share/lua/media/youtube.lua

   1 -- libquvi-scripts
   2 -- Copyright (C) 2010-2012  Toni Gundogdu <legatvs@gmail.com>
   3 --
   4 -- This file is part of libquvi-scripts <http://quvi.sourceforge.net/>.
   5 --
   6 -- This library is free software; you can redistribute it and/or
   7 -- modify it under the terms of the GNU Lesser General Public
   8 -- License as published by the Free Software Foundation; either
   9 -- version 2.1 of the License, or (at your option) any later version.
  10 --
  11 -- This library 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 GNU
  14 -- Lesser General Public License for more details.
  15 --
  16 -- You should have received a copy of the GNU Lesser General Public
  17 -- License along with this library; if not, write to the Free Software
  18 -- Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
  19 -- 02110-1301  USA
  20 --
  21
  22 local YouTube = {} -- Utility functions unique to this script
  23
  24 -- <http://en.wikipedia.org/wiki/YouTube#Quality_and_codecs>
  25
  26 -- Identify the script.
  27 function ident(self)
  28   local A = require 'quvi/accepts'
  29   local Y = require 'quvi/youtube'
  30   local C = require 'quvi/const'
  31   local u = Y.normalize(qargs.input_url)
  32   local r = {
  33     accepts = A.accepts(u, {"youtube%.com"}, {"/watch"}, {"v=[%w-_]+"}),
  34     categories = C.proto_http
  35   }
  36   return r
  37 end
  38
  39 -- Parse media properties.
  40 function parse(self)
  41     local p_url = YouTube.normalize(self.page_url)
  42     self.start_time = p_url:match('#a?t=(.+)') or ''
  43
  44     return YouTube.get_video_info(self)
  45 end
  46
  47 --
  48 -- Utility functions
  49 --
  50
  51 function YouTube.get_config(self)
  52     local sch = self.page_url:match('^(%w+)://')
  53                   or error("no match: scheme")
  54
  55     local p_url = YouTube.normalize(self.page_url)
  56
  57     self.id = p_url:match("v=([%w-_]+)")
  58                 or error("no match: media ID")
  59
  60     local s_fmt = "%s://www.youtube.com/get_video_info?&video_id=%s"
  61                     .. "&el=detailpage&ps=default&eurl=&gl=US&hl=en"
  62
  63     local c_url = string.format(s_fmt, sch, self.id)
  64
  65     local U = require 'quvi/util'
  66     local c = U.decode(quvi.fetch(c_url, {fetch_type='config'}))
  67
  68     if c['reason'] then
  69         local reason = U.unescape(c['reason'])
  70         local code = c['errorcode']
  71         error(string.format("%s (code=%s)", reason, code))
  72     end
  73
  74     return c, U
  75 end
  76
  77 function YouTube.iter_formats(config, U)
  78     local fmt_stream_map = config['url_encoded_fmt_stream_map']
  79                         or error("no match: url_encoded_fmt_stream_map")
  80
  81     fmt_stream_map = U.unescape(fmt_stream_map) .. ","
  82
  83     local urls = {}
  84     for f in fmt_stream_map:gmatch('([^,]*),') do
  85         local d = U.decode(f)
  86         if d['itag'] and d['url'] then
  87             urls[U.unescape(d['itag'])] = U.unescape(d['url'])
  88         end
  89     end
  90
  91     local fmt_map = config['fmt_list'] or error("no match: fmt_list")
  92     fmt_map = U.unescape(fmt_map)
  93
  94     local r = {}
  95     for f,w,h in fmt_map:gmatch('(%d+)/(%d+)x(%d+)') do
  96 --        print(f,w,h)
  97         table.insert(r, {fmt_id=tonumber(f),    url=urls[f],
  98                           width=tonumber(w), height=tonumber(h)})
  99     end
 100
 101     return r
 102 end
 103
 104 function YouTube.get_video_info(self)
 105     local config,U = YouTube.get_config(self)
 106
 107     self.title = config['title'] or error('no match: media title')
 108     self.title = U.unescape(self.title)
 109
 110     self.thumbnail_url = config['thumbnail_url'] or ''
 111     if #self.thumbnail_url > 0 then
 112         self.thumbnail_url = U.unescape(self.thumbnail_url)
 113     end
 114
 115     self.duration = (config['length_seconds'] or 0)*1000 -- to msec
 116
 117     self.requested_format =
 118         YouTube.convert_deprecated_id(self.requested_format)
 119
 120     local formats = YouTube.iter_formats(config, U)
 121     local format  = U.choose_format(self, formats,
 122                                     YouTube.choose_best,
 123                                     YouTube.choose_default,
 124                                     YouTube.to_s)
 125                         or error("unable to choose format")
 126     local url     = format.url or error("no match: media url")
 127
 128     if url and #self.start_time > 0 then
 129         local min, sec = self.start_time:match("^(%d+)m(%d+)s$")
 130         min = tonumber(min) or 0
 131         sec = tonumber(sec) or 0
 132         local msec = (min * 60000) + (sec * 1000)
 133         if msec > 0 then
 134             url = url .. "&begin=" .. msec
 135         end
 136     end
 137
 138     self.url = {url}
 139     return self
 140 end
 141
 142 function YouTube.choose_best(formats) -- Highest quality available
 143     local r = {width=0, height=0, url=nil}
 144     local U = require 'quvi/util'
 145     for _,v in pairs(formats) do
 146         if U.is_higher_quality(v,r) then
 147             r = v
 148         end
 149     end
 150 --    for k,v in pairs(r) do print(k,v) end
 151     return r
 152 end
 153
 154 function YouTube.choose_default(formats)
 155     local r = formats[1] -- Either whatever YouTube returns as the first.
 156     for _,v in pairs(formats) do
 157         if v.height == 480 then -- Or, whichever is of 480p and found first.
 158             r = v
 159             break
 160         end
 161     end
 162     return r
 163 end
 164
 165 YouTube.conv_table = { -- Deprecated.
 166     -- flv
 167     flv_240p =  '5',
 168     flv_360p = '34',
 169     flv_480p = '35',
 170     -- mp4
 171      mp4_360p = '18',
 172      mp4_720p = '22',
 173     mp4_1080p = '37',
 174     mp4_3072p = '38'
 175 }
 176
 177 function YouTube.convert_deprecated_id(r_fmt)
 178     if YouTube.conv_table[r_fmt] then
 179         local s = string.format("fmt%02d_", YouTube.conv_table[r_fmt])
 180         r_fmt = r_fmt:gsub("^(%w+)_", s)
 181     end
 182     return r_fmt
 183 end
 184
 185 function YouTube.to_s(t)
 186     return string.format("fmt%02d_%sp", t.fmt_id, t.height)
 187 end
 188
 189 --[[
 190 local a = {
 191   {u='http://youtu.be/3WSQH__H1XE',             -- u=page url
 192    e='http://youtube.com/watch?v=3WSQH__H1XE'}, -- e=expected url
 193   {u='http://youtu.be/v/3WSQH__H1XE?hl=en',
 194    e='http://youtube.com/watch?v=3WSQH__H1XE'},
 195   {u='http://youtu.be/watch?v=3WSQH__H1XE',
 196    e='http://youtube.com/watch?v=3WSQH__H1XE'},
 197   {u='http://youtu.be/embed/3WSQH__H1XE',
 198    e='http://youtube.com/watch?v=3WSQH__H1XE'},
 199   {u='http://youtu.be/v/3WSQH__H1XE',
 200    e='http://youtube.com/watch?v=3WSQH__H1XE'},
 201   {u='http://youtu.be/e/3WSQH__H1XE',
 202    e='http://youtube.com/watch?v=3WSQH__H1XE'},
 203   {u='http://youtube.com/watch?v=3WSQH__H1XE',
 204    e='http://youtube.com/watch?v=3WSQH__H1XE'},
 205   {u='http://youtube.com/embed/3WSQH__H1XE',
 206    e='http://youtube.com/watch?v=3WSQH__H1XE'},
 207   {u='http://jp.youtube.com/watch?v=3WSQH__H1XE',
 208    e='http://jp.youtube.com/watch?v=3WSQH__H1XE'},
 209   {u='http://jp.youtube-nocookie.com/e/3WSQH__H1XE',
 210    e='http://jp.youtube.com/watch?v=3WSQH__H1XE'},
 211   {u='http://jp.youtube.com/embed/3WSQH__H1XE',
 212    e='http://jp.youtube.com/watch?v=3WSQH__H1XE'},
 213   {u='http://youtube.com/3WSQH__H1XE', -- invalid page url
 214    e='http://youtube.com/watch?v=3WSQH__H1XE'}
 215 }
 216 local e = 0
 217 for i,v in pairs(a) do
 218   local s = YouTube.normalize(v.u)
 219   if s ~= v.e then
 220     print('\n   input: ' .. v.u .. " (#" .. i .. ")")
 221     print('expected: '   .. v.e)
 222     print('     got: '   .. s)
 223     e = e + 1
 224   end
 225 end
 226 print((e == 0) and 'Tests OK' or ('\nerrors: ' .. e))
 227 ]]--
 228
 229 -- vim: set ts=4 sw=4 tw=72 expandtab: