]> gitweb.michael.orlitzky.com - dead/whatever-dl.git/blob - src/websites/youtube.rb
Fix the Youtube class for the latest update to the site.
[dead/whatever-dl.git] / src / websites / youtube.rb
1 #
2 # Copyright Michael Orlitzky
3 #
4 # http://michael.orlitzky.com/
5 #
6 # This program is free software: you can redistribute it and/or modify
7 # it under the terms of the GNU General Public License as published by
8 # the Free Software Foundation, either version 3 of the License, or
9 # (at your option) any later version.
10 #
11 # This program is distributed in the hope that it will be useful,
12 # but WITHOUT ANY WARRANTY; without even the implied warranty of
13 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 # GNU General Public License for more details.
15 #
16 # http://www.fsf.org/licensing/licenses/gpl.html
17 #
18
19 require 'src/website'
20 require 'cgi'
21
22 class Youtube < Website
23
24 VALID_YOUTUBE_URL_REGEX = /^(http:\/\/)?([a-z0-9]+\.)?youtube\.com\/((watch\?v=)|(v\/)|([a-z]+\#[a-z]\/[a-z]\/[0-9]\/))[a-z0-9_\-]+(\&.*)?\#?$/i
25
26 def self.owns_url?(url)
27 return url =~ VALID_YOUTUBE_URL_REGEX
28 end
29
30
31 def initialize(url)
32 super
33
34 # The @format variable just caches the format of the video we're
35 # downloading. Storing it will prevent us from having to calculate
36 # it twice.
37 @format = 0
38 end
39
40
41 def get_video_url()
42 video_id = self.parse_video_id()
43
44 # The video's URL (the "page data" URL) may be different from the
45 # URL that was passed to the program. We support the /v/video_id
46 # URL format, but that is *not* the main video page where we can
47 # retrieve the "t" parameter. We can only get that from the
48 # /watch?v=video_id form.
49 page_data_url = "http://www.youtube.com/watch?v=#{video_id}"
50 page_data = self.get_page_data(page_data_url)
51
52 # Get the URL map from the page.
53 fmt_url_map = get_format_url_map(page_data)
54
55 # Figure out which formats are available, and if any are,
56 # choose the best one.
57 available_formats = fmt_url_map.keys()
58 desired_format = get_desired_format(available_formats)
59
60 # First we cache the format so that when we're asked for the
61 # video filename later, we don't have to recompute the format.
62 @format = desired_format
63
64 # And then use whatever URL is available for the desired format.
65 # We assume that all available formats will have an entry in the
66 # fmt_url_map hash.
67 video_url = fmt_url_map[desired_format]
68
69 return video_url
70 end
71
72
73 def get_video_filename()
74 # The format -> extension mapping is available on Wikipedia:
75 #
76 # http://en.wikipedia.org/wiki/YouTube#Quality_and_codecs
77 #
78 # The default extension is .flv.
79 extension = '.flv'
80
81 if [18, 22, 35, 37].include?(@format)
82 extension = '.mp4'
83 elsif (@format == 17)
84 extension = '.3gp'
85 end
86
87 return (self.parse_video_id() + extension)
88 end
89
90
91 protected;
92
93 # Get the video id from the URL. Should be relatively easy,
94 # unless Youtube supports some URL formats of which I'm unaware.
95 def parse_video_id()
96 # Both URLs are fairly easy to parse if you handle
97 # them one at a time. The only tricky situation is when
98 # parameters like "&hl=en" are tacked on to the end.
99 # We'll call /watch?v=video_id the "first form."
100 first_form_video_id_regex = /v=([0-9a-z_\-]+)/i
101 first_form_matches = first_form_video_id_regex.match(@url)
102 return first_form_matches[1] if not (first_form_matches.nil? ||
103 first_form_matches.length < 2)
104
105 # First form didn't work? Try the second.
106 second_form_video_id_regex = /\/v\/([0-9a-z_\-]+)/i
107 second_form_matches = second_form_video_id_regex.match(@url)
108 return second_form_matches[1] if not (second_form_matches.nil? ||
109 second_form_matches.length < 2)
110
111 # ...and the third.
112 third_form_video_id_regex = /\/([[:alnum:]]+)$/i
113 third_form_matches = third_form_video_id_regex.match(@url)
114 return third_form_matches[1] if not (third_form_matches.nil? ||
115 third_form_matches.length < 2)
116
117 # If we made it here, we couldn't figure out the video id. Yes,
118 # this is fatal, since we don't know where the video file is
119 # located.
120 raise StandardError.new("Could not parse the video id.")
121 end
122
123
124
125 def get_format_url_map(page_data)
126 # Youtube has implemented a new fmt_url_map that (perhaps
127 # unsurprisingly) maps formats to video URLs. This makes it
128 # easyish to parse the video URLs.
129 url_map = {}
130 url_map_regex = /fmt_url_map=([^&\"]+)/
131
132 matches = url_map_regex.match(page_data)
133
134 if (matches.nil? || matches.length < 1)
135 raise StandardError.new("Could not parse the fmt_url_map Flash variable.")
136 end
137
138 # The map is stored entirely in one Flash variable. The format is
139 # key|value,key|value,...
140 maptext = CGI::unescape(matches[1])
141 entries = maptext.split(',')
142 entries.each do |entry|
143 key = entry.split('|')[0].to_i
144 value = entry.split('|')[1]
145 url_map[key] = value
146 end
147
148 if (url_map.length < 1)
149 raise StandardError.new("Could not find any valid format URLs.")
150 end
151
152 return url_map
153 end
154
155
156 def get_desired_format(available_formats)
157 # Check for the presence of formats, in order of preference
158 # (quality). That is, we check for the best formats first. As soon
159 # as a format is found to be available, we return it as the
160 # desired format, since the first format we find is going to be
161 # the best available format.
162 return 37 if available_formats.include?(37)
163 return 22 if available_formats.include?(22)
164 return 35 if available_formats.include?(35)
165 return 18 if available_formats.include?(18)
166 return 34 if available_formats.include?(34)
167 return 17 if available_formats.include?(17)
168
169 # Available formats can't be empty (we would have raised an error
170 # in get_available_formats), so if there's some unknown format
171 # here we might as well return it as a last resort.
172 return available_formats[0]
173 end
174
175 end