Пример #1
0
// command main
func command(opt *option) error {
	var dic tokenizer.Dic
	if opt.dic == "" {
		dic = tokenizer.SysDic()
	} else {
		var err error
		dic, err = tokenizer.NewDic(opt.dic)
		if err != nil {
			return err
		}
	}
	var udic tokenizer.UserDic
	if opt.udic != "" {
		var err error
		udic, err = tokenizer.NewUserDic(opt.udic)
		if err != nil {
			return err
		}
	}
	var fp = os.Stdin
	if opt.file != "" {
		var err error
		fp, err = os.Open(opt.file)
		if err != nil {
			return err
		}
		defer fp.Close()
	}

	t := tokenizer.NewWithDic(dic)
	t.SetUserDic(udic)

	mode := tokenizer.Normal
	switch opt.mode {
	case "normal":
		mode = tokenizer.Normal
		break
	case "search":
		mode = tokenizer.Search
	case "extended":
		mode = tokenizer.Extended
	}

	scanner := bufio.NewScanner(fp)
	for scanner.Scan() {
		line := scanner.Text()
		tokens := t.Analyze(line, mode)
		for i, size := 1, len(tokens); i < size; i++ {
			tok := tokens[i]
			c := tok.Features()
			if tok.Class == tokenizer.DUMMY {
				fmt.Printf("%s\n", tok.Surface)
			} else {
				fmt.Printf("%s\t%v\n", tok.Surface, strings.Join(c, ","))
			}
		}
	}
	return scanner.Err()
}
Пример #2
0
func init() {
	_ = tokenizer.SysDic() // prepare system dictionary
}
Пример #3
0
func init() {
	tokenizer.SysDic()
}
Пример #4
0
func init() {
	rand.Seed(time.Now().UnixNano())
	_ = tokenizer.SysDic()
}