'\r[download] %s of %s at %s ETA %s' %
236                                 (percent_str, data_len_str, speed_str, eta_str), skip_eol=True)
237         
238         def report_finish(self):
239                 """Report download finished."""
240                 self.to_stdout(u'')
241
242         def download(self, url_list):
243                 """Download a given list of URLs."""
244                 retcode = 0
245                 if len(url_list) > 1 and self.fixed_template():
246                         raise SameFileError(self.params['outtmpl'])
247
248                 for url in url_list:
249                         suitable_found = False
250                         for ie in self._ies:
251                                 if not ie.suitable(url):
252                                         continue
253                                 # Suitable InfoExtractor found
254                                 suitable_found = True
255                                 all_results = ie.extract(url)
256                                 results = [x for x in all_results if x is not None]
257                                 if len(results) != len(all_results):
258                                         retcode = self.trouble()
259
260                                 if len(results) > 1 and self.fixed_template():
261                                         raise SameFileError(self.params['outtmpl'])
262
263                                 for result in results:
264                                         # Forced printings
265                                         if self.params.get('forcetitle', False):
266                                                 print result['title']
267                                         if self.params.get('forceurl', False):
268                                                 print result['url']
269                                                 
270                                         # Do nothing else if in simulate mode
271                                         if self.params.get('simulate', False):
272                                                 continue
273
274                                         try:
275                                                 filename = self.params['outtmpl'] % result
276                                                 self.report_destination(filename)
277                                         except (ValueError, KeyError), err:
278                                                 retcode = self.trouble('ERROR: invalid output template or system charset: %s' % str(err))
279                                                 continue
280                                         if self.params['nooverwrites'] and os.path.exists(filename):
281                                                 self.to_stderr('WARNING: file exists: %s; skipping' % filename)
282                                                 continue
283                                         try:
284                                                 self.pmkdir(filename)
285                                         except (OSError, IOError), err:
286                                                 retcode = self.trouble('ERROR: unable to create directories: %s' % str(err))
287                                                 continue
288                                         try:
289                                                 outstream = open(filename, 'wb')
290                                         except (OSError, IOError), err:
291                                                 retcode = self.trouble('ERROR: unable to open for writing: %s' % str(err))
292                                                 continue
293                                         try:
294                                                 self._do_download(outstream, result['url'])
295                                                 outstream.close()
296                                         except (OSError, IOError), err:
297                                                 retcode = self.trouble('ERROR: unable to write video data: %s' % str(err))
298                                                 continue
299                                         except (urllib2.URLError, httplib.HTTPException, socket.error), err:
300                                                 retcode = self.trouble('ERROR: unable to download video data: %s' % str(err))
301                                                 continue
302                                         try:
303                                                 self.post_process(filename, result)
304                                         except (PostProcessingError), err:
305                                                 retcode = self.trouble('ERROR: postprocessing: %s' % str(err))
306                                                 continue
307
308                                 break
309                         if not suitable_found:
310                                 retcode = self.trouble('ERROR: no suitable InfoExtractor: %s' % url)
311
312                 return retcode
313
314         def post_process(self, filename, ie_info):
315                 """Run the postprocessing chain on the given file."""
316                 info = dict(ie_info)
317                 info['filepath'] = filename
318                 for pp in self._pps:
319                         info = pp.run(info)
320                         if info is None:
321                                 break
322         
323         def _do_download(self, stream, url):
324                 request = urllib2.Request(url, None, std_headers)
325                 data = urllib2.urlopen(request)
326                 data_len = data.info().get('Content-length', None)
327                 data_len_str = self.format_bytes(data_len)
328                 byte_counter = 0
329                 block_size = 1024
330                 start = time.time()
331                 while True:
332                         # Progress message
333                         percent_str = self.calc_percent(byte_counter, data_len)
334                         eta_str = self.calc_eta(start, time.time(), data_len, byte_counter)
335                         speed_str = self.calc_speed(start, time.time(), byte_counter)
336                         self.report_progress(percent_str, data_len_str, speed_str, eta_str)
337
338                         # Download and write
339                         before = time.time()
340                         data_block = data.read(block_size)
341                         after = time.time()
342                         data_block_len = len(data_block)
343                         if data_block_len == 0:
344                                 break
345                         byte_counter += data_block_len
346                         stream.write(data_block)
347                         block_size = self.best_block_size(after - before, data_block_len)
348
349                         # Apply rate limit
350                         self.slow_down(start, byte_counter)
351
352                 self.report_finish()
353                 if data_len is not None and str(byte_counter) != data_len:
354                         raise ValueError('Content too short: %s/%s bytes' % (byte_counter, data_len))
355
356 class InfoExtractor(object):
357         """Information Extractor class.
358
359         Information extractors are the classes that, given a URL, extract
360         information from the video (or videos) the URL refers to. This
361         information includes the real video URL, the video title and simplified
362         title, author and others. It is returned in a list of dictionaries when
363         calling its extract() method. It is a list because a URL can refer to
364         more than one video (think of playlists). The dictionaries must include
365         the following fields:
366
367         id:             Video identifier.
368         url:            Final video URL.
369         uploader:       Nickname of the video uploader.
370         title:          Literal title.
371         stitle:         Simplified title.
372         ext:            Video filename extension.
373
374         Subclasses of this one should re-define the _real_initialize() and
375         _real_extract() methods, as well as the suitable() static method.
376         Probably, they should also be instantiated and added to the main
377         downloader.
378         """
379
380         _ready = False
381         _downloader = None
382
383         def __init__(self, downloader=None):
384                 """Constructor. Receives an optional downloader."""
385                 self._ready = False
386                 self.set_downloader(downloader)
387
388         @staticmethod
389         def suitable(url):
390                 """Receives a URL and returns True if suitable for this IE."""
391                 return False
392
393         def initialize(self):
394                 """Initializes an instance (authentication, etc)."""
395                 if not self._ready:
396                         self._real_initialize()
397                         self._ready = True
398
399         def extract(self, url):
400                 """Extracts URL information and returns it in list of dicts."""
401                 self.initialize()
402                 return self._real_extract(url)
403
404         def set_downloader(self, downloader):
405                 """Sets the downloader for this IE."""
406                 self._downloader = downloader
407         
408         def to_stdout(self, message):
409                 """Print message to stdout if downloader is not in quiet mode."""
410                 if self._downloader is None or not self._downloader.params.get('quiet', False):
411                         print message
412         
413         def to_stderr(self, message):
414                 """Print message to stderr."""
415                 print >>sys.stderr, message
416
417         def _real_initialize(self):
418                 """Real initialization process. Redefine in subclasses."""
419                 pass
420
421         def _real_extract(self, url):
422                 """Real extraction process. Redefine in subclasses."""
423                 pass
424
425 class YoutubeIE(InfoExtractor):
426         """Information extractor for youtube.com."""
427
428         _VALID_URL = r'^((?:http://)?(?:\w+\.)?youtube\.com/(?:(?:v/)|(?:(?:watch(?:\.php)?)?\?(?:.+&)?v=)))?([0-9A-Za-z_-]+)(?(1).+)?$'
429         _LANG_URL = r'http://uk.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
430         _LOGIN_URL = 'http://www.youtube.com/signup?next=/&gl=US&hl=en'
431         _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
432         _NETRC_MACHINE = 'youtube'
433
434         @staticmethod
435         def suitable(url):
436                 return (re.match(YoutubeIE._VALID_URL, url) is not None)
437
438         @staticmethod
439         def htmlentity_transform(matchobj):
440                 """Transforms an HTML entity to a Unicode character."""
441                 entity = matchobj.group(1)
442
443                 # Known non-numeric HTML entity
444                 if entity in htmlentitydefs.name2codepoint:
445                         return unichr(htmlentitydefs.name2codepoint[entity])
446
447                 # Unicode character
448                 mobj = re.match(ur'(?u)#(x?\d+)', entity)
449                 if mobj is not None:
450                         numstr = mobj.group(1)
451                         if numstr.startswith(u'x'):
452                                 base = 16
453                                 numstr = u'0%s' % numstr
454                         else:
455                                 base = 10
456                         return unichr(long(numstr, base))
457
458                 # Unknown entity in name, return its literal representation
459                 return (u'&%s;' % entity)
460
461         def report_lang(self):
462                 """Report attempt to set language."""
463                 self.to_stdout(u'[youtube] Setting language')
464
465         def report_login(self):
466                 """Report attempt to log in."""
467                 self.to_stdout(u'[youtube] Logging in')
468         
469         def report_age_confirmation(self):
470                 """Report attempt to confirm age."""
471                 self.to_stdout(u'[youtube] Confirming age')
472         
473         def report_webpage_download(self, video_id):
474                 """Report attempt to download webpage."""
475                 self.to_stdout(u'[youtube] %s: Downloading video webpage' % video_id)
476         
477         def report_information_extraction(self, video_id):
478                 """Report attempt to extract video information."""
479                 self.to_stdout(u'[youtube] %s: Extracting video information' % video_id)
480         
481         def report_video_url(self, video_id, video_real_url):
482                 """Report extracted video URL."""
483                 self.to_stdout(u'[youtube] %s: URL: %s' % (video_id, video_real_url))
484         
485         def _real_initialize(self):
486                 if self._downloader is None:
487                         return
488
489                 username = None
490                 password = None
491                 downloader_params = self._downloader.params
492
493                 # Attempt to use provided username and password or .netrc data
494                 if downloader_params.get('username', None) is not None:
495                         username = downloader_params['username']
496                         password = downloader_params['password']
497                 elif downloader_params.get('usenetrc', False):
498                         try:
499                                 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
500                                 if info is not None:
501                                         username = info[0]
502                                         password = info[2]
503                                 else:
504                                         raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
505                         except (IOError, netrc.NetrcParseError), err:
506                                 self.to_stderr(u'WARNING: parsing .netrc: %s' % str(err))
507                                 return
508
509                 # Set language
510                 request = urllib2.Request(self._LANG_URL, None, std_headers)
511                 try:
512                         self.report_lang()
513                         urllib2.urlopen(request).read()
514                 except (urllib2.URLError, httplib.HTTPException, socket.error), err:
515                         self.to_stderr(u'WARNING: unable to set language: %s' % str(err))
516                         return
517
518                 # No authentication to be performed
519                 if username is None:
520                         return
521
522                 # Log in
523                 login_form = {
524                                 'current_form': 'loginForm',
525                                 'next':         '/',
526                                 'action_login': 'Log In',
527                                 'username':     username,
528                                 'password':     password,
529                                 }
530                 request = urllib2.Request(self._LOGIN_URL, urllib.urlencode(login_form), std_headers)
531                 try:
532                         self.report_login()
533                         login_results = urllib2.urlopen(request).read()
534                         if re.search(r'(?i)<form[^>]* name="loginForm"', login_results) is not None:
535                                 self.to_stderr(u'WARNING: unable to log in: bad username or password')
536                                 return
537                 except (urllib2.URLError, httplib.HTTPException, socket.error), err:
538                         self.to_stderr(u'WARNING: unable to log in: %s' % str(err))
539                         return
540         
541                 # Confirm age
542                 age_form = {
543                                 'next_url':             '/',
544                                 'action_confirm':       'Confirm',
545                                 }
546                 request = urllib2.Request(self._AGE_URL, urllib.urlencode(age_form), std_headers)
547                 try:
548                         self.report_age_confirmation()
549                         age_results = urllib2.urlopen(request).read()
550                 except (urllib2.URLError, httplib.HTTPException, socket.error), err:
551                         self.to_stderr(u'ERROR: unable to confirm age: %s' % str(err))
552                         return
553
554         def _real_extract(self, url):
555                 # Extract video id from URL
556                 mobj = re.match(self._VALID_URL, url)
557                 if mobj is None:
558                         self.to_stderr(u'ERROR: invalid URL: %s' % url)
559                         return [None]
560                 video_id = mobj.group(2)
561
562                 # Downloader parameters
563                 format_param = None
564                 if self._downloader is not None:
565                         params = self._downloader.params
566                         format_param = params.get('format', None)
567
568                 # Extension
569                 video_extension = {
570                         '17': '3gp',
571                         '18': 'mp4',
572                         '22': 'mp4',
573                 }.get(format_param, 'flv')
574
575                 # Normalize URL, including format
576                 normalized_url = 'http://www.youtube.com/watch?v=%s&gl=US&hl=en' % video_id
577                 if format_param is not None:
578                         normalized_url = '%s&fmt=%s' % (normalized_url, format_param)
579                 request = urllib2.Request(normalized_url, None, std_headers)
580                 try:
581                         self.report_webpage_download(video_id)
582                         video_webpage = urllib2.urlopen(request).read()
583                 except (urllib2.URLError, httplib.HTTPException, socket.error), err:
584                         self.to_stderr(u'ERROR: unable to download video webpage: %s' % str(err))
585                         return [None]
586                 self.report_information_extraction(video_id)
587                 
588                 # "t" param
589                 mobj = re.search(r', "t": "([^"]+)"', video_webpage)
590                 if mobj is None:
591                         self.to_stderr(u'ERROR: unable to extract "t" parameter')
592                         return [None]
593                 video_real_url = 'http://www.youtube.com/get_video?video_id=%s&t=%s' % (video_id, mobj.group(1))
594                 if format_param is not None:
595                         video_real_url = '%s&fmt=%s' % (video_real_url, format_param)
596                 self.report_video_url(video_id, video_real_url)
597
598                 # uploader
599                 mobj = re.search(r"var watchUsername = '([^']+)';", video_webpage)
600                 if mobj is None:
601                         self.to_stderr(u'ERROR: unable to extract uploader nickname')
602                         return [None]
603                 video_uploader = mobj.group(1)
604
605                 # title
606                 mobj = re.search(r'(?im)<title>YouTube - ([^<]*)</title>', video_webpage)
607                 if mobj is None:
608                         self.to_stderr(u'ERROR: unable to extract video title')
609                         return [None]
610                 video_title = mobj.group(1).decode('utf-8')
611                 video_title = re.sub(ur'(?u)&(.+?);', self.htmlentity_transform, video_title)
612                 video_title = video_title.replace(os.sep, u'%')
613
614                 # simplified title
615                 simple_title = re.sub(ur'(?u)([^%s]+)' % simple_title_chars, ur'_', video_title)
616                 simple_title = simple_title.strip(ur'_')
617
618                 # Return information
619                 return [{
620                         'id':           video_id.decode('utf-8'),
621                         'url':          video_real_url.decode('utf-8'),
622                         'uploader':     video_uploader.decode('utf-8'),
623                         'title':        video_title,
624                         'stitle':       simple_title,
625                         'ext':          video_extension.decode('utf-8'),
626                         }]
627
628 class MetacafeIE(InfoExtractor):
629         """Information Extractor for metacafe.com."""
630
631         _VALID_URL = r'(?:http://)?(?:www\.)?metacafe\.com/watch/([^/]+)/([^/]+)/.*'
632         _DISCLAIMER = 'http://www.metacafe.com/family_filter/'
633         _youtube_ie = None
634
635         def __init__(self, youtube_ie, downloader=None):
636                 InfoExtractor.__init__(self, downloader)
637                 self._youtube_ie = youtube_ie
638
639         @staticmethod
640         def suitable(url):
641                 return (re.match(MetacafeIE._VALID_URL, url) is not None)
642
643         def report_disclaimer(self):
644                 """Report disclaimer retrieval."""
645                 self.to_stdout(u'[metacafe] Retrieving disclaimer')
646
647         def report_age_confirmation(self):
648                 """Report attempt to confirm age."""
649                 self.to_stdout(u'[metacafe] Confirming age')
650         
651         def report_download_webpage(self, video_id):
652                 """Report webpage download."""
653                 self.to_stdout(u'[metacafe] %s: Downloading webpage' % video_id)
654         
655         def report_extraction(self, video_id):
656                 """Report information extraction."""
657                 self.to_stdout(u'[metacafe] %s: Extracting information' % video_id)
658
659         def _real_initialize(self):
660                 # Retrieve disclaimer
661                 request = urllib2.Request(self._DISCLAIMER, None, std_headers)
662                 try:
663                         self.report_disclaimer()
664                         disclaimer = urllib2.urlopen(request).read()
665                 except (urllib2.URLError, httplib.HTTPException, socket.error), err:
666                         self.to_stderr(u'ERROR: unable to retrieve disclaimer: %s' % str(err))
667                         return
668
669                 # Confirm age
670                 disclaimer_form = {
671                         'filters': '0',
672                         'submit': "Continue - I'm over 18",
673                         }
674                 request = urllib2.Request('http://www.metacafe.com/', urllib.urlencode(disclaimer_form), std_headers)
675                 try:
676                         self.report_age_confirmation()
677                         disclaimer = urllib2.urlopen(request).read()
678                 except (urllib2.URLError, httplib.HTTPException, socket.error), err:
679                         self.to_stderr(u'ERROR: unable to confirm age: %s' % str(err))
680                         return
681         
682         def _real_extract(self, url):
683                 # Extract id and simplified title from URL
684                 mobj = re.match(self._VALID_URL, url)
685                 if mobj is None:
686                         self.to_stderr(u'ERROR: invalid URL: %s' % url)
687                         return [None]
688
689                 video_id = mobj.group(1)
690
691                 # Check if video comes from YouTube
692                 mobj2 = re.match(r'^yt-(.*)$', video_id)
693                 if mobj2 is not None:
694                         return self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % mobj2.group(1))
695
696                 simple_title = mobj.group(2).decode('utf-8')
697                 video_extension = 'flv'
698
699                 # Retrieve video webpage to extract further information
700                 request = urllib2.Request('http://www.metacafe.com/watch/%s/' % video_id)
701                 try:
702                         self.report_download_webpage(video_id)
703                         webpage = urllib2.urlopen(request).read()
704                 except (urllib2.URLError, httplib.HTTPException, socket.error), err:
705                         self.to_stderr(u'ERROR: unable retrieve video webpage: %s' % str(err))
706                         return [None]
707
708                 # Extract URL, uploader and title from webpage
709                 self.report_extraction(video_id)
710                 mobj = re.search(r'(?m)"mediaURL":"(http.*?\.flv)"', webpage)
711                 if mobj is None:
712                         self.to_stderr(u'ERROR: unable to extract media URL')
713                         return [None]
714                 mediaURL = mobj.group(1).replace('\\', '')
715
716                 mobj = re.search(r'(?m)"gdaKey":"(.*?)"', webpage)
717                 if mobj is None:
718                         self.to_stderr(u'ERROR: unable to extract gdaKey')
719                         return [None]
720                 gdaKey = mobj.group(1)
721
722                 video_url = '%s?__gda__=%s' % (mediaURL, gdaKey)
723
724                 mobj = re.search(r'(?im)<title>(.*) - Video</title>', webpage)
725                 if mobj is None:
726                         self.to_stderr(u'ERROR: unable to extract title')
727                         return [None]
728                 video_title = mobj.group(1).decode('utf-8')
729
730                 mobj = re.search(r'(?m)<li id="ChnlUsr">.*?Submitter:<br />(.*?)</li>', webpage)
731                 if mobj is None:
732                         self.to_stderr(u'ERROR: unable to extract uploader nickname')
733                         return [None]
734                 video_uploader = re.sub(r'<.*?>', '', mobj.group(1))
735
736                 # Return information
737                 return [{
738                         'id':           video_id.decode('utf-8'),
739                         'url':          video_url.decode('utf-8'),
740                         'uploader':     video_uploader.decode('utf-8'),
741                         'title':        video_title,
742                         'stitle':       simple_title,
743                         'ext':          video_extension.decode('utf-8'),
744                         }]
745
746
747 class YoutubeSearchIE(InfoExtractor):
748         """Information Extractor for YouTube search queries."""
749         _VALID_QUERY = r'ytsearch(\d+|all)?:[\s\S]+'
750         _TEMPLATE_URL = 'http://www.youtube.com/results?search_query=%s&page=%s&gl=US&hl=en'
751         _VIDEO_INDICATOR = r'href="/watch\?v=.+?"'
752         _MORE_PAGES_INDICATOR = r'>Next</a>'
753         _youtube_ie = None
754
755         def __init__(self, youtube_ie, downloader=None):
756                 InfoExtractor.__init__(self, downloader)
757                 self._youtube_ie = youtube_ie
758         
759         @staticmethod
760         def suitable(url):
761                 return (re.match(YoutubeSearchIE._VALID_QUERY, url) is not None)
762
763         def report_download_page(self, query, pagenum):
764                 """Report attempt to download playlist page with given number."""
765                 self.to_stdout(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
766
767         def _real_initialize(self):
768                 self._youtube_ie.initialize()
769         
770         def _real_extract(self, query):
771                 mobj = re.match(self._VALID_QUERY, query)
772                 if mobj is None:
773                         self.to_stderr(u'ERROR: invalid search query "%s"' % query)
774                         return [None]
775
776                 prefix, query = query.split(':')
777                 prefix = prefix[8:]
778                 if prefix == '':
779                         return self._download_n_results(query, 1)
780                 elif prefix == 'all':
781                         return self._download_n_results(query, -1)
782                 else:
783                         try:
784                                 n = int(prefix)
785                                 if n <= 0:
786                                         self.to_stderr(u'ERROR: invalid download number %s for query "%s"' % (n, query))
787                                         return [None]
788                                 return self._download_n_results(query, n)
789                         except ValueError: # parsing prefix as int fails
790                                 return self._download_n_results(query, 1)
791
792         def _download_n_results(self, query, n):
793                 """Downloads a specified number of results for a query"""
794
795                 video_ids = []
796                 already_seen = set()
797                 pagenum = 1
798
799                 while True:
800                         self.report_download_page(query, pagenum)
801                         result_url = self._TEMPLATE_URL % (urllib.quote_plus(query), pagenum)
802                         request = urllib2.Request(result_url, None, std_headers)
803                         try:
804                                 page = urllib2.urlopen(request).read()
805                         except (urllib2.URLError, httplib.HTTPException, socket.error), err:
806                                 self.to_stderr(u'ERROR: unable to download webpage: %s' % str(err))
807                                 return [None]
808
809                         # Extract video identifiers
810                         for mobj in re.finditer(self._VIDEO_INDICATOR, page):
811                                 video_id = page[mobj.span()[0]:mobj.span()[1]].split('=')[2][:-1]
812                                 if video_id not in already_seen:
813                                         video_ids.append(video_id)
814                                         already_seen.add(video_id)
815                                         if len(video_ids) == n:
816                                                 # Specified n videos reached
817                                                 information = []
818                                                 for id in video_ids:
819                                                         information.extend(self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id))
820                                                 return information
821
822                         if self._MORE_PAGES_INDICATOR not in page:
823                                 information = []
824                                 for id in video_ids:
825                                         information.extend(self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id))
826                                 return information
827
828                         pagenum = pagenum + 1
829
830 class YoutubePlaylistIE(InfoExtractor):
831         """Information Extractor for YouTube playlists."""
832
833         _VALID_URL = r'(?:http://)?(?:\w+\.)?youtube.com/view_play_list\?p=(.+)'
834         _TEMPLATE_URL = 'http://www.youtube.com/view_play_list?p=%s&page=%s&gl=US&hl=en'
835         _VIDEO_INDICATOR = r'/watch\?v=(.+?)&'
836         _MORE_PAGES_INDICATOR = r'/view_play_list?p=%s&amp;page=%s'
837         _youtube_ie = None
838
839         def __init__(self, youtube_ie, downloader=None):
840                 InfoExtractor.__init__(self, downloader)
841                 self._youtube_ie = youtube_ie
842         
843         @staticmethod
844         def suitable(url):
845                 return (re.match(YoutubePlaylistIE._VALID_URL, url) is not None)
846
847         def report_download_page(self, playlist_id, pagenum):
848                 """Report attempt to download playlist page with given number."""
849                 self.to_stdout(u'[youtube] PL %s: Downloading page #%s' % (playlist_id, pagenum))
850
851         def _real_initialize(self):
852                 self._youtube_ie.initialize()
853         
854         def _real_extract(self, url):
855                 # Extract playlist id
856                 mobj = re.match(self._VALID_URL, url)
857                 if mobj is None:
858                         self.to_stderr(u'ERROR: invalid url: %s' % url)
859                         return [None]
860
861                 # Download playlist pages
862                 playlist_id = mobj.group(1)
863                 video_ids = []
864                 pagenum = 1
865
866                 while True:
867                         self.report_download_page(playlist_id, pagenum)
868                         request = urllib2.Request(self._TEMPLATE_URL % (playlist_id, pagenum), None, std_headers)
869                         try:
870                                 page = urllib2.urlopen(request).read()
871                         except (urllib2.URLError, httplib.HTTPException, socket.error), err:
872                                 self.to_stderr(u'ERROR: unable to download webpage: %s' % str(err))
873                                 return [None]
874
875                         # Extract video identifiers
876                         ids_in_page = []
877                         for mobj in re.finditer(self._VIDEO_INDICATOR, page):
878                                 if mobj.group(1) not in ids_in_page:
879                                         ids_in_page.append(mobj.group(1))
880                         video_ids.extend(ids_in_page)
881
882                         if (self._MORE_PAGES_INDICATOR % (playlist_id, pagenum + 1)) not in page:
883                                 break
884                         pagenum = pagenum + 1
885
886                 information = []
887                 for id in video_ids:
888                         information.extend(self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id))
889                 return information
890
891 class PostProcessor(object):
892         """Post Processor class.
893
894         PostProcessor objects can be added to downloaders with their
895         add_post_processor() method. When the downloader has finished a
896         successful download, it will take its internal chain of PostProcessors
897         and start calling the run() method on each one of them, first with
898         an initial argument and then with the returned value of the previous
899         PostProcessor.
900
901         The chain will be stopped if one of them ever returns None or the end
902         of the chain is reached.
903
904         PostProcessor objects follow a "mutual registration" process similar
905         to InfoExtractor objects.
906         """
907
908         _downloader = None
909
910         def __init__(self, downloader=None):
911                 self._downloader = downloader
912
913         def to_stdout(self, message):
914                 """Print message to stdout if downloader is not in quiet mode."""
915                 if self._downloader is None or not self._downloader.params.get('quiet', False):
916                         print message
917         
918         def to_stderr(self, message):
919                 """Print message to stderr."""
920                 print >>sys.stderr, message
921
922         def set_downloader(self, downloader):
923                 """Sets the downloader for this PP."""
924                 self._downloader = downloader
925         
926         def run(self, information):
927                 """Run the PostProcessor.
928
929                 The "information" argument is a dictionary like the ones
930                 returned by InfoExtractors. The only difference is that this
931                 one has an extra field called "filepath" that points to the
932                 downloaded file.
933
934                 When this method returns None, the postprocessing chain is
935                 stopped. However, this method may return an information
936                 dictionary that will be passed to the next postprocessing
937                 object in the chain. It can be the one it received after
938                 changing some fields.
939
940                 In addition, this method may raise a PostProcessingError
941                 exception that will be taken into account by the downloader
942                 it was called from.
943                 """
944                 return information # by default, do nothing
945         
946 ### MAIN PROGRAM ###
947 if __name__ == '__main__':
948         try:
949                 # Modules needed only when running the main program
950                 import getpass
951                 import optparse
952
953                 # General configuration
954                 urllib2.install_opener(urllib2.build_opener(urllib2.ProxyHandler()))
955                 urllib2.install_opener(urllib2.build_opener(urllib2.HTTPCookieProcessor()))
956                 socket.setdefaulttimeout(300) # 5 minutes should be enough (famous last words)
957
958                 # Parse command line
959                 parser = optparse.OptionParser(
960                                 usage='Usage: %prog [options] url...',
961                                 version='2009.04.06',
962                                 conflict_handler='resolve',
963                                 )
964                 parser.add_option('-h', '--help',
965                                 action='help', help='print this help text and exit')
966                 parser.add_option('-v', '--version',
967                                 action='version', help='print program version and exit')
968                 parser.add_option('-u', '--username',
969                                 dest='username', metavar='UN', help='account username')
970                 parser.add_option('-p', '--password',
971                                 dest='password', metavar='PW', help='account password')
972                 parser.add_option('-o', '--output',
973                                 dest='outtmpl', metavar='TPL', help='output filename template')
974                 parser.add_option('-q', '--quiet',
975                                 action='store_true', dest='quiet', help='activates quiet mode', default=False)
976                 parser.add_option('-s', '--simulate',
977                                 action='store_true', dest='simulate', help='do not download video', default=False)
978                 parser.add_option('-t', '--title',
979                                 action='store_true', dest='usetitle', help='use title in file name', default=False)
980                 parser.add_option('-l', '--literal',
981                                 action='store_true', dest='useliteral', help='use literal title in file name', default=False)
982                 parser.add_option('-n', '--netrc',
983                                 action='store_true', dest='usenetrc', help='use .netrc authentication data', default=False)
984                 parser.add_option('-g', '--get-url',
985                                 action='store_true', dest='geturl', help='simulate, quiet but print URL', default=False)
986                 parser.add_option('-e', '--get-title',
987                                 action='store_true', dest='gettitle', help='simulate, quiet but print title', default=False)
988                 parser.add_option('-f', '--format',
989                                 dest='format', metavar='FMT', help='video format code')
990                 parser.add_option('-m', '--mobile-version',
991                                 action='store_const', dest='format', help='alias for -f 17', const='17')
992                 parser.add_option('-d', '--high-def',
993                                 action='store_const', dest='format', help='alias for -f 22', const='22')
994                 parser.add_option('-i', '--ignore-errors',
995                                 action='store_true', dest='ignoreerrors', help='continue on download errors', default=False)
996                 parser.add_option('-r', '--rate-limit',
997                                 dest='ratelimit', metavar='L', help='download rate limit (e.g. 50k or 44.6m)')
998                 parser.add_option('-a', '--batch-file',
999                                 dest='batchfile', metavar='F', help='file containing URLs to download')
1000                 parser.add_option('-w', '--no-overwrites',
1001                                 action='store_true', dest='nooverwrites', help='do not overwrite files', default=False)
1002                 (opts, args) = parser.parse_args()
1003
1004                 # Batch file verification
1005                 batchurls = []
1006                 if opts.batchfile is not None:
1007                         try:
1008                                 batchurls = [line.strip() for line in open(opts.batchfile, 'r')]
1009                         except IOError:
1010                                 sys.exit(u'ERROR: batch file could not be read')
1011                 all_urls = batchurls + args
1012
1013                 # Conflicting, missing and erroneous options
1014                 if len(all_urls) < 1:
1015                         sys.exit(u'ERROR: you must provide at least one URL')
1016                 if opts.usenetrc and (opts.username is not None or opts.password is not None):
1017                         sys.exit(u'ERROR: using .netrc conflicts with giving username/password')
1018                 if opts.password is not None and opts.username is None:
1019                         sys.exit(u'ERROR: account username missing')
1020                 if opts.outtmpl is not None and (opts.useliteral or opts.usetitle):
1021                         sys.exit(u'ERROR: using output template conflicts with using title or literal title')
1022                 if opts.usetitle and opts.useliteral:
1023                         sys.exit(u'ERROR: using title conflicts with using literal title')
1024                 if opts.username is not None and opts.password is None:
1025                         opts.password = getpass.getpass(u'Type account password and press return:')
1026                 if opts.ratelimit is not None:
1027                         numeric_limit = FileDownloader.parse_bytes(opts.ratelimit)
1028                         if numeric_limit is None:
1029                                 sys.exit(u'ERROR: invalid rate limit specified')
1030                         opts.ratelimit = numeric_limit
1031
1032                 # Information extractors
1033                 youtube_ie = YoutubeIE()
1034                 metacafe_ie = MetacafeIE(youtube_ie)
1035                 youtube_pl_ie = YoutubePlaylistIE(youtube_ie)
1036                 youtube_search_ie = YoutubeSearchIE(youtube_ie)
1037
1038                 # File downloader
1039                 charset = locale.getdefaultlocale()[1]
1040                 if charset is None:
1041                         charset = 'ascii'
1042                 fd = FileDownloader({
1043                         'usenetrc': opts.usenetrc,
1044                         'username': opts.username,
1045                         'password': opts.password,
1046                         'quiet': (opts.quiet or opts.geturl or opts.gettitle),
1047                         'forceurl': opts.geturl,
1048                         'forcetitle': opts.gettitle,
1049                         'simulate': (opts.simulate or opts.geturl or opts.gettitle),
1050                         'format': opts.format,
1051                         'outtmpl': ((opts.outtmpl is not None and opts.outtmpl.decode(charset))
1052                                 or (opts.usetitle and u'%(stitle)s-%(id)s.%(ext)s')
1053                                 or (opts.useliteral and u'%(title)s-%(id)s.%(ext)s')
1054                                 or u'%(id)s.%(ext)s'),
1055                         'ignoreerrors': opts.ignoreerrors,
1056                         'ratelimit': opts.ratelimit,
1057                         'nooverwrites': opts.nooverwrites,
1058                         })
1059                 fd.add_info_extractor(youtube_search_ie)
1060                 fd.add_info_extractor(youtube_pl_ie)
1061                 fd.add_info_extractor(metacafe_ie)
1062                 fd.add_info_extractor(youtube_ie)
1063                 retcode = fd.download(all_urls)
1064                 sys.exit(retcode)
1065
1066         except DownloadError:
1067                 sys.exit(1)
1068         except SameFileError:
1069                 sys.exit(u'ERROR: fixed output name but more than one file to download')
1070         except KeyboardInterrupt:
1071                 sys.exit(u'\nERROR: Interrupted by user')