share/lua/website/wdrmaus.lua

   1
   2 -- libquvi-scripts
   3 -- Copyright (C) 2013  Guido Leisker <guido@guido-leisker.de>
   4 --
   5 -- This file is part of libquvi-scripts <http://quvi.sourceforge.net/>.
   6 --
   7 -- This library is free software; you can redistribute it and/or
   8 -- modify it under the terms of the GNU Lesser General Public
   9 -- License as published by the Free Software Foundation; either
  10 -- version 2.1 of the License, or (at your option) any later version.
  11 --
  12 -- This library is distributed in the hope that it will be useful,
  13 -- but WITHOUT ANY WARRANTY; without even the implied warranty of
  14 -- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  15 -- Lesser General Public License for more details.
  16 --
  17 -- You should have received a copy of the GNU Lesser General Public
  18 -- License along with this library; if not, write to the Free Software
  19 -- Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
  20 -- 02110-1301  USA
  21
  22 -- About
  23 --  wrdmaus.de falls into different sections each working
  24 --  quite differently
  25 --
  26 --  elefantenseite:
  27 --    flash gallery with link target to proper pages for each video
  28 --    last part of url is id, configuration.php5
  29 --    http://www.wdrmaus.de/elefantenseite/
  30 --    http://www.wdrmaus.de/elefantenseite/#/anke_tanzt_zooztiere
  31 --
  32 --  kaeptnblaubaerseite:
  33 --    single flash, users need to use html only page (html gallery)
  34 --    and copy target links of videos
  35 --    http://www.wdrmaus.de/kaeptnblaubaerseite/baerchen/tv.php5
  36 --    http://www.wdrmaus.de/kaeptnblaubaerseite/baerchen/tv.php5?mid=1&dslSrc=rtmp://gffstream.fcod.llnwd.net/a792/e2/blaubaer/flash/oink_web-m.flv&isdnSrc=rtmp://gffstream.fcod.llnwd.net/a792/e2/blaubaer/flash/oink_web-s.flv
  37 --
  38 --  sachgeschichten:
  39 --    html video gallery, users need to copy target links of videos
  40 --    http://www.wdrmaus.de/sachgeschichten/sachgeschichten/
  41 --    http://www.wdrmaus.de/sachgeschichten/sachgeschichten/sachgeschichte.php5?id=2702
  42 --
  43 --  entenseite (very few videos, not supported!):
  44 --    http://www.wdrmaus.de/enteseite/index.php5
  45 --    http://www.wdrmaus.de/enteseite/tuerenauf/index.php5?v=3#v2
  46 --
  47 --  There are more sections but they seem to provide no videos.
  48
  49
  50 local WdrMaus = {} -- Utility functions unique to this script
  51
  52 -- Identify the script.
  53 function ident(self)
  54   package.path = self.script_dir .. '/?.lua'
  55   local C      = require 'quvi/const'
  56   local r      = {}
  57   r.domain     = "wdrmaus%.de"
  58   r.formats    = "default"
  59   r.categories = C.proto_rtmp
  60   local U      = require 'quvi/util'
  61   -- there seems to be no possiblity to really
  62   -- make a decision here: we are using a less
  63   -- strict pattern here
  64   r.handles    =  U.handles(self.page_url, {r.domain})
  65   return r
  66 end
  67
  68 -- Query available formats.
  69 function query_formats(self)
  70     self.formats = 'default'
  71     return self
  72 end
  73
  74 -- Parse media URL.
  75 function parse(self)
  76   self.host_id = "wdrmaus"
  77
  78   local U = require 'quvi/url'
  79   local t = U.parse(self.page_url)
  80
  81   if t.path:match('elefantenseite') then
  82     WdrMaus.parseElefantenseite(self)
  83   elseif t.path:match('kaeptnblaubaerseite') then
  84     WdrMaus.parseKaeptnblaubaerseite(self)
  85   elseif t.path:match('sachgeschichten') then
  86     WdrMaus.parseSachgeschichten(self)
  87   end
  88
  89   return self
  90 end
  91
  92 --
  93 -- Utility functions
  94 --
  95
  96 function WdrMaus.parseElefantenseite(self)
  97   self.id = self.page_url:match('\/([%w_]-)$')
  98       or error('no match: media id')
  99
 100   local rooturl = self.page_url:match('(%w+://.+/%w+)/.*')
 101   local configuration = quvi.fetch(rooturl .. '/data/configuration.php5')
 102   local streamingServerPath = WdrMaus.getXMLvalue(configuration, 'streamingServerPath')
 103   local toc = quvi.fetch(rooturl .. '/data/tableOfContents.php5')
 104   local metadataPath = WdrMaus.getMetadataPathFromToc(toc, self.id)
 105   local metadata = quvi.fetch(rooturl .. '/' .. metadataPath);
 106
 107   streamingServerPath = string.gsub(streamingServerPath, '/$', '')
 108   self.url = { streamingServerPath
 109                .. WdrMaus.getXMLvalue(metadata, 'file') }
 110   self.title = WdrMaus.getXMLvalue(metadata, 'title')
 111   -- no idea why this url has a diffent host
 112   self.thumbnail_url = 'http://www.wdr.de/bilder/mediendb/elefant_online'
 113         .. WdrMaus.getXMLvalue(metadata, 'image')
 114   self.format = 'default'
 115
 116   return self
 117 end
 118
 119 function WdrMaus.parseKaeptnblaubaerseite(self)
 120   self.id = self.page_url:match('/([^/]-)$')
 121   local metadatasite = quvi.fetch(
 122       'http://www.wdrmaus.de/kaeptnblaubaerseite/baerchen/tv.php5')
 123   local matcher = string.gsub(self.id, '-' , '.')
 124   metadata = metadatasite:match('.*(<img.-' .. matcher .. ')')
 125   self.title = metadata:match('<p>%s-(.-)%s-<br')
 126   local thumb_rel = metadata:match('<img src="(.-)"') or ''
 127   thumb_rel = string.gsub(thumb_rel, '%.%.', '')
 128   self.thumbnail_url =
 129       'http://www.wdrmaus.de/kaeptnblaubaerseite'
 130       .. thumb_rel
 131   self.url = { self.page_url:match('.-(rtmp.-flv)') }
 132   self.format = 'default'
 133
 134   return self
 135 end
 136
 137 function WdrMaus.parseSachgeschichten(self)
 138   self.id = self.page_url:match('%d+$')
 139   local metadata = quvi.fetch(
 140       self.page_url)
 141   self.title = metadata:match('h1_019DCE">(.-)<')
 142   self.url = { metadata:match('(rtmp.-mp4)') }
 143   self.format = 'default'
 144   return self
 145 end
 146
 147 function WdrMaus.getXMLvalue(str, value)
 148   ret = str:match('<' .. value .. '>(.-)</' .. value .. '>')
 149       or error('Cannot match ' .. value)
 150   ret = ret:match('<!%[CDATA%[(.+)]]>') or ret
 151   return ret
 152 end
 153
 154 function WdrMaus.getMetadataPathFromToc(toc, id)
 155   -- the toc contains all paths for this root (f.i. elefantenkino)
 156   local page = string.match(toc, '<id>.-(' .. id .. '.-</xmlPath>)')
 157   return WdrMaus.getXMLvalue(page, 'xmlPath')
 158 end
 159
 160 -- vim: set ts=2 sw=2 tw=72 expandtab:
 161