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+$"})
  33     return r
  34 end
  35
  36 function query_formats(self)
  37     local config = Ard.get_config(self)
  38     local formats = Ard.iter_formats(config)
  39
  40     local t = {}
  41     for _,v in pairs(formats) do
  42         table.insert(t, Ard.to_s(v))
  43     end
  44
  45     table.sort(t)
  46     self.formats = table.concat(t, "|")
  47
  48     return self
  49 end
  50
  51 function parse(self)
  52
  53     local config = Ard.get_config(self)
  54     local Util  = require 'quvi/util'
  55
  56     Ard.test_availability(config)
  57
  58     self.host_id = 'ard'
  59     self.title = config:match(
  60                      '<meta property="og:title" content="([^"]*)'
  61                  ):gsub(
  62                     '%s*%- %w-$', '' -- remove name of station
  63                  ):gsub(
  64                     '%s*%(FSK.*', '' -- remove FSK nonsense
  65                  )
  66                  or error('no match: media title')
  67     self.thumbnail_url = config:match(
  68                              '<meta property="og:image" content="([^"]*)'
  69                          ) or ''
  70
  71     local formats = Ard.iter_formats(config)
  72     local format  = Util.choose_format(self,
  73                                        formats,
  74                                        Ard.choose_best,
  75                                        Ard.choose_default,
  76                                        Ard.to_s)
  77                     or error('unable to choose format')
  78
  79     if not format.url then error('no match: media url') end
  80     self.url = { format.url }
  81
  82     return self
  83 end
  84
  85 function Ard.test_availability(page)
  86     -- some films are only scrapable at certain times
  87     local fsk_pattern =
  88         'Der Clip ist deshalb nur von (%d%d?) bis (%d%d?) Uhr verfügbar'
  89     local fsk = {}
  90     fsk.from, fsk.to = page:match(fsk_pattern)
  91     if fsk.from and fsk.to then
  92         error('film only available from ' ..fsk.from.. ':00 to '
  93               ..fsk.to.. ':00 CET (UTC+2)')
  94     end
  95 end
  96
  97 function Ard.get_config(self)
  98     local c = quvi.fetch(self.page_url)
  99     self.id = self.page_url:match('documentId=(%d*)')
 100               or error('no match: media id')
 101     if c:match('<title>ARD Mediathek %- Fehlerseite</title>') then
 102         error('invalid URL, maybe the media is no longer available')
 103     end
 104
 105     return c
 106 end
 107
 108 function Ard.choose_best(t)
 109     return t[#t] -- return the last from the array
 110 end
 111
 112 function Ard.choose_default(t)
 113     return t[1] -- return the first from the array
 114 end
 115
 116 function Ard.to_s(t)
 117     s = t.container
 118
 119     if t.encoding then
 120         s = s .. '_' .. t.encoding
 121     end
 122
 123     if t.height then
 124         s = s .. '_' .. t.height
 125     elseif t.webx then
 126         s = s .. '_' .. t.webx
 127     else
 128         s = s .. '_' .. t.quality
 129     end
 130
 131     return s
 132 end
 133
 134 function Ard.iter_formats(page)
 135     local r = {}
 136     local s = 'mediaCollection%.addMediaStream'
 137                 .. '%(0, (%d), "(.-)", "(.-)", "%w+"%);'
 138
 139     for quality, prefix, suffix in  page:gmatch(s) do
 140         local u = prefix .. suffix
 141         -- remove querystring
 142         u = u:match('^(.-)?') or u
 143
 144         -- .webs. or Web-S or .s.
 145         local webx = suffix:match('%.web(%w)%.') or suffix:match('%.(%w)%.')
 146                         or suffix:match('[=%.]Web%-(%w)')
 147         if webx then webx = 'web' .. webx:lower() end
 148
 149         local height = suffix:match('_%d+x(%d+)[_%.]')
 150         if height then height = 'p' .. height end
 151
 152         local t = {
 153              url = u,
 154              container = suffix:match('^(...):') or suffix:match('%.(...)$')
 155                          or suffix:match('%.(...)$') or 'mp4',
 156              quality = tonumber(quality),
 157              encoding = suffix:match('%.(h264)%.'),
 158              height = height,
 159              webx = webx
 160            }
 161         table.insert(r,t)
 162     end
 163
 164     if #r == 0 then
 165         error('no media urls found')
 166     end
 167
 168     return r
 169 end
 170
 171 -- vim: set ts=4 sw=4 sts=4 tw=72 expandtab: