mikefab / clause_extractor Goto Github PK

clause_extractor receives a string and returns an hash of tense names and verbal clauses. It was created specifically for use by Audioverb.com, a search engine of subtitles. clause_extractor does not sentencize, therefore only submit one sentence at a time. 

Usage in irb: 
1.9.3-p194 :008 > require "clause_extractor"

ClauseExtractor.get_clauses("She would have grilled her. She had been grilling. I have gone.")

prints: tense:clause:word index range
conditional perfect:she would have grilled:0..3
past perfect progressive:she had been grilling:5..8
present perfect:i have gone:9..11

And returns:
 {"conditional perfect:she would have grilled:0..3"=>0..3, "past perfect progressive:she had been grilling:5..8"=>5..8, "present perfect:i have gone:9..11"=>9..11}