I am following the instructions at the How_to_bootstrap_a_new_pair and
The_quick_and_dirty_guide_to_making_a_new_language_pair wikis using
HFST for both languages. After a bit of tweaking the following test
command in the bilingual directory works:
echo “Houses” | apertium -d . XXX-YYY
Casas
echo “Casas” | apertium -d . YYY-XXX
Houses
I then attempted to follow the instructions at the
Lexc_and_flag_diacritics_for_prefix_tagging wiki in order to manage
the prefixing. The tests above continue to work (I didn’t remove their
respective lines from the .lexc files). However, none of the new noun
roots work. E.g.
echo “Abambo” | apertium -d . XXX-YYY
* Abambo
echo “Acibaba” | apertium -d . YYY-XXX
* Acibaba
When I run a few hfst commands in each monolingual directory, however,
it shows that the required reordering of the POS tags is working. E.g.
echo “Abambo” | hfst-proc -w XXX.automorph.hfst
^Abambo/bambo<n><pl>$
echo “Acibaba” | hfst-proc -w YYY.automorph.hfst
^Acibaba/baba<n><pl>$
To continue debugging, I would like to go step by step through the
rest of the translation pipeline in order to pinpoint where the
problem is (using hfst-proc instead of lt-proc as defined in modes.xml
file)! However, there are no *.hfst files in the XXX-YYY bilingual
directory! And when I try:
echo “Acibaba” | hfst-proc -w YYY-XXX.automorf.bin
I get the following error:
terminate called after throwing an instance of 'TransducerHasWrongTypeException'
How do I generate *.hfst files in the XXX-YYY bilingual directory? Or,
what is it that I am missing?
_______________________________________________
Apertium-stuff mailing list
Apertium-stuff@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/apertium-stuff
