More for developers

Here are some more documentation if you want to know more than just make the demos.

The linguistic files

Files for OAHPA are in the ped-directory, here is used sme (North Saami) and nob (Norwegian) as example.

Generating the database of word forms, the files are in sme/src/ or sme/meta/:

  • N_smenob.xml and N_paradigms.txt (in paradigms are the tags for generation, one pair for each PoS)
  • Adv_smenob.xml, Mwe_smenob.xml (only for Leksa, no paradigms)
  • semantic_sets.xml (semantic categories of semantic sets)
  • Pron_sme.xml (containing paradigms, only for Morfa)
  • Names.xml (person names, only for Morfa)
  • fillings.xml (words of other POS)
  • tags.txt
  • gt/bin/isme-GG.restr.fst (for generation of the western dialect)
  • gt/bin/isme-KJ.restr.fst (for generation of the eastern dialect)
  • gt/bin/isme-norm.fst (for generation of the key answers)

Read more about generation of word forms.

Files especially for Leksa, from Norwegian to North Sami, in nobsrc/:

  • A_nobsme.xml, (for each PoS)

Automata for Numra:

  • sme-Num.fst

Files for generating tasks in Morfa-C and Vasta, in sme/meta/:

  • N_questions.xml (one file for each PoS in Morfa C)
  • vastas_questions.xml
  • grammar_defaults.xml (takes care of agreement in Morfa C tasks)

Dialogue files for Sahka, in sme/meta/:

  • dialogue_firstmeeting.xml (one of many dialogues)
  • dialogues.dtd

Files for giving feedback about morphology in Morfa-S:

The feedback-files add feedbacktags based on a combination of the morphological features specified in the lexicon, and the task itself. With help of the tags the system generates feedback messages for the user. In sme/meta/:

  • feedback_N.xml (one file for each Morfa C PoS
  • messages.xml (one file for each metalanguage)

Other files offering help to the user, in sme/meta/:

comments.xml (nob, eng, sme, swe, fin) (comments to the user aobut how she is succeeding with the tasks.)

grammatikklinker.txt (nob, eng, sme, swe, fin) (links to the grammar documentation. The grammar files are in userdoc/)

Read more about feedback.

Files for analysing the free input in Sahka and Vasta:

ped-sme.fst gives morphological analysis to the user's input. It is normative, but it allows proper nouns written with small initial letter, but marks it with an error tag.

sme-ped.cg3 disambuguates and adds grammar error and navigation tags to the user's input.

Files for giving feedback in Vasta and Sahka:

The feedback messages are generated with help of the grammartags in the analysis done with the vislcg3-file, in sme/meta/:

  • messages_vasta.xml, (nob, eng, sme, swe, fin)

Localisation

Documentation is here

System files

In univ_oahpa/ are files system files:

  • installation scripts
  • media/ : images
  • templates/ : html-files for basic look for all the programs
  • drill/ : scripts for each program
  • drill/templates/ : html-files for each program
  • locale/ : localisation

In userdoc/ are files for userdocumentation and the North Sami grammar.