[mythtv-users] Q for XML guru - extracting selected channel info from XMLTV grabber file

[Nick]

BACKGROUND

If mythfilldatabase is run against a video source using an XML
listings file containing channels not defined for the source, it
automatically creates new channels with the XMLTVID information from
the XML listings. Questions about this non-intuitive behaviour
frequently pop up on this list.

SCENARIO

Currently I use 3 separate runs of the grabber for each source type,
which means duplication in connecting to the BBC server for listings
information, and duplication in the listings files. Ideally I would
like to be able to just have to run the grabber and MFDB once, and
have MFDB use the file intelligently, so that it notes which XMLTVIDs
are available for each video source it is run against and only populas
listings data based on this.

DESIRED FUNCTIONALITY

I want to be able to use a single XML listings file (that contains
information for all channels on all sources) and then split out this
information on a per source basis so that this works with the current
MFDB behaviour (having an option --dont-create-channels would be
better IMO though) This way a single XML file can be used not just
with MythTV, but with other XML based apps.

In the meantime, is there an easy shell-based XML method to extract
listings info from a single source file, based on a given list of
XMLTVID for each source, so that MFDB can have individual files for
each source and avoid the channel creation issues? I could probably
hack something very kludgy in Java to do this, but that means learning
XML parsing first ...

Cheers,
Nick