share/common/quvi/youtube.lua

   1 -- libquvi-scripts
   2 -- Copyright (C) 2012-2013  Toni Gundogdu <legatvs@gmail.com>
   3 --
   4 -- This file is part of libquvi-scripts <http://quvi.sourceforge.net/>.
   5 --
   6 -- This program is free software: you can redistribute it and/or
   7 -- modify it under the terms of the GNU Affero General Public
   8 -- License as published by the Free Software Foundation, either
   9 -- version 3 of the License, or (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 Affero General Public License for more details.
  15 --
  16 -- You should have received a copy of the GNU Affero General
  17 -- Public License along with this program.  If not, see
  18 -- <http://www.gnu.org/licenses/>.
  19 --
  20
  21 local M = {}
  22
  23 --[[
  24 Return the `ident' data for the {media,subtitle} scripts.
  25 Parameters:
  26   qargs .. quvi args
  27 Returns:
  28   A table containing the values expected by the library.
  29 ]]--
  30 function M.ident(qargs)
  31   local u = M.normalize(qargs.input_url)
  32   return {
  33     domains = table.concat({'youtube.com'}, ','),
  34     can_parse_url = M.can_parse_url(u)
  35   }
  36 end
  37
  38 --[[
  39 Check if script can parse the URL.
  40 Parameters:
  41   url .. URL to check
  42 Returns:
  43   A boolean value.
  44 ]]--
  45 function M.can_parse_url(url)
  46   local U = require 'socket.url'
  47   local t = U.parse(url)
  48   if t and t.scheme and t.scheme:lower():match('^https?$')
  49        and t.host   and t.host:lower():match('youtube%.com$')
  50        and t.query  and t.query:lower():match('v=[%w-_]+')
  51        and t.path   and t.path:lower():match('^/watch$')
  52   then
  53     return true
  54   else
  55     return false
  56   end
  57 end
  58
  59 --[[
  60 "Normalize" URL to YouTube media URL. See the test URLs for examples.
  61 Parameters:
  62   s  .. URL to normalize
  63 Returns:
  64   Normalized URL
  65 ]]--
  66 function M.normalize(url)
  67   -- Leave if url is undefined for some reason.
  68   if not url then
  69     return url
  70   end
  71   local U = require 'socket.url'
  72   local u = U.parse(url)
  73   -- Leave if parsing fails for some reason.
  74   if not u.host or not u.path then
  75     return url
  76   end
  77   -- Unroll youtu.be
  78   u.host = u.host:gsub('youtu%.be', 'youtube.com')
  79   -- Process URLs with the youtube domain name only.
  80   if not u.host:match('youtube%.com$') then
  81     return url
  82   end
  83   -- Try to lookup the video/media ID.
  84   for _,p in pairs({'/embed/([-_%w]+)', '/%w/([-_%w]+)', '/([-_%w]+)'}) do
  85     local v_id = u.path:match(p)
  86     if v_id and #v_id ==11 then -- Convert the URL into a YouTube media URL.
  87       u.query = table.concat({'v=',v_id})
  88       u.path = '/watch'
  89     end
  90   end
  91   return U.build(u) -- Rebuild and return the media URL.
  92 end
  93
  94 --[[
  95 Append URL to qargs.media_url if it is unique by comparing video IDs.
  96 Parameters:
  97   url .. URL to append
  98 ]]--
  99 function M.append_if_unique(qargs, url)
 100   if not url then return end
 101
 102   url = M.normalize(url)
 103
 104   local U = require 'socket.url'
 105   local t = U.parse(url)
 106
 107   if not t.host or not t.query then return end
 108
 109   local p = 'v=([%w-_]+)'
 110   local v = t.query:match(p)
 111
 112   for _,u in pairs(qargs.media_url) do
 113     local tt = U.parse(u)
 114     if tt.query and  v == tt.query:match(p) then
 115       return -- Found duplicate. Ignore URL.
 116     end
 117   end
 118
 119   table.insert(qargs.media_url, url)
 120 end
 121
 122 --
 123 -- Tests
 124 --
 125
 126 --[[
 127 local function test_normalize()
 128   local test_cases = {
 129     {url = 'http://youtu.be/3WSQH__H1XE',
 130      expect = 'http://youtube.com/watch?v=3WSQH__H1XE'},
 131     {url = 'https://youtu.be/3WSQH__H1XE',
 132      expect = 'https://youtube.com/watch?v=3WSQH__H1XE'},
 133     {url='http://youtu.be/v/3WSQH__H1XE?hl=en',
 134      expect='http://youtube.com/watch?v=3WSQH__H1XE'},
 135     {url='http://youtu.be/watch?v=3WSQH__H1XE',
 136      expect='http://youtube.com/watch?v=3WSQH__H1XE'},
 137     {url='http://youtu.be/embed/3WSQH__H1XE',
 138      expect='http://youtube.com/watch?v=3WSQH__H1XE'},
 139     {url='http://youtu.be/v/3WSQH__H1XE',
 140      expect='http://youtube.com/watch?v=3WSQH__H1XE'},
 141     {url='http://youtu.be/e/3WSQH__H1XE',
 142      expect='http://youtube.com/watch?v=3WSQH__H1XE'},
 143     {url='http://youtube.com/watch?v=3WSQH__H1XE',
 144      expect='http://youtube.com/watch?v=3WSQH__H1XE'},
 145     {url='http://youtube.com/embed/3WSQH__H1XE',
 146      expect='http://youtube.com/watch?v=3WSQH__H1XE'},
 147     {url='http://jp.youtube.com/watch?v=3WSQH__H1XE',
 148      expect='http://jp.youtube.com/watch?v=3WSQH__H1XE'},
 149     {url='http://jp.youtube.com/embed/3WSQH__H1XE',
 150      expect='http://jp.youtube.com/watch?v=3WSQH__H1XE'},
 151     {url='https://jp.youtube.com/embed/3WSQH__H1XE',
 152      expect='https://jp.youtube.com/watch?v=3WSQH__H1XE'},
 153     {url='http://youtube.com/3WSQH__H1XE', -- invalid page url
 154      expect='http://youtube.com/watch?v=3WSQH__H1XE'}
 155   }
 156   local i,e = 0,0
 157   for _,v in pairs(test_cases) do
 158     local r = M.normalize(v.url)
 159     if r ~= v.expect then
 160       print(string.format('input: %s (#%s)\nexpected: %s\ngot: %s',
 161                           v.url, i, v.expect, r))
 162       e = e+1
 163     end
 164     i = i+1
 165   end
 166   print((e==0) and 'tests OK' or error('failed tests: '..e))
 167 end
 168
 169 test_normalize()
 170 ]]--
 171
 172 return M
 173
 174 -- vim: set ts=2 sw=2 tw=72 expandtab: