Sha256: 11af4363ed9d94c8b14cc9f2bb653f159db1c37cb9644849b75ae0fd4b41f00a
Contents?: true
Size: 870 Bytes
Versions: 18
Compression:
Stored size: 870 Bytes
Contents
module PragmaticSegmenter
module Languages
class Burmese
class Process < PragmaticSegmenter::Process
private
def sentence_boundary_punctuation(txt)
PragmaticSegmenter::Languages::Burmese::SentenceBoundaryPunctuation.new(text: txt).split
end
def punctuation_array
PragmaticSegmenter::Languages::Burmese::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