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