func NewSearchServer(corpusFile, sgmtDictFile string) *SearchServer {
sgmt := new(sego.Segmenter)
sgmt.LoadDictionary(sgmtDictFile)
idx := weakand.NewIndexFromFile(corpusFile, sgmt, "") // no index dump
return &SearchServer{index: idx, sgmt: sgmt}
}
func guaranteeSegmenter(sgmt **sego.Segmenter) {
if *sgmt == nil {
s := new(sego.Segmenter)
s.LoadDictionary(path.Join(gosrc(),
"github.com/huichen/sego/data/dictionary.txt"))
*sgmt = s
}
}
func main() {
flag.Parse()
var seg sego.Segmenter
seg.LoadDictionary("../data/dictionary.txt")
segments := seg.Segment([]byte(*text))
fmt.Println(sego.SegmentsToString(segments, true))
}
func tokenize(doc string, sgmt *sego.Segmenter) []string {
var terms []string
for _, seg := range sgmt.Segment([]byte(doc)) {
term := seg.Token().Text()
if !AllPunctOrSpace(term) {
terms = append(terms, term)
}
}
return terms
}
func Setup(t *telegram.Telegram, config map[string]interface{}, modules map[string]bool, cmds *types.CommandMap) types.Command {
if val, ok := modules["chinese"]; !ok || val {
var s sego.Segmenter
s.LoadDictionary(config["dict"].(string))
c := &Chinese{
tg: t,
redis: redis.NewClient(&redis.Options{
Addr: config["redis"].(string),
Password: "",
DB: int64(config["redis_db"].(float64)),
}),
seg: s,
}
if config["debug"] != nil {
c.debug = config["debug"].(bool)
}
(*cmds)["learn"] = types.Command{
Name: "learn",
Args: "<expr>",
ArgNum: -1,
Desc: "Learn a Chinese expression",
Processor: c,
}
(*cmds)["speak"] = types.Command{
Name: "speak",
ArgNum: 0,
Desc: "Speak a Chinese sentence based on previously learned data",
Processor: c,
}
(*cmds)["answer"] = types.Command{
Name: "answer",
Args: "[question]",
ArgNum: -1,
Desc: "Answer to [question]. If no [question] provided, answer to the message you reply to.",
Processor: c,
}
pong, err := c.redis.Ping().Result()
if (err != nil) || (pong != "PONG") {
panic(errors.New("Cannot PING redis"))
}
return types.Command{
Name: "chn",
Processor: c,
}
}
return types.Command{}
}
func main() {
// 载入词典
var segmenter sego.Segmenter
segmenter.LoadDictionary("/home/bryce/code/gocode/src/github.com/huichen/sego/data/dictionary.txt")
// 分词
text := []byte("中华人民共和国中央人民政府")
segments := segmenter.Segment(text)
// 处理分词结果
// 支持普通模式和搜索模式两种分词，见代码中SegmentsToString函数的注释。
fmt.Println(sego.SegmentsToString(segments, true))
}
func main() {
var seg sego.Segmenter
seg.LoadDictionary("data/dictionary.txt")
text1 := "我喜欢看电视，也喜欢接电话"
text2 := "我不喜欢看电视，也不喜欢看电影"
type num struct {
V, W int
}
var words = make(map[string]*num, 10)
segments1 := seg.Segment([]byte(text1))
segments2 := seg.Segment([]byte(text2))
for _, v := range sego.SegmentsToSlice(segments1, true) {
_, ok := words[v]
if ok {
p := words[v]
p.V = p.V + 1
words[v] = p
} else {
words[v] = &num{1, 0}
}
}
for _, v := range sego.SegmentsToSlice(segments2, true) {
_, ok := words[v]
if ok {
p := words[v]
p.W = p.W + 1
words[v] = p
} else {
words[v] = &num{0, 1}
}
}
//余弦相似度计算
var a, b, c int
for k, v := range words {
a = a + v.V*v.W
b = b + v.V*v.V
c = c + v.W*v.W
fmt.Println(k, v)
}
fmt.Println(a, b, c, float64(a)/(math.Sqrt(float64(b))*math.Sqrt(float64(c))))
}
func main() {
// 确保单线程，因为Go从1.5开始默认多线程
runtime.GOMAXPROCS(1)
// 解析命令行参数
flag.Parse()
// 记录时间
t0 := time.Now()
var segmenter sego.Segmenter
segmenter.LoadDictionary("../data/dictionary.txt")
// 记录时间
t1 := time.Now()
log.Printf("载入词典花费时间 %v", t1.Sub(t0))
// 写入内存profile文件
if *memprofile != "" {
f, err := os.Create(*memprofile)
if err != nil {
log.Fatal(err)
}
pprof.WriteHeapProfile(f)
defer f.Close()
}
// 打开将要分词的文件
file, err := os.Open("../testdata/bailuyuan.txt")
if err != nil {
log.Fatal(err)
}
defer file.Close()
// 逐行读入
scanner := bufio.NewScanner(file)
size := 0
lines := [][]byte{}
for scanner.Scan() {
var text string
fmt.Sscanf(scanner.Text(), "%s", &text)
content := []byte(text)
size += len(content)
lines = append(lines, content)
}
// 当指定输出文件时打开输出文件
var of *os.File
if *output != "" {
of, err = os.Create(*output)
if err != nil {
log.Fatal(err)
}
defer of.Close()
}
// 记录时间
t2 := time.Now()
// 打开处理器profile文件
if *cpuprofile != "" {
f, err := os.Create(*cpuprofile)
if err != nil {
log.Fatal(err)
}
pprof.StartCPUProfile(f)
defer pprof.StopCPUProfile()
}
// 分词
for i := 0; i < numRuns; i++ {
for _, l := range lines {
segments := segmenter.Segment(l)
if *output != "" {
of.WriteString(sego.SegmentsToString(segments, false))
of.WriteString("\n")
}
}
}
// 停止处理器profile
if *cpuprofile != "" {
defer pprof.StopCPUProfile()
}
// 记录时间并计算分词速度
t3 := time.Now()
log.Printf("分词花费时间 %v", t3.Sub(t2))
log.Printf("分词速度 %f MB/s", float64(size*numRuns)/t3.Sub(t2).Seconds()/(1024*1024))
}
