Sha256: 46f49d2083f031afdaabcdca49b8cce3b7199147af4b559ce3ae3815d1bab905
Contents?: true
Size: 862 Bytes
Versions: 18
Compression:
Stored size: 862 Bytes
Contents
module PragmaticSegmenter
module Languages
class Armenian
class Process < PragmaticSegmenter::Process
private
def sentence_boundary_punctuation(txt)
PragmaticSegmenter::Languages::Armenian::SentenceBoundaryPunctuation.new(text: txt).split
end
def punctuation_array
PragmaticSegmenter::Languages::Armenian::Punctuation.new.punct
end
end
class Cleaner < PragmaticSegmenter::Cleaner
end
class SentenceBoundaryPunctuation < PragmaticSegmenter::SentenceBoundaryPunctuation
SENTENCE_BOUNDARY = /.*?[։՜:]|.*?$/
def split
text.scan(SENTENCE_BOUNDARY)
end
end
class Punctuation < PragmaticSegmenter::Punctuation
PUNCT = ['։', '՜', ':']
def punct
PUNCT
end
end
end
end
end
Version data entries
18 entries across 18 versions & 1 rubygems