func init() {
// prepare tokenizer
b, _ := data.Asset("data/english.json")
// load the training data
training, _ := sentences.LoadTraining(b)
// create the default sentence tokenizer
tokenizer = sentences.NewSentenceTokenizer(training)
}
// English customized sentence tokenizer.
func NewSentenceTokenizer(s *sentences.Storage) (*sentences.DefaultSentenceTokenizer, error) {
training := s
if training == nil {
b, err := data.Asset("data/english.json")
if err != nil {
return nil, err
}
training, err = sentences.LoadTraining(b)
if err != nil {
return nil, err
}
}
// supervisor abbreviations
abbrevs := []string{"sgt", "gov", "no"}
for _, abbr := range abbrevs {
training.AbbrevTypes.Add(abbr)
}
lang := sentences.NewPunctStrings()
word := NewWordTokenizer(lang)
annotations := sentences.NewAnnotations(training, lang, word)
ortho := &sentences.OrthoContext{
Storage: training,
PunctStrings: lang,
TokenType: word,
TokenFirst: word,
}
multiPunct := &MultiPunctWordAnnotation{
Storage: training,
TokenParser: word,
TokenGrouper: &sentences.DefaultTokenGrouper{},
Ortho: ortho,
}
annotations = append(annotations, multiPunct)
tokenizer := &sentences.DefaultSentenceTokenizer{
Storage: training,
PunctStrings: lang,
WordTokenizer: word,
Annotations: annotations,
}
return tokenizer, nil
}
