Raphaël G. Git Repositories - youtubedl/blob - youtube-dl

   1 #!/usr/bin/env python
   2 #
   3 # Copyright (c) 2006 Ricardo Garcia Gonzalez
   4 #
   5 # Permission is hereby granted, free of charge, to any person obtaining a
   6 # copy of this software and associated documentation files (the "Software"),
   7 # to deal in the Software without restriction, including without limitation
   8 # the rights to use, copy, modify, merge, publish, distribute, sublicense,
   9 # and/or sell copies of the Software, and to permit persons to whom the
  10 # Software is furnished to do so, subject to the following conditions:
  11 #
  12 # The above copyright notice and this permission notice shall be included
  13 # in all copies or substantial portions of the Software.
  14 #
  15 # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  16 # IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  17 # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
  18 # THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR
  19 # OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
  20 # ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  21 # OTHER DEALINGS IN THE SOFTWARE.
  22 #
  23 # Except as contained in this notice, the name(s) of the above copyright
  24 # holders shall not be used in advertising or otherwise to promote the
  25 # sale, use or other dealings in this Software without prior written
  26 # authorization.
  27 #
  28 import sys
  29 import optparse
  30 import httplib
  31 import urllib2
  32 import re
  33 import string
  34 import os
  35
  36 # First off, check Python and refuse to run
  37 if sys.hexversion < 0x020400f0:
  38         sys.exit('Error: Python 2.4 or later needed to run the program')
  39
  40 # Global constants
  41 const_video_url_str = 'http://www.youtube.com/watch?v=%s'
  42 const_video_url_re = re.compile(r'(?:http://)?(?:www\d*\.)?youtube\.com/(?:v/|(?:watch(?:\.php)?)?\?v=)([^&]+).*')
  43 const_login_url_str = 'http://www.youtube.com/login?next=/watch%%3Fv%%3D%s'
  44 const_login_post_str = 'current_form=loginForm&next=%%2Fwatch%%3Fv%%3D%s&username=%s&password=%s&action_login=Log+In'
  45 const_age_url_str = 'http://www.youtube.com/verify_age?next_url=/watch%%3Fv%%3D%s'
  46 const_age_post_str = 'next_url=%%2Fwatch%%3Fv%%3D%s&action_confirm=Confirm'
  47 const_video_url_params_re = re.compile(r'player2\.swf\?([^"]+)"', re.M)
  48 const_video_url_real_str = 'http://www.youtube.com/get_video?%s'
  49 const_video_title_re = re.compile(r'<title>YouTube - ([^<]*)</title>', re.M | re.I)
  50 const_1k = 1024
  51 const_block_size = 10 * const_1k
  52
  53 # Print error message, followed by standard advice information, and then exit
  54 def error_advice_exit(error_text):
  55         sys.stderr.write('Error: %s.\n' % error_text)
  56         sys.stderr.write('Try again several times. It may be a temporal problem.\n')
  57         sys.stderr.write('Other typical problems:\n\n')
  58         sys.stderr.write('\tVideo no longer exists.\n')
  59         sys.stderr.write('\tVideo requires age confirmation but you did not provide an account.\n')
  60         sys.stderr.write('\tYou provided the account data, but it is not valid.\n')
  61         sys.stderr.write('\tThe connection was cut suddenly for some reason.\n')
  62         sys.stderr.write('\tYouTube changed their system, and the program no longer works.\n')
  63         sys.stderr.write('\nTry to confirm you are able to view the video using a web browser.\n')
  64         sys.stderr.write('Use the same video URL and account information, if needed, with this program.\n')
  65         sys.stderr.write('When using a proxy, make sure http_proxy has http://host:port format.\n')
  66         sys.stderr.write('Try again several times and contact me if the problem persists.\n')
  67         sys.exit('\n')
  68
  69 # Wrapper to create custom requests with typical headers
  70 def request_create(url, data=None):
  71         retval = urllib2.Request(url)
  72         if not data is None:
  73                 retval.add_data(data)
  74         # Try to mimic Firefox, at least a little bit
  75         retval.add_header('User-Agent', 'Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.1) Gecko/20061010 Firefox/2.0')
  76         retval.add_header('Accept-Charset', 'ISO-8859-1,utf-8;q=0.7,*;q=0.7')
  77         retval.add_header('Accept', 'text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5')
  78         retval.add_header('Accept-Language', 'en-us,en;q=0.5')
  79         return retval
  80
  81 # Perform a request, process headers and return response
  82 def perform_request(url, data=None):
  83         request = request_create(url, data)
  84         response = urllib2.urlopen(request)
  85         return response
  86
  87 # Convert bytes to KiB
  88 def to_k(bytes):
  89         global const_1k
  90         return bytes / const_1k
  91
  92 # Conditional print
  93 def cond_print(str):
  94         global cmdl_opts
  95         if not cmdl_opts.quiet:
  96                 sys.stdout.write(str)
  97                 sys.stdout.flush()
  98
  99 # Title string normalization
 100 def title_string_norm(title):
 101         title = title.lower()
 102         title = map(lambda x: x in string.whitespace and '_' or x, title)
 103         title = filter(lambda x: x in string.lowercase or x in string.digits or x == '_', title)
 104         return ''.join(title)
 105
 106 # Generic download step
 107 def download_step(return_data_flag, step_title, step_error, url, post_data=None):
 108         try:
 109                 cond_print('%s... ' % step_title)
 110                 data = perform_request(url, post_data).read()
 111                 cond_print('done.\n')
 112                 if return_data_flag:
 113                         return data
 114                 return None
 115
 116         except (urllib2.URLError, ValueError, httplib.HTTPException, TypeError):
 117                 cond_print('failed.\n')
 118                 error_advice_exit(step_error)
 119
 120         except KeyboardInterrupt:
 121                 sys.exit('\n')
 122
 123 # Generic extract step
 124 def extract_step(step_title, step_error, regexp, data):
 125         try:
 126                 cond_print('%s... ' % step_title)
 127                 match = regexp.search(data)
 128
 129                 if match is None:
 130                         cond_print('failed.\n')
 131                         error_advice_exit(step_error)
 132
 133                 extracted_data = match.group(1)
 134                 cond_print('done.\n')
 135                 return extracted_data
 136
 137         except KeyboardInterrupt:
 138                 sys.exit('\n')
 139
 140 # Create the command line options parser and parse command line
 141 cmdl_usage = 'usage: %prog [options] video_url'
 142 cmdl_version = '2006.12.07'
 143 cmdl_parser = optparse.OptionParser(usage=cmdl_usage, version=cmdl_version, conflict_handler='resolve')
 144 cmdl_parser.add_option('-h', '--help', action='help', help='print this help text and exit')
 145 cmdl_parser.add_option('-v', '--version', action='version', help='print program version and exit')
 146 cmdl_parser.add_option('-u', '--username', dest='username', metavar='USERNAME', help='account username')
 147 cmdl_parser.add_option('-p', '--password', dest='password', metavar='PASSWORD', help='account password')
 148 cmdl_parser.add_option('-o', '--output', dest='outfile', metavar='FILE', help='output video file name')
 149 cmdl_parser.add_option('-q', '--quiet', action='store_true', dest='quiet', help='activates quiet mode')
 150 cmdl_parser.add_option('-s', '--simulate', action='store_true', dest='simulate', help='do not download video')
 151 cmdl_parser.add_option('-t', '--title', action='store_true', dest='use_title', help='use title in file name')
 152 (cmdl_opts, cmdl_args) = cmdl_parser.parse_args()
 153
 154 # Get video URL
 155 if len(cmdl_args) != 1:
 156         cmdl_parser.print_help()
 157         sys.exit('\n')
 158 video_url_cmdl = cmdl_args[0]
 159
 160 # Verify video URL format and convert to "standard" format
 161 video_url_mo = const_video_url_re.match(video_url_cmdl)
 162 if video_url_mo is None:
 163         sys.exit('Error: URL does not seem to be a youtube video URL. If it is, report a bug.')
 164 video_url_id = video_url_mo.group(1)
 165 video_url = const_video_url_str % video_url_id
 166
 167 # Check conflicting options
 168 if not cmdl_opts.outfile is None and cmdl_opts.simulate:
 169         sys.stderr.write('Warning: video file name given but will not be used.\n')
 170
 171 if not cmdl_opts.outfile is None and cmdl_opts.use_title:
 172         sys.exit('Error: using the video title conflicts with using a given file name.')
 173
 174 # Verify both or none present
 175 if ((cmdl_opts.username is None and not cmdl_opts.password is None) or
 176     (not cmdl_opts.username is None and cmdl_opts.password is None)):
 177         sys.exit('Error: both username and password must be given, or none.')
 178
 179 # Get output file name
 180 if cmdl_opts.outfile is None:
 181         video_filename = '%s.flv' % video_url_id
 182 else:
 183         video_filename = cmdl_opts.outfile
 184
 185 # Check name
 186 if not video_filename.lower().endswith('.flv'):
 187         sys.stderr.write('Warning: video file name does not end in .flv\n')
 188
 189 # Test writable file
 190 if not cmdl_opts.simulate:
 191         try:
 192                 disk_test = open(video_filename, 'wb')
 193                 disk_test.close()
 194
 195         except (OSError, IOError):
 196                 sys.exit('Error: unable to open %s for writing.' % video_filename)
 197
 198 # Install cookie and proxy handlers
 199 urllib2.install_opener(urllib2.build_opener(urllib2.ProxyHandler()))
 200 urllib2.install_opener(urllib2.build_opener(urllib2.HTTPCookieProcessor()))
 201
 202 # Log in and confirm age if needed
 203 if not cmdl_opts.username is None:
 204         url = const_login_url_str % video_url_id
 205         post = const_login_post_str % (video_url_id, cmdl_opts.username, cmdl_opts.password)
 206         download_step(False, 'Logging in', 'unable to log in', url, post)
 207
 208         url = const_age_url_str % video_url_id
 209         post = const_age_post_str % video_url_id
 210         download_step(False, 'Confirming age', 'unable to confirm age', url, post)
 211
 212 # Retrieve video webpage
 213 video_webpage = download_step(True, 'Retrieving video webpage', 'unable to retrieve video webpage', video_url)
 214
 215 # Extract video title if needed
 216 if cmdl_opts.use_title:
 217         video_title = extract_step('Extracting video title', 'unable to extract video title', const_video_title_re, video_webpage)
 218
 219 # Extract needed video URL parameters
 220 video_url_params = extract_step('Extracting video URL parameters', 'unable to extract URL parameters', const_video_url_params_re, video_webpage)
 221 video_url_real = const_video_url_real_str % video_url_params
 222
 223 # Retrieve video data
 224 try:
 225         video_data = perform_request(video_url_real)
 226         cond_print('Video data found at %s\n' % video_data.geturl())
 227
 228         # Abort here if in simulate mode
 229         if cmdl_opts.simulate:
 230                 sys.exit()
 231
 232         cond_print('Retrieving video data... ')
 233         video_file = open(video_filename, 'wb')
 234         try:
 235                 video_len_str = '%sk' % to_k(long(video_data.info()['Content-length']))
 236         except KeyError:
 237                 video_len_str = '(unknown)'
 238
 239         byte_counter = 0
 240         video_block = video_data.read(const_block_size)
 241         while len(video_block) != 0:
 242                 byte_counter += len(video_block)
 243                 video_file.write(video_block)
 244                 cond_print('\rRetrieving video data... %sk of %s ' % (to_k(byte_counter), video_len_str))
 245                 video_block = video_data.read(const_block_size)
 246
 247         video_file.close()
 248         cond_print('done.\n')
 249         cond_print('Video data saved to %s\n' % video_filename)
 250
 251 except (urllib2.URLError, ValueError, httplib.HTTPException, TypeError):
 252         cond_print('failed.\n')
 253         error_advice_exit('unable to download video data')
 254
 255 except KeyboardInterrupt:
 256         sys.exit('\n')
 257
 258 # Rename video file if needed
 259 if cmdl_opts.use_title:
 260         try:
 261                 final_filename = '%s-%s.flv' % (title_string_norm(video_title), video_url_id)
 262                 os.rename(video_filename, final_filename)
 263                 cond_print('Video file renamed to %s\n' % final_filename)
 264
 265         except OSError:
 266                 sys.stderr.write('Warning: unable to rename file.\n')
 267
 268         except KeyboardInterrupt:
 269                 sys.exit('\n')
 270
 271 # Finish
 272 sys.exit()