share/lua/website/ard.lua

   1 -- libquvi-scripts
   2 -- Copyright (C) 2013  Thomas Weißschuh
   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 local Ard = {}
  22
  23 function ident(self)
  24     package.path = self.script_dir .. '/?.lua'
  25     local C      = require 'quvi/const'
  26     local U      = require 'quvi/util'
  27     local B      = require 'quvi/bit'
  28     local r      = {}
  29     r.domain     = 'www%.ardmediathek%.de'
  30     r.formats    = 'default|best'
  31     r.categories = B.bit_or(C.proto_http, C.proto_rtmp)
  32     r.handles    = U.handles(self.page_url, {r.domain},
  33                                nil, {"documentId=%d+$"})
  34     return r
  35 end
  36
  37 function query_formats(self)
  38     local config = Ard.get_config(self)
  39     local formats = Ard.iter_formats(config)
  40
  41     local t = {}
  42     for _,v in pairs(formats) do
  43         table.insert(t, Ard.to_s(v))
  44     end
  45
  46     table.sort(t)
  47     self.formats = table.concat(t, "|")
  48
  49     return self
  50 end
  51
  52 function parse(self)
  53
  54     local config = Ard.get_config(self)
  55     local Util  = require 'quvi/util'
  56
  57     self.host_id = 'ard'
  58     self.title = config:match(
  59                      '<meta property="og:title" content="([^"]*)'
  60                  ):gsub(
  61                     '%s*%- %w-$', '' -- remove name of station
  62                  ):gsub(
  63                     '%s*%(FSK.*', '' -- remove FSK nonsense
  64                  )
  65                  or error('no match: media title')
  66     self.thumbnail_url = config:match(
  67                              '<meta property="og:image" content="([^"]*)'
  68                          ) or ''
  69
  70     local formats = Ard.iter_formats(config)
  71     local format  = Util.choose_format(self,
  72                                        formats,
  73                                        Ard.choose_best,
  74                                        Ard.choose_default,
  75                                        Ard.to_s)
  76                     or error('unable to choose format')
  77
  78     if not format.url then error('no match: media url') end
  79     self.url = { format.url }
  80
  81     return self
  82 end
  83
  84 function Ard.test_availability(page)
  85     -- some videos are only scrapable at certain times
  86     local fsk_pattern =
  87         'Der Clip ist deshalb nur von (%d%d?) bis (%d%d?) Uhr verfügbar'
  88     local from, to = page:match(fsk_pattern)
  89     if from and to then
  90         error('video only available from ' ..from.. ':00 to '
  91               ..to.. ':00 CET')
  92     end
  93 end
  94
  95 function Ard.get_config(self)
  96     local c = quvi.fetch(self.page_url)
  97     self.id = self.page_url:match('documentId=(%d*)')
  98               or error('no match: media id')
  99     if c:match('<title>ARD Mediathek %- Fehlerseite</title>') then
 100         error('invalid URL, maybe the media is no longer available')
 101     end
 102
 103     return c
 104 end
 105
 106 function Ard.choose_best(t)
 107     return t[#t] -- return the last from the array
 108 end
 109
 110 function Ard.choose_default(t)
 111     return t[1] -- return the first from the array
 112 end
 113
 114 function Ard.to_s(t)
 115     return string.format("%s_%s_i%02d%s%s",
 116               (t.quality) and t.quality or 'sd',
 117               t.container, t.stream_id,
 118               (t.encoding) and '_'..t.encoding or '',
 119               (t.height) and '_'..t.height or '')
 120 end
 121
 122 function Ard.quality_from(suffix)
 123     local q = suffix:match('%.web(%w)%.') or suffix:match('%.(%w)%.')
 124                 or suffix:match('[=%.]Web%-(%w)') -- .webs. or Web-S or .s
 125     if q then
 126         q = q:lower()
 127         local t = {s='ld', m='md', l='sd', xl='hd'}
 128         for k,v in pairs(t) do
 129             if q == k then return v end
 130         end
 131     end
 132     return q
 133 end
 134
 135 function Ard.height_from(suffix)
 136     local h = suffix:match('_%d+x(%d+)[_%.]')
 137     if h then return h..'p' end
 138 end
 139
 140 function Ard.container_from(suffix)
 141     return suffix:match('^(...):') or suffix:match('%.(...)$') or 'mp4'
 142 end
 143
 144 function Ard.iter_formats(page)
 145     local r = {}
 146     local s = 'mediaCollection%.addMediaStream'
 147                 .. '%(0, (%d+), "(.-)", "(.-)", "%w+"%);'
 148
 149     Ard.test_availability(page)
 150
 151     for s_id, prefix, suffix in  page:gmatch(s) do
 152         local u = prefix .. suffix
 153         u = u:match('^(.-)?') or u  -- remove querystring
 154         local t = {
 155             container = Ard.container_from(suffix),
 156             encoding = suffix:match('%.(h264)%.'),
 157             quality = Ard.quality_from(suffix),
 158             height = Ard.height_from(suffix),
 159             stream_id = s_id, -- internally (by service) used stream ID
 160             url = u
 161         }
 162         table.insert(r,t)
 163     end
 164     if #r == 0 then error('no media urls found') end
 165     return r
 166 end
 167
 168 -- vim: set ts=4 sw=4 sts=4 tw=72 expandtab: