Hello - I have tried to post this on chat, but I am being frequently disconnected, so I've decided to use the mailing list.
I need to add some kind of a "choose" parameter to a morphological dictionary or a rule. What I am trying to do is make the system translate first and second person of "widzisz" as "see", and the third person as "sees". At the moment the third person works correctly, but the first and second result in "#widzieć" echo "see" | lt-proc en-anal.bin | ./gawk | apertium-transfer trans-enpl.t1x trans-enpl.bin en-to-pl.bin | lt-proc -g pl-gen.bin #widzieć The third person however, works fine echo "sees" | lt-proc en-anal.bin | ./gawk | apertium-transfer trans-enpl.t1x trans-enpl.bin en-to-pl.bin | lt-proc -g pl-gen.bin widzi The same situation appears with nouns - I have added inflected forms of three cases (Nom, Gen, Acc) in Polish, which all correspond to one equivalent (house) in English - everything goes well in the analysis echo "house" | lt-proc en-anal.bin | ./gawk | apertium-transfer trans-enpl.t1x trans-enpl.bin en-to-pl.bin ^dom<n><sg>$ but in generation I get an error (#dom) echo "house" -anal.bin | ./gawk | apertium-transfer trans-enpl.t1x trans-enpl.bin en-to-pl.bin | lt-proc -g pl-gen.bin #dom I suppose this is somehow connected with the fact, that in analysis I get only ^dom<n><sg> while in the Polish morphological dictionary there is also an entry for the case (<s n="acc"/>). My transfer rule file - http://apertium.codepad.org/l56nJHhx Polish morph dict - http://codepad.org/mwiLzKxt English morph dict - http://codepad.org/udyfbCRj Thanks in advance Regards ------------------------------------------------------------------------------ Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev _______________________________________________ Apertium-stuff mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/apertium-stuff
