Golang 개발자를 위한 필수 기술: Baidu AI 인터페이스에 쉽게 연결하여 음성 인식 달성

go get -u github.com/go-resty/resty/v2
go get -u github.com/json-iterator/go

package main

import (
    "fmt"
    "io/ioutil"
    "net/http"
    "os"
    "strings"

    "github.com/go-resty/resty/v2"
    "github.com/json-iterator/go"
)

const (
    TokenURL     = "https://aip.baidubce.com/oauth/2.0/token"
    APIURL       = "http://vop.baidu.com/server_api"
    APIKey       = "your_api_key"    // 替换成你的API Key
    SecretKey    = "your_secret_key" // 替换成你的Secret Key
    AudioFile    = "audio.wav"       // 替换成你的音频文件路径
    DevUserID    = "user01"          // 替换成你的用户标识
)

type TokenResponse struct {
    AccessToken string `json:"access_token"`
    ExpiresIn   int    `json:"expires_in"`
}

type RecognitionResult struct {
    ErrNo  int      `json:"err_no"`
    ErrMsg string   `json:"err_msg"`
    Result []string `json:"result"`
}

func main() {
    accessToken := getAccessToken()

    audioData, err := ioutil.ReadFile(AudioFile)
    if err != nil {
        fmt.Printf("读取音频文件失败：%s
", err.Error())
        os.Exit(1)
    }

    boundary := "12345678901234567890"
    body := fmt.Sprintf("--%s
Content-Disposition: form-data; name="dev_pid"

1537
--%s
Content-Disposition: form-data; name="format"

wav
--%s
Content-Disposition: form-data; name="channel"

1
--%s
Content-Disposition: form-data; name="token"

%s
--%s
Content-Disposition: form-data; name="cuid"

%s
--%s
Content-Disposition: form-data; name="len"

%d
--%s
Content-Disposition: form-data; name="speech"; filename="%s"
Content-Type: application/octet-stream

%s
--%s--",
        boundary, boundary, boundary, boundary, accessToken, boundary, DevUserID, boundary, len(audioData), AudioFile, audioData, boundary)
    resp, err := resty.New().R().
        SetHeader("Content-Type", "multipart/form-data; boundary="+boundary).
        SetBody(body).
        Post(APIURL)
    if err != nil {
        fmt.Printf("请求百度AI接口失败：%s
", err.Error())
        os.Exit(1)
    }

    result := RecognitionResult{}
    if err := jsoniter.Unmarshal(resp.Body(), &result); err != nil {
        fmt.Printf("解析返回结果失败：%s
", err.Error())
        os.Exit(1)
    }

    if result.ErrNo != 0 {
        fmt.Printf("识别失败：%s
", result.ErrMsg)
    } else {
        text := strings.Join(result.Result, "")
        fmt.Printf("识别结果：%s
", text)
    }
}

func getAccessToken() string {
    resp, err := resty.New().R().
        SetQueryParams(map[string]string{
            "grant_type":    "client_credentials",
            "client_id":     APIKey,
            "client_secret": SecretKey,
        }).
        Get(TokenURL)
    if err != nil {
        fmt.Printf("获取百度AI接口Token失败：%s
", err.Error())
        os.Exit(1)
    }

    token := TokenResponse{}
    if err := jsoniter.Unmarshal(resp.Body(), &token); err != nil {
        fmt.Printf("解析Token失败：%s
", err.Error())
        os.Exit(1)
    }

    return token.AccessToken
}

go build speech_recognition.go
./speech_recognition