share/lua/website/youtube.lua

   1
   2 -- quvi
   3 -- Copyright (C) 2010,2011  Toni Gundogdu <legatvs@gmail.com>
   4 --
   5 -- This file is part of quvi <http://quvi.sourceforge.net/>.
   6 --
   7 -- This library is free software; you can redistribute it and/or
   8 -- modify it under the terms of the GNU Lesser General Public
   9 -- License as published by the Free Software Foundation; either
  10 -- version 2.1 of the License, or (at your option) any later version.
  11 --
  12 -- This library is distributed in the hope that it will be useful,
  13 -- but WITHOUT ANY WARRANTY; without even the implied warranty of
  14 -- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  15 -- Lesser General Public License for more details.
  16 --
  17 -- You should have received a copy of the GNU Lesser General Public
  18 -- License along with this library; if not, write to the Free Software
  19 -- Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
  20 -- 02110-1301  USA
  21 --
  22
  23 local YouTube = {} -- Utility functions unique to this script
  24
  25 -- <http://en.wikipedia.org/wiki/YouTube#Quality_and_codecs>
  26
  27 -- Identify the script.
  28 function ident(self)
  29     package.path = self.script_dir .. '/?.lua'
  30     local C      = require 'quvi/const'
  31     local r      = {}
  32     r.domain     = "youtube.com"
  33     r.formats    = "default|best"
  34     r.categories  = C.proto_http
  35     self.page_url = YouTube.normalize(self.page_url)
  36     local U       = require 'quvi/util'
  37     r.handles     = U.handles(self.page_url,
  38                         {r.domain}, {"/watch"}, {"v=[%w-_]+"})
  39     return r
  40 end
  41
  42 -- Query available formats.
  43 function query_formats(self)
  44     local config,U  = YouTube.get_config(self)
  45     local formats   = YouTube.iter_formats(config, U)
  46
  47     local t = {}
  48     for _,v in pairs(formats) do
  49         table.insert(t, YouTube.to_s(v))
  50     end
  51
  52     table.sort(t)
  53     self.formats = table.concat(t, "|")
  54
  55     return self
  56 end
  57
  58 -- Parse URL.
  59 function parse(self)
  60     self.host_id   = "youtube"
  61     local page_url = YouTube.normalize(self.page_url)
  62
  63     local _,_,s = page_url:find('#a?t=(.+)')
  64     self.start_time = s or ''
  65
  66     return YouTube.get_video_info(self)
  67 end
  68
  69 --
  70 -- Utility functions
  71 --
  72
  73 function YouTube.normalize(url)
  74     if not url then return url end
  75     url = url:gsub("-nocookie", "")        -- youtube-nocookie.com
  76     url = url:gsub("/v/", "/watch?v=")     -- embedded
  77     url = url:gsub("/embed/", "/watch?v=") -- embedded
  78     url = url:gsub("/e/", "/watch?v=")     -- embedded
  79     url = url:gsub("youtu%.be/","youtube.com/watch?v=") -- youtu.be
  80     return url
  81 end
  82
  83 function YouTube.get_config(self)
  84     local _,_,s  = self.page_url:find('^(%w+)://')
  85     local scheme = s or error("no match: scheme")
  86
  87     local page_url = YouTube.normalize(self.page_url)
  88
  89     local _,_,s = page_url:find("v=([%w-_]+)")
  90     self.id = s or error("no match: media id")
  91
  92     local s_fmt = "%s://www.youtube.com/get_video_info?&video_id=%s"
  93                     .. "&el=detailpage&ps=default&eurl=&gl=US&hl=en"
  94
  95     local config_url = string.format(s_fmt, scheme, self.id)
  96
  97     local U      = require 'quvi/util'
  98     local config = U.decode(quvi.fetch(config_url, {fetch_type='config'}))
  99
 100     if config['reason'] then
 101         local reason = U.unescape(config['reason'])
 102         local code   = config['errorcode']
 103         error(string.format("%s (code=%s)", reason, code))
 104     end
 105
 106     return config,U
 107 end
 108
 109 function YouTube.iter_formats(config, U)
 110     local fmt_stream_map = config['url_encoded_fmt_stream_map']
 111                         or error("no match: url_encoded_fmt_stream_map")
 112
 113     fmt_stream_map = U.unescape(fmt_stream_map) .. ","
 114
 115     local urls = {}
 116     for f in fmt_stream_map:gfind('([^,]*),') do
 117         local d = U.decode(f)
 118         if d['itag'] and d['url'] then
 119             urls[U.unescape(d['itag'])] = U.unescape(d['url'])
 120         end
 121     end
 122
 123     local fmt_map = config['fmt_list'] or error("no match: fmt_list")
 124     fmt_map = U.unescape(fmt_map)
 125
 126     local r = {}
 127     for f,w,h in fmt_map:gfind('(%d+)/(%d+)x(%d+)') do
 128 --        print(f,w,h)
 129         table.insert(r, {fmt_id=tonumber(f),    url=urls[f],
 130                           width=tonumber(w), height=tonumber(h)})
 131     end
 132
 133     return r
 134 end
 135
 136 function YouTube.get_video_info(self)
 137     local config,U = YouTube.get_config(self)
 138
 139     self.title = config['title'] or error('no match: media title')
 140     self.title = U.unescape(self.title)
 141
 142     self.thumbnail_url = config['thumbnail_url'] or ''
 143     if #self.thumbnail_url > 0 then
 144         self.thumbnail_url = U.unescape(self.thumbnail_url)
 145     end
 146
 147     self.duration = (config['length_seconds'] or 0)*1000 -- to msec
 148
 149     self.requested_format =
 150         YouTube.convert_deprecated_id(self.requested_format)
 151
 152     local formats = YouTube.iter_formats(config, U)
 153     local url     = U.choose_format(self, formats,
 154                                     YouTube.choose_best,
 155                                     YouTube.choose_default,
 156                                     YouTube.to_s).url
 157                         or error("no match: media url")
 158
 159     if url and #self.start_time > 0 then
 160         local min, sec = self.start_time:match("^(%d+)m(%d+)s$")
 161         min = tonumber(min) or 0
 162         sec = tonumber(sec) or 0
 163         local msec = (min * 60000) + (sec * 1000)
 164         if msec > 0 then
 165             url = url .. "&begin=" .. msec
 166         end
 167     end
 168
 169     self.url = {url}
 170
 171     return self
 172 end
 173
 174 function YouTube.choose_best(formats) -- Highest quality available
 175     local r = {width=0, height=0, url=nil}
 176     local U = require 'quvi/util'
 177     for _,v in pairs(formats) do
 178         if U.is_higher_quality(v,r) then
 179             r = v
 180         end
 181     end
 182 --    for k,v in pairs(r) do print(k,v) end
 183     return r
 184 end
 185
 186 function YouTube.choose_default(formats) -- Lowest quality available
 187     local r = {width=0xffff, height=0xffff, url=nil}
 188     local U = require 'quvi/util'
 189     for _,v in pairs(formats) do
 190         if U.is_lower_quality(v,r) then
 191             r = v
 192         end
 193     end
 194 --    for k,v in pairs(r) do print(k,v) end
 195     return r
 196 end
 197
 198 YouTube.conv_table = { -- Deprecated.
 199     -- flv
 200     flv_240p =  '5',
 201     flv_360p = '34',
 202     flv_480p = '35',
 203     -- mp4
 204      mp4_360p = '18',
 205      mp4_720p = '22',
 206     mp4_1080p = '37',
 207     mp4_3072p = '38'
 208 }
 209
 210 function YouTube.convert_deprecated_id(r_fmt)
 211     if YouTube.conv_table[r_fmt] then
 212         local s = string.format("fmt%02d_", YouTube.conv_table[r_fmt])
 213         r_fmt = r_fmt:gsub("^(%w+)_", s)
 214     end
 215     return r_fmt
 216 end
 217
 218 function YouTube.to_s(t)
 219     return string.format("fmt%02d_%sp", t.fmt_id, t.height)
 220 end
 221
 222 -- vim: set ts=4 sw=4 tw=72 expandtab: