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 local WdrMaus = {} -- Utility functions unique to this script
  50
  51 -- Identify the script.
  52 function ident(self)
  53   package.path = self.script_dir .. '/?.lua'
  54   local C      = require 'quvi/const'
  55   local r      = {}
  56   r.domain     = "wdrmaus%.de"
  57   r.formats    = "default"
  58   r.categories = C.proto_rtmp
  59   local U      = require 'quvi/util'
  60   -- there seems to be no possiblity to really
  61   -- make a decision here: we are using a less
  62   -- strict pattern here
  63   r.handles    =  U.handles(self.page_url, {r.domain})
  64   return r
  65 end
  66
  67 -- Query available formats.
  68 function query_formats(self)
  69     self.formats = 'default'
  70     return self
  71 end
  72
  73 -- Parse media URL.
  74 function parse(self)
  75   self.host_id = "wdrmaus"
  76
  77   local a = {
  78       {pat='kaeptnblaubaerseite', func=WdrMaus.parseKaeptnblaubaerseite},
  79       {pat='sachgeschichten',     func=WdrMaus.parseSachgeschichten},
  80       {pat='elefantenseite',      func=WdrMaus.parseElefantenseite}
  81   }
  82
  83   local U = require 'quvi/url'
  84   local t = U.parse(self.page_url)
  85   local s = {}
  86
  87   for _,v in pairs(a) do
  88       if t.path:match(v.pat) then return v.func(self) end
  89       table.insert(s, v.pat)
  90   end
  91
  92   error(string.format("limited support for the {%s} sections only",
  93                         table.concat(s, ',')))
  94 end
  95
  96 --
  97 -- Utility functions
  98 --
  99
 100 function WdrMaus.parseElefantenseite(self)
 101   self.id = self.page_url:match('/([%w_]-)$')
 102                 or error('no match: media ID')
 103
 104   local rooturl = self.page_url:match('(%w+://.+/%w+)/.*')
 105                       or error('no match: root url')
 106
 107   local qo = {fetch_type='config'}
 108   local configuration = quvi.fetch(rooturl .. '/data/configuration.php5', qo)
 109
 110   local streamingServerPath =
 111           WdrMaus.getXMLvalue(configuration, 'streamingServerPath')
 112
 113   local toc = quvi.fetch(rooturl .. '/data/tableOfContents.php5', qo)
 114
 115   local metadataPath = WdrMaus.getMetadataPathFromToc(toc, self.id)
 116   local metadata = quvi.fetch(rooturl .. '/' .. metadataPath, qo);
 117
 118   streamingServerPath = string.gsub(streamingServerPath, '/$', '')
 119   self.url = { streamingServerPath .. WdrMaus.getXMLvalue(metadata, 'file') }
 120   self.title = WdrMaus.getXMLvalue(metadata, 'title')
 121
 122   -- no idea why this url has a diffent host
 123   self.thumbnail_url = 'http://www.wdr.de/bilder/mediendb/elefant_online'
 124         .. WdrMaus.getXMLvalue(metadata, 'image')
 125
 126   return self
 127 end
 128
 129 function WdrMaus.parseKaeptnblaubaerseite(self)
 130   self.id = self.page_url:match('/([^/]-)$') or error('no match: media ID')
 131
 132   local qo = {fetch_type='config'}
 133   local metadatasite = quvi.fetch(
 134       'http://www.wdrmaus.de/kaeptnblaubaerseite/baerchen/tv.php5', qo)
 135
 136   local matcher = string.gsub(self.id, '-' , '.')
 137   metadata = metadatasite:match('.*(<img.-' .. matcher .. ')')
 138                 or error('no match: metadata')
 139
 140   self.title = metadata:match('<p>%s-(.-)%s-<br') or error('no match: title')
 141   local thumb_rel = metadata:match('<img src="(.-)"') or ''
 142
 143   thumb_rel = string.gsub(thumb_rel, '%.%.', '')
 144   self.thumbnail_url = 'http://www.wdrmaus.de/kaeptnblaubaerseite' ..thumb_rel
 145
 146   self.url = { self.page_url:match('.-(rtmp.-flv)')
 147                   or error('no match: media stream URL') }
 148   return self
 149 end
 150
 151 function WdrMaus.parseSachgeschichten(self)
 152   self.id = self.page_url:match('%d+$') or error('no match: media ID')
 153   local metadata = quvi.fetch(self.page_url)
 154   self.title = metadata:match('h1_019DCE">(.-)<') or error('no match: title')
 155   self.url = { metadata:match('(rtmp.-mp4)')
 156                 or error('no match: media stream URL') }
 157   return self
 158 end
 159
 160 function WdrMaus.getXMLvalue(str, value)
 161   ret = str:match('<' .. value .. '>(.-)</' .. value .. '>')
 162             or error('Cannot match ' .. value)
 163   return ret:match('<!%[CDATA%[(.+)]]>') or ret
 164 end
 165
 166 function WdrMaus.getMetadataPathFromToc(toc, id)
 167   -- the toc contains all paths for this root (f.i. elefantenkino)
 168   local page = toc:match('<id>.-(' .. id .. '.-</xmlPath>)')
 169                   or error('no match: metadata path')
 170   return WdrMaus.getXMLvalue(page, 'xmlPath')
 171 end
 172
 173 -- vim: set ts=2 sw=2 tw=72 expandtab: