V3SentenceAsrExample.java

package com.tencent.trtcasr.examples;

import java.nio.file.Files;
import java.nio.file.Path;

import com.tencent.trtcasr.common.Credential;
import com.tencent.trtcasr.v3.SentenceRecognizer;
import com.tencent.trtcasr.v3.TranscribeRequest;
import com.tencent.trtcasr.v3.TranscribeResponse;
import com.tencent.trtcasr.v3.Wire;

/**
 * Sentence (one-shot) speech recognition example over the v3 protocol
 * (POST /v3/transcribe).
 *
 * <p>Recognizes a local audio file (<=60s, <=3MB).
 *
 * <p>Credentials come from environment variables:
 * {@code TRTC_ASR_SDK_APP_ID}, {@code TRTC_ASR_SECRET_KEY}
 * (v3 does not need the Tencent Cloud APPID.)
 *
 * <p>Usage: {@code java ... V3SentenceAsrExample <audio.pcm> [engine]}
 */
public class V3SentenceAsrExample {
    public static void main(String[] args) throws Exception {
        long sdkAppId = Long.parseLong(env("TRTC_ASR_SDK_APP_ID", "0"));
        String secretKey = env("TRTC_ASR_SECRET_KEY", "");
        if (sdkAppId == 0 || secretKey.isEmpty()) {
            System.err.println("Set TRTC_ASR_SDK_APP_ID and TRTC_ASR_SECRET_KEY first.");
            System.exit(1);
        }
        String path = args.length > 0 ? args[0] : "audio.pcm";
        if (args.length <= 1) {
            System.err.println("error: engine is required (engine model type, e.g. bigmodel)");
            System.exit(2);
        }
        String engine = args[1];
        // Empty lang falls back to server-side language detection.
        String lang = args.length > 2 ? args[2] : "";
        // The bigmodel engine is best used with an explicit language; every
        // other engine falls back to server-side detection unless lang is given.
        if (lang.isEmpty() && "bigmodel".equals(engine)) {
            lang = "zh";
        }

        // v3 credentials need only SDKAppID + SecretKey (no Tencent Cloud APPID).
        Credential credential = Wire.newCredential(sdkAppId, secretKey);
        SentenceRecognizer recognizer = new SentenceRecognizer(credential);

        byte[] data = Files.readAllBytes(Path.of(path));
        TranscribeRequest req = new TranscribeRequest();
        req.engineModelType = engine;
        req.voiceFormat = "pcm";
        if (!lang.isEmpty()) {
            req.language = lang;
        }
        TranscribeResponse resp = recognizer.recognizeDataWithOptions(data, req);

        System.out.println("Result: " + resp.result);
        System.out.println("Duration: " + resp.audioDuration + " ms  RequestId: " + resp.requestId);
        if (resp.wordList != null) {
            for (TranscribeResponse.Word w : resp.wordList) {
                System.out.printf("  [%6d - %6d] %s%n", w.startTime, w.endTime, w.word);
            }
        }
    }

    private static String env(String name, String fallback) {
        String v = System.getenv(name);
        return v == null || v.isEmpty() ? fallback : v;
    }
}