[developers] XML input mode for PET

[Bernd Kiefer]

Following up to the discussion about external processors, i want to
show you what Ulrich Schaefer and me developed for PET. What is still
missing is a way to specify non-input tokens, i.e, use PET to
re-construct parse trees, but i'm thinking about that. You can not
really specify feature structures in this format, but you can supply
an arbitrary list of path/value pairs to modify a lexicon feature
structure, which we thought would suffice.

The XML mode includes and extends the functionality of the YY mode and
covers all possibilities to build input tokens that are available in
PET. The morphological part follows the rule based approach, which i
personally prefer over the tokenization.

This is just a preliminary piece of documentation, though. I hope you
don't mind its quality too much.

Bernd


--
**********************************************************************
Bernd Kiefer                                            Am Blauberg 16
kiefer at dfki.de                                      66119 Saarbruecken
+49-681/302-5301 (office)                      +49-681/3904507  (home)
**********************************************************************
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cheap-doc.tex
Type: application/x-tex
Size: 17373 bytes
Desc: not available
URL: <http://lists.delph-in.net/archives/developers/attachments/20050221/2b59f5b0/attachment.tex>