This EPG grabber or "scraper" allows extracting TV programme listings in XMLTV format from teletext pages as broadcast by most European TV networks. The grabber works by reading teletext pages via a TV card from all individual networks, then extracting programme title and starting times from programme tables (e.g. in Germany usually pages 301-304) and finally adding description texts from pages which are referenced in the overviews.

The grabber can be used stand-alone, but is best used as part of the Nextview EPG grabber nxtvepg (version 2.9.0 or later), which can also be used for immediately browsing the programme listings in the generated XMLTV files. The grabber is integrated into the nxtvepg acquisition control and can be activated via the configuration menus. (Version 2.9 of nxtvepg still defaults to acquisition via Nextview EPG, but that service is defunct now; use nxtvepg version 3.0 or later for optimized support.)

Since the grabber has to read each individual network's teletext stream and many networks use sub-pages at least for descriptions, the process if done sequentially takes relatively long, namely at least 90-120 seconds per network for grabbing all data including description texts. However, when capturing from DVB, it is possible to capture from all channels sharing a transponder, which in average speeds up acquisition by more than factor 5 (see description of script capall.sh below). There is also a fast scan mode which requires only 10-20 seconds per network in average, however in this mode overview and descriptions may be incomplete if sub-pages are used by the respective network.

For now, only nation-wide German networks are supported well. For other networks, the parser would need to be adapted to different page formats. That's because almost every network formats tables, dates, times and descriptions slightly differently, so that the parser will have trouble locating the EPG data among all the advertisements and other content in teletext.

Development was resumed with release 2.0 after a 9 year gap. The first release mainly focused on enhancing working with digital TV capture cards (e.g. by adapting the capall.pl example automation script). Also tools have been updated (e.g. removed the need for the Boost regex library by requiring C 11) and the parser was slightly enhanced and adapted to changes in teletext content by networks.

Subsequent releases provided further enhancements and fixes in teletext page parser and tools (e.g. DVB channel tuning), as well as interface adaption for integration in nxtvepg. See file "changelog" for details.

If you got the Debian package (or created one yourself using shell script mkdebian.pl), you can install the executable and manual page by typing:

    dpkg -i PACKAGE

Alternatively, you can compile and install the package from sources directly by typing make install prefix=/usr/local (optionally replacing /usr/local with another directory which contains bin/ and man/, where to install the executable and manual page respectively.)

For use with nxtvepg: The grabber is actually compiled into the nxtvepg executable: Both nxtvepg release 2.9.0 and 3.0.0 include the current latest grabber release 2.3. See nxtvepg grabber documentation.

This grabber captures EPG data from live TV broadcast, therefore a TV capture device (e.g. a DVB capture device) is required.

When not used from within nxtvepg, the grabber requires the ZVBI library for capturing teletext from a TV capture device. The library is available for Linux and BSD only. Most Linux distributions provide the library as a package (e.g. for Debian, Ubuntu et.al. use apt-get install libzvbi-dev).

For building the executable, a C compiler supporting release C 11 or later is required.

Note up to version 1.x of this package Perl and the Video::ZVBI Perl extension module were used for implementation (actually that extension was developed for this purpose). Due to increased size and complexity of the implementation however, C is used instead since version 2.0, so these dependencies are moot.

The following is a usage example: First the channel is switched to network ARD using mplayer (you could use other tools such as dvb5-zap for this purpose, depending on your DVB setup.) Then the grabber is run: With the shown options, it will capture pages 300-399 from DVB stream 104 (which is the teletext PID of network "Das Erste", formerly known as "ARD"; you can obtain the PID using tool tune_dvb in this package). Capturing will stop after all sub-pages of the page range have been acquired. Afterwards the grabber extracts EPG data and generates an XMLTV file which is written to stdout, which is redirected to a file in the example. Finally the XMLTV file is loaded into nxtvepg for browsing the programme schedules. (Any other XMLTV-capable browser will work, too.)

  mplayer dvb://"Das Erste" &
  ./tv_grab_ttx -dvbpid 104 -page 301-399 > ttx-ARD.xml
  nxtvepg ttx-ARD.xml

For additional options invoke the grabber with command line option "-help".

Using the above will produce a separate XMLTV file per channel, which is not practical for browsing. Use script merge.pl for merging XML files of all channels into a single XMLTV file. When using the grabber from within nxtvepg, the merge can be configured to be done automatically.

See script capall.pl for an example how to automate capturing and merging EPG data from multiple channels. (Note this script requires make tune_dvb for tuning into channels and you will have to edit it for your channels.conf path and channel names used therein.) In particular the script speeds-up grabbing significantly by running tv_grab_ttx multiple times concurrently for all DVB channels sharing a transponder.

Copyright 2006-2011, 2020-2021 by T. Zoerner

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation. This program is distributed in the hope that it will be useful, but without any warranty; without even the implied warranty of merchantability or fitness for a particular purpose.

Content copyright: Please note that content providers do hold a copyright on the programme schedules which can be extracted by means of this software. The data is free for personal use, but must not be publicly redistributed (e.g. made available on the Internet) without prior permission.