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     return string.format("%s_%s%s%s",
 118               (t.quality) and t.quality or 'sd',
 119               t.container,
 120               (t.encoding) and '_'..t.encoding or '',
 121               (t.height) and '_'..t.height or '')
 122 end
 123
 124 function Ard.quality_from(suffix)
 125     local q = suffix:match('%.web(%w)%.') or suffix:match('%.(%w)%.')
 126                 or suffix:match('[=%.]Web%-(%w)') -- .webs. or Web-S or .s
 127     if q then
 128         q = q:lower()
 129         local t = {s='ld', m='md', l='sd', xl='hd'}
 130         for k,v in pairs(t) do
 131             if q == k then return v end
 132         end
 133     end
 134     return q
 135 end
 136
 137 function Ard.height_from(suffix)
 138     local h = suffix:match('_%d+x(%d+)[_%.]')
 139     if h then return h..'p' end
 140 end
 141
 142 function Ard.container_from(suffix)
 143     return suffix:match('^(...):') or suffix:match('%.(...)$')
 144               or suffix:match('%.(...)$') or 'mp4'
 145 end
 146
 147 function Ard.iter_formats(page)
 148     local r = {}
 149     local s = 'mediaCollection%.addMediaStream'
 150                 .. '%(0, %d+, "(.-)", "(.-)", "%w+"%);'
 151
 152     for prefix, suffix in  page:gmatch(s) do
 153         local u = prefix .. suffix
 154         -- remove querystring
 155         u = u:match('^(.-)?') or u
 156
 157         -- .webs. or Web-S or .s.
 158         local webx = suffix:match('%.web(%w)%.') or suffix:match('%.(%w)%.')
 159                         or suffix:match('[=%.]Web%-(%w)')
 160         if webx then webx = 'web' .. webx:lower() end
 161
 162         local t = {
 163              container = Ard.container_from(suffix),
 164              encoding = suffix:match('%.(h264)%.'),
 165              quality = Ard.quality_from(suffix),
 166              height = Ard.height_from(suffix),
 167              webx = webx,
 168              url = u
 169            }
 170         table.insert(r,t)
 171     end
 172
 173     if #r == 0 then
 174         error('no media urls found')
 175     end
 176
 177     return r
 178 end
 179
 180 -- vim: set ts=4 sw=4 sts=4 tw=72 expandtab: