On Tue, 14 Oct 2014, Kamil Żyta wrote:
On Tue, Oct 14, 2014 at 12:32:52PM +0100, Nick Burch wrote:
On Tue, 14 Oct 2014, Kamil Żyta wrote:
All you should need to do is provide a tika-external-parsers.xml file on
your classpath (in the appropriate directory), which defines how to talk
to your command line tool. Tika will find that and wire it up to the
external parser for you
where is the appropriate directory?
# find . -name tika-external-parsers.xml
./tika-parsers/src/main/resources/org/apache/tika/parser/external/tika-external-parsers.xml
./tika-parsers/target/classes/org/apache/tika/parser/external/tika-external-parsers.xml
It's org/apache/tika/parser/external/ - the rest before is just maven
project conventions
Tika has tests for the external parser included in it, you can try
looking at those for inspiration
I can not find it
tika-core/src/test/java/org/apache/tika/parser/external/ExternalParserTest.java
sample tika-external-parsers.xml include ffmpeg for 'video/x-matroska'
Not on my version of Tika it doesn't...
If you edit that file to do so, you still need to provide the external
parsers xml file to Tika in the right place, so Tika will find it
Nick
