Help us shape the future of Go! Take the 2026 Go Developer Survey

gojieba

package module
v1.0.1 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Sep 8, 2019 License: MIT Imports: 5 Imported by: 0

README

GoJieba English

Build Status Author Performance License GoDoc Coverage Status codebeat badge Go Report Card Awesome

logo

GoJieba是"结巴"中文分词的Golang语言版本。

简介

  • 支持多种分词方式，包括: 最大概率模式, HMM新词发现模式, 搜索引擎模式, 全模式
  • 核心算法底层由C++实现，性能高效。
  • 无缝集成到 bleve 到进行搜索引擎的中文分词功能。
  • 字典路径可配置，NewJieba(...string), NewExtractor(...string) 可变形参，当参数为空时使用默认词典(推荐方式)

用法

go get github.com/yanyiwu/gojieba

分词示例

package main

import (
	"fmt"
	"strings"

	"github.com/yanyiwu/gojieba"
)

func main() {
	var s string
	var words []string
	use_hmm := true
	x := gojieba.NewJieba()
	defer x.Free()

	s = "我来到北京清华大学"
	words = x.CutAll(s)
	fmt.Println(s)
	fmt.Println("全模式:", strings.Join(words, "/"))

	words = x.Cut(s, use_hmm)
	fmt.Println(s)
	fmt.Println("精确模式:", strings.Join(words, "/"))
	s = "比特币"
	words = x.Cut(s, use_hmm)
	fmt.Println(s)
	fmt.Println("精确模式:", strings.Join(words, "/"))

	x.AddWord("比特币")
	s = "比特币"
	words = x.Cut(s, use_hmm)
	fmt.Println(s)
	fmt.Println("添加词典后,精确模式:", strings.Join(words, "/"))

	s = "他来到了网易杭研大厦"
	words = x.Cut(s, use_hmm)
	fmt.Println(s)
	fmt.Println("新词识别:", strings.Join(words, "/"))

	s = "小明硕士毕业于中国科学院计算所，后在日本京都大学深造"
	words = x.CutForSearch(s, use_hmm)
	fmt.Println(s)
	fmt.Println("搜索引擎模式:", strings.Join(words, "/"))

	s = "长春市长春药店"
	words = x.Tag(s)
	fmt.Println(s)
	fmt.Println("词性标注:", strings.Join(words, ","))

	s = "区块链"
	words = x.Tag(s)
	fmt.Println(s)
	fmt.Println("词性标注:", strings.Join(words, ","))

	s = "长江大桥"
	words = x.CutForSearch(s, !use_hmm)
	fmt.Println(s)
	fmt.Println("搜索引擎模式:", strings.Join(words, "/"))

	wordinfos := x.Tokenize(s, gojieba.SearchMode, !use_hmm)
	fmt.Println(s)
	fmt.Println("Tokenize:(搜索引擎模式)", wordinfos)

	wordinfos = x.Tokenize(s, gojieba.DefaultMode, !use_hmm)
	fmt.Println(s)
	fmt.Println("Tokenize:(默认模式)", wordinfos)

	keywords := x.ExtractWithWeight(s, 5)
	fmt.Println("Extract:", keywords)
}

我来到北京清华大学
全模式: 我/来到/北京/清华/清华大学/华大/大学
我来到北京清华大学
精确模式: 我/来到/北京/清华大学
比特币
精确模式: 比特/币
比特币
添加词典后,精确模式: 比特币
他来到了网易杭研大厦
新词识别: 他/来到/了/网易/杭研/大厦
小明硕士毕业于中国科学院计算所，后在日本京都大学深造
搜索引擎模式: 小明/硕士/毕业/于/中国/科学/学院/科学院/中国科学院/计算/计算所/，/后/在/日本/京都/大学/日本京都大学/深造
长春市长春药店
词性标注: 长春市/ns,长春/ns,药店/n
区块链
词性标注: 区块链/nz
长江大桥
搜索引擎模式: 长江/大桥/长江大桥
长江大桥
Tokenize: [{长江 0 6} {大桥 6 12} {长江大桥 0 12}]

See example in jieba_test, extractor_test

Bleve 中文分词插件用法

package main

import (
	"fmt"
	"os"

	"github.com/blevesearch/bleve"
	"github.com/yanyiwu/gojieba"
	_ "github.com/yanyiwu/gojieba/bleve"
)

func Example() {
	INDEX_DIR := "gojieba.bleve"
	messages := []struct {
		Id   string
		Body string
	}{
		{
			Id:   "1",
			Body: "你好",
		},
		{
			Id:   "2",
			Body: "世界",
		},
		{
			Id:   "3",
			Body: "亲口",
		},
		{
			Id:   "4",
			Body: "交代",
		},
	}

	indexMapping := bleve.NewIndexMapping()
	os.RemoveAll(INDEX_DIR)
	// clean index when example finished
	defer os.RemoveAll(INDEX_DIR)

	err := indexMapping.AddCustomTokenizer("gojieba",
		map[string]interface{}{
			"dictpath":     gojieba.DICT_PATH,
			"hmmpath":      gojieba.HMM_PATH,
			"userdictpath": gojieba.USER_DICT_PATH,
			"idf":          gojieba.IDF_PATH,
			"stop_words":   gojieba.STOP_WORDS_PATH,
			"type":         "gojieba",
		},
	)
	if err != nil {
		panic(err)
	}
	err = indexMapping.AddCustomAnalyzer("gojieba",
		map[string]interface{}{
			"type":      "gojieba",
			"tokenizer": "gojieba",
		},
	)
	if err != nil {
		panic(err)
	}
	indexMapping.DefaultAnalyzer = "gojieba"

	index, err := bleve.New(INDEX_DIR, indexMapping)
	if err != nil {
		panic(err)
	}
	for _, msg := range messages {
		if err := index.Index(msg.Id, msg); err != nil {
			panic(err)
		}
	}

	querys := []string{
		"你好世界",
		"亲口交代",
	}

	for _, q := range querys {
		req := bleve.NewSearchRequest(bleve.NewQueryStringQuery(q))
		req.Highlight = bleve.NewHighlight()
		res, err := index.Search(req)
		if err != nil {
			panic(err)
		}
		fmt.Println(res)
	}
}

func main() {
	Example()
}

Output:

2 matches, showing 1 through 2, took 360.584µs
    1. 2 (0.423287)
    Body
        <mark>世界</mark>
    2. 1 (0.423287)
    Body
        <mark>你好</mark>

2 matches, showing 1 through 2, took 131.055µs
    1. 4 (0.423287)
    Body
        <mark>交代</mark>
    2. 3 (0.423287)
    Body
        <mark>亲口</mark>

See example in bleve_test

性能评测

Jieba中文分词系列性能评测

测试

Unittest

go test ./...

Benchmark

go test -bench "Jieba" -test.benchtime 10s
go test -bench "Extractor" -test.benchtime 10s

客服

  • Email: i@yanyiwu.com
  • QQ: 64162451
  • WeChat:
  • image

Documentation

Index

Examples

Constants

View Source 
const TOTAL_DICT_PATH_NUMBER = 5

Variables

View Source 
var (
	DICT_DIR        string
	DICT_PATH       string
	HMM_PATH        string
	USER_DICT_PATH  string
	IDF_PATH        string
	STOP_WORDS_PATH string
)

Functions

This section is empty.

Types

type Jieba 

type Jieba struct {
	// contains filtered or unexported fields
}
Example 
var s string
var words []string
use_hmm := true
//equals with x := NewJieba(DICT_PATH, HMM_PATH, USER_DICT_PATH)
x := NewJieba()
defer x.Free()

s = "我来到北京清华大学"
words = x.CutAll(s)
fmt.Println(s)
fmt.Println("全模式:", strings.Join(words, "/"))

words = x.Cut(s, use_hmm)
fmt.Println(s)
fmt.Println("精确模式:", strings.Join(words, "/"))

s = "比特币"
words = x.Cut(s, use_hmm)
fmt.Println(s)
fmt.Println("精确模式:", strings.Join(words, "/"))

x.AddWord("比特币")
s = "比特币"
words = x.Cut(s, use_hmm)
fmt.Println(s)
fmt.Println("添加词典后,精确模式:", strings.Join(words, "/"))

s = "他来到了网易杭研大厦"
words = x.Cut(s, use_hmm)
fmt.Println(s)
fmt.Println("新词识别:", strings.Join(words, "/"))

s = "小明硕士毕业于中国科学院计算所，后在日本京都大学深造"
words = x.CutForSearch(s, use_hmm)
fmt.Println(s)
fmt.Println("搜索引擎模式:", strings.Join(words, "/"))

s = "长春市长春药店"
words = x.Tag(s)
fmt.Println(s)
fmt.Println("词性标注:", strings.Join(words, ","))

s = "区块链"
words = x.Tag(s)
fmt.Println(s)
fmt.Println("词性标注:", strings.Join(words, ","))

s = "长江大桥"
words = x.CutForSearch(s, !use_hmm)
fmt.Println(s)
fmt.Println("搜索引擎模式:", strings.Join(words, "/"))

wordinfos := x.Tokenize(s, SearchMode, !use_hmm)
fmt.Println(s)
fmt.Println("Tokenize:", wordinfos)

Output:
我来到北京清华大学
全模式: 我/来到/北京/清华/清华大学/华大/大学
我来到北京清华大学
精确模式: 我/来到/北京/清华大学
比特币
精确模式: 比特/币
比特币
添加词典后,精确模式: 比特币
他来到了网易杭研大厦
新词识别: 他/来到/了/网易/杭研/大厦
小明硕士毕业于中国科学院计算所，后在日本京都大学深造
搜索引擎模式: 小明/硕士/毕业/于/中国/科学/学院/科学院/中国科学院/计算/计算所/，/后/在/日本/京都/大学/日本京都大学/深造
长春市长春药店
词性标注: 长春市/ns,长春/ns,药店/n
区块链
词性标注: 区块链/nz
长江大桥
搜索引擎模式: 长江/大桥/长江大桥
长江大桥
Tokenize: [{长江 0 6} {大桥 6 12} {长江大桥 0 12}]

func NewJieba 

func NewJieba(paths ...string) *Jieba

func (*Jieba) AddWord 

func (x *Jieba) AddWord(s string)

func (*Jieba) Cut 

func (x *Jieba) Cut(s string, hmm bool) []string

func (*Jieba) CutAll 

func (x *Jieba) CutAll(s string) []string

func (*Jieba) CutForSearch 

func (x *Jieba) CutForSearch(s string, hmm bool) []string

func (*Jieba) Extract 

func (x *Jieba) Extract(s string, topk int) []string

func (*Jieba) ExtractWithWeight 

func (x *Jieba) ExtractWithWeight(s string, topk int) []WordWeight

func (*Jieba) Free 

func (x *Jieba) Free()

func (*Jieba) Tag 

func (x *Jieba) Tag(s string) []string

func (*Jieba) Tokenize 

func (x *Jieba) Tokenize(s string, mode TokenizeMode, hmm bool) []Word

type TokenizeMode 

type TokenizeMode int
const (
	DefaultMode TokenizeMode = iota
	SearchMode
)

type Word 

type Word struct {
	Str   string
	Start int
	End   int
}

type WordWeight 

type WordWeight struct {
	Word   string
	Weight float64
}

Source Files

View all Source files

Directories

Path Synopsis

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.