This is a (currently very incomplete!) parser for the Na'vi language.

It uses Reykunyu's API (at https://reykunyu.wimiso.nl) to parse individual words in the input sentence, and then uses a nearley grammar to parse these tokens into a parse tree.

Usage

• Install nearley: sudo npm install -g nearley
• Compile the grammar: nearleyc navi.ne -o navi.js
• Parse a sentence: node runParser.js "<sentence to parse>" (or use it programmatically: tslamyu.doParse(...))

Supported grammar features

Only a very limited set of grammar features are supported at the moment:

• nouns
  • including conjugated nouns with fì-/tsa-, -fkeyk and similar affixes
  • subclauses attached to nouns by a
  • genitive attached to nouns
• pronouns
• intransitive verbs
  • subjective
• transitive verbs
  • agentive
  • patientive
  • intransitive usage of transitive verb
• copula verbs
  • with adjectives or nouns as the predicate
• modal verbs (to do: require in the dependent verb)
• si verbs
• adjectives
  • predicative usage
  • attributive usage
• adverbials
  • adverbs
  • dative
  • adpositions
• negation (ke / rä'ä)

Anything else is unsupported for now, including:

• adverbials
  • adpositions attached to a noun
  • vocatives
• topical
• numbers
• f-words
  • also taluna, krra, etc.
• interjections (kaltxì, wiya)
• san / sìk
• fte and tsnì phrases
• verb infixes
  • especially and <äp> don't change transitivity yet

Many of these can only be implemented after they are supported in navi-reykunyu first.