ruiyun-human 1.0.11 → 1.0.13

package/dist/ruiyun-human.cjs.js

@@ -385,8 +385,11 @@ var recorder = { exports: {} };
 var recorderExports = recorder.exports;
 var jsAudioRecorder = recorderExports;
 const Recorder = /* @__PURE__ */ getDefaultExportFromCjs(jsAudioRecorder);
-const baseTTSUrl = "http://192.168.3.182:8002";
-const ASR_URL = "http://192.168.3.182:9988";
+const VOICE_CLIENT_ID = "e5cd2b";
+const VOICE_TENANT_ID = "000000";
+const TTS_USER = "web-test-user";
+const TTS_PCM_SAMPLE_RATE = 16e3;
+const TTS_MAX_RETRIES = 5;
 const _sfc_main = {
   __name: "index",
   props: {
@@ -397,6 +400,8 @@ const _sfc_main = {
   },
   emits: ["loaded", "changed", "paused", "resumed", "audio-ready"],
   setup(__props, { expose: __expose, emit: __emit }) {
+    const TTS_URL = "http://localhost:6039";
+    const ASR_URL = "http://localhost:6039";
     const props = __props;
     const emit = __emit;
     const containerRef = vue.ref(null);
@@ -649,112 +654,93 @@ const _sfc_main = {
         currentAudio.value = null;
       }
     }
-    const voiceFileCache = /* @__PURE__ */ new Map();
-    const FILEDATA_META = { _type: "gradio.FileData" };
-    const uploadReferenceAudio = async (blob, filename, signal) => {
-      const form = new FormData();
-      form.append("files", blob, filename);
-      const res = await fetch(`${baseTTSUrl}/gradio_api/upload`, {
-        method: "POST",
-        body: form,
-        signal
-      });
-      if (!res.ok) throw new Error(`参考音频上传失败 HTTP ${res.status}`);
-      const paths = await res.json();
-      const path = Array.isArray(paths) ? paths[0] : paths;
-      if (!path) throw new Error("参考音频上传失败：服务未返回文件路径");
-      return { path, orig_name: filename, meta: FILEDATA_META };
+    const base64ToBlob = (b64, mime = "audio/wav") => {
+      const raw = String(b64).replace(/^data:audio\/[\w+-]+;base64,/, "");
+      const binary = atob(raw);
+      const bytes = new Uint8Array(binary.length);
+      for (let i = 0; i < binary.length; i++) bytes[i] = binary.charCodeAt(i);
+      return new Blob([bytes], { type: mime });
     };
-    const resolveVoiceFileData = async (voiceSource, signal) => {
-      if (voiceSource && typeof voiceSource === "object" && !Array.isArray(voiceSource) && voiceSource.path) {
-        return {
-          path: voiceSource.path,
-          orig_name: voiceSource.orig_name || "reference.mp3",
-          meta: FILEDATA_META
-        };
-      }
-      if (voiceSource instanceof Blob) {
-        const name = voiceSource.name || `reference_${Date.now()}.mp3`;
-        return await uploadReferenceAudio(voiceSource, name, signal);
-      }
-      if (typeof voiceSource === "string" && voiceSource) {
-        if (voiceSource.startsWith(baseTTSUrl)) {
-          const match = voiceSource.match(/\/gradio_api\/file=(.+)$/);
-          if (match) {
-            const path = decodeURIComponent(match[1]);
-            const origName = path.split(/[\\/]/).pop() || "reference.mp3";
-            return { path, orig_name: origName, meta: FILEDATA_META };
-          }
-        }
-        const res = await fetch(voiceSource, { signal });
-        if (!res.ok) throw new Error(`参考音频下载失败 HTTP ${res.status}`);
-        const blob = await res.blob();
-        const fileName = decodeURIComponent(voiceSource.split("?")[0].split("/").pop()) || "reference.mp3";
-        return await uploadReferenceAudio(blob, fileName, signal);
-      }
-      throw new Error("不支持的参考音频来源：voiceUrl 应为音频 URL 字符串、File/Blob 或 { path } 对象");
+    const pcmToWavBlob = (pcmBuffer, sampleRate = TTS_PCM_SAMPLE_RATE) => {
+      const dataLength = pcmBuffer.byteLength;
+      const header = new ArrayBuffer(44);
+      const view = new DataView(header);
+      const writeStr = (offset, str) => {
+        for (let i = 0; i < str.length; i++) view.setUint8(offset + i, str.charCodeAt(i));
+      };
+      writeStr(0, "RIFF");
+      view.setUint32(4, 36 + dataLength, true);
+      writeStr(8, "WAVE");
+      writeStr(12, "fmt ");
+      view.setUint32(16, 16, true);
+      view.setUint16(20, 1, true);
+      view.setUint16(22, 1, true);
+      view.setUint32(24, sampleRate, true);
+      view.setUint32(28, sampleRate * 2, true);
+      view.setUint16(32, 2, true);
+      view.setUint16(34, 16, true);
+      writeStr(36, "data");
+      view.setUint32(40, dataLength, true);
+      return new Blob([header, pcmBuffer], { type: "audio/wav" });
     };
-    const submitCloneJob = async (data, signal) => {
-      const res = await fetch(`${baseTTSUrl}/gradio_api/call/_clone_fn`, {
-        method: "POST",
-        headers: { "Content-Type": "application/json" },
-        body: JSON.stringify({ data }),
-        signal
-      });
-      if (!res.ok) {
-        const text = await res.text().catch(() => "");
-        throw new Error(`克隆任务提交失败 HTTP ${res.status} ${text}`);
+    const ensurePlayableWavBlob = async (blob) => {
+      const buf = await blob.arrayBuffer();
+      if (buf.byteLength < 4) return blob;
+      const head = String.fromCharCode(...new Uint8Array(buf, 0, 4));
+      if (head === "RIFF") {
+        return new Blob([buf], { type: "audio/wav" });
       }
-      const json = await res.json();
-      if (!(json == null ? void 0 : json.event_id)) throw new Error("克隆任务提交失败：未返回 event_id");
-      return json.event_id;
+      console.log("[TTS] 检测到裸 PCM，补 WAV 头后播放", { size: buf.byteLength });
+      return pcmToWavBlob(buf);
     };
-    const waitCloneResult = async (eventId, signal) => {
-      const res = await fetch(`${baseTTSUrl}/gradio_api/call/_clone_fn/${eventId}`, {
-        signal,
-        headers: { Accept: "text/event-stream" }
-      });
-      if (!res.ok) throw new Error(`合成结果监听失败 HTTP ${res.status}`);
-      const reader = res.body.getReader();
-      const decoder = new TextDecoder("utf-8");
-      let buffer = "";
-      let eventName = "message";
-      while (true) {
-        const { done, value } = await reader.read();
-        if (done) break;
-        buffer += decoder.decode(value, { stream: true }).replace(/\r\n/g, "\n");
-        let sep;
-        while ((sep = buffer.indexOf("\n\n")) !== -1) {
-          const frame = buffer.slice(0, sep);
-          buffer = buffer.slice(sep + 2);
-          const dataLines = [];
-          for (const line of frame.split("\n")) {
-            if (line.startsWith("event:")) eventName = line.slice(6).trim();
-            else if (line.startsWith("data:")) dataLines.push(line.slice(5).replace(/^ /, ""));
-          }
-          if (dataLines.length === 0) continue;
-          const payload = dataLines.join("\n");
-          if (eventName === "complete") {
-            return JSON.parse(payload);
-          }
-          if (eventName === "error") {
-            let msg = payload;
-            try {
-              const parsed = JSON.parse(payload);
-              msg = typeof parsed === "string" ? parsed : (parsed == null ? void 0 : parsed.message) || payload;
-            } catch {
-            }
-            throw new Error(msg || "OmniVoice 合成失败");
+    const requestTTSAudio = async (text, signal) => {
+      let lastError = null;
+      for (let attempt = 1; attempt <= TTS_MAX_RETRIES; attempt++) {
+        if (signal == null ? void 0 : signal.aborted) throw new DOMException("Aborted", "AbortError");
+        const res = await fetch(`${TTS_URL}/web/voice/jd/text-to-speech`, {
+          method: "POST",
+          headers: {
+            "Content-Type": "application/json",
+            clientid: VOICE_CLIENT_ID,
+            "tenant-id": VOICE_TENANT_ID
+          },
+          body: JSON.stringify({ text, user: TTS_USER, timbre: 47 }),
+          signal
+        });
+        if (!res.ok) {
+          const errText = await res.text().catch(() => "");
+          lastError = new Error(`TTS HTTP ${res.status} ${errText}`);
+          if (res.status >= 500 && attempt < TTS_MAX_RETRIES) {
+            console.warn(`[TTS] ${res.status}，重试 ${attempt}/${TTS_MAX_RETRIES}`);
+            await new Promise((r) => setTimeout(r, 400 * attempt));
+            continue;
           }
-          eventName = "message";
+          throw lastError;
         }
+        const contentType = (res.headers.get("content-type") || "").toLowerCase();
+        if (contentType.includes("audio") || contentType.includes("octet-stream")) {
+          const raw = await res.blob();
+          console.log("[TTS] 收到音频", { size: raw.size, contentType });
+          return ensurePlayableWavBlob(raw);
+        }
+        const json = await res.json();
+        const data = json == null ? void 0 : json.data;
+        const audioUrl = typeof data === "object" && ((data == null ? void 0 : data.url) || (data == null ? void 0 : data.audioUrl)) || (json == null ? void 0 : json.url) || (json == null ? void 0 : json.audioUrl);
+        const b64 = (typeof data === "string" ? data : null) || (data == null ? void 0 : data.audio) || (data == null ? void 0 : data.base64) || (json == null ? void 0 : json.audio) || (json == null ? void 0 : json.base64);
+        if (audioUrl && typeof audioUrl === "string" && /^https?:\/\//.test(audioUrl)) {
+          const audioRes = await fetch(audioUrl, { signal });
+          if (!audioRes.ok) throw new Error(`TTS 音频下载失败 HTTP ${audioRes.status}`);
+          return ensurePlayableWavBlob(await audioRes.blob());
+        }
+        if (b64 && typeof b64 === "string") {
+          const mime = (data == null ? void 0 : data.format) ? `audio/${data.format}` : "audio/wav";
+          return ensurePlayableWavBlob(base64ToBlob(b64, mime));
+        }
+        throw new Error((json == null ? void 0 : json.msg) || (json == null ? void 0 : json.message) || "TTS 未返回可播放音频");
       }
-      throw new Error("OmniVoice 合成结果流提前中断");
+      throw lastError || new Error("TTS 请求失败");
     };
-    const playAudioFromUrl = async (audioUrl, requestId, signal) => {
-      const audioRes = await fetch(audioUrl, { signal });
-      if (!audioRes.ok) throw new Error(`合成音频下载失败 HTTP ${audioRes.status}`);
-      const audioBlob = await audioRes.blob();
+    const playAudioBlob = async (audioBlob, requestId, signal) => {
       const objectUrl = URL.createObjectURL(audioBlob);
       if (signal.aborted || currentTtsRequestId.value !== requestId) {
         URL.revokeObjectURL(objectUrl);
@@ -767,32 +753,29 @@ const _sfc_main = {
         if (currentAudio.value === audio) currentAudio.value = null;
       };
       audio.onplaying = () => {
-        console.log("[OmniVoice] 音频开始播放");
+        console.log("[TTS] 音频开始播放");
         speak();
       };
       audio.onended = () => {
-        console.log("[OmniVoice] 音频播放结束");
+        console.log("[TTS] 音频播放结束");
         cleanup();
         stopSpeaking();
       };
       audio.onerror = () => {
-        console.error("[OmniVoice] 音频播放失败");
+        console.error("[TTS] 音频播放失败", audio.error);
         cleanup();
         stopSpeaking();
       };
-      await audio.play();
+      try {
+        await audio.play();
+      } catch (e) {
+        console.error("[TTS] audio.play 被拒绝或失败:", e);
+        cleanup();
+        throw e;
+      }
     };
     const playTTS = async (text) => {
-      var _a, _b, _c, _d;
-      console.log(props.humans);
-      let voiceUrl = "";
-      let voiceUrl2 = "http://192.168.3.182:8002/gradio_api/file=C:\\Users\\Administrator\\AppData\\Local\\Temp\\gradio\\4eef6d9157de756de02bccca2f66389382e2ac4acb0c674981aa9fe13dcc294c\\李梓萌.mp3";
-      let voiceUrl1 = "http://192.168.3.182:8002/gradio_api/file=C:\\Users\\Administrator\\AppData\\Local\\Temp\\gradio\\ff2665f538a4b0e5efbcd79f9f699236f5f0e52cad7e17f0664ddf431169be30\\guozj.mp3";
-      if (props.humans[0].speakSet.sex == "Female / 女") {
-        voiceUrl = voiceUrl2;
-      } else {
-        voiceUrl = voiceUrl1;
-      }
+      var _a;
       if (!text || !((_a = currentHuman.value) == null ? void 0 : _a.isMuted)) return;
       abortTTS();
       ttsAbortController = new AbortController();
@@ -802,39 +785,19 @@ const _sfc_main = {
       try {
         const cleanText = text.replace(/```[\s\S]*?```/g, "").replace(/`[^`]+`/g, "").replace(/[#*_~]/g, "").replace(/\n+/g, "，").replace(/\s+/g, " ").replace(/[，。！？]{2,}/g, "，").trim();
         if (!cleanText) return;
-        const speed = Number((_c = (_b = currentHuman.value) == null ? void 0 : _b.speakSet) == null ? void 0 : _c.speed) || 1;
-        const voiceSource = voiceUrl || ((_d = currentHuman.value) == null ? void 0 : _d.voiceUrl);
-        if (!voiceSource) {
-          console.error("[OmniVoice] 未配置参考音频 voiceUrl，无法进行音色克隆");
-          standby();
-          return;
-        }
-        let refFile = voiceFileCache.get(voiceSource);
-        if (!refFile) {
-          console.log("[OmniVoice] 准备参考音频:", typeof voiceSource === "string" ? voiceSource : "Blob/File");
-          refFile = await resolveVoiceFileData(voiceSource, signal);
-          voiceFileCache.set(voiceSource, refFile);
-        }
-        console.log("[OmniVoice] 参考音频就绪:", refFile);
-        const data = [cleanText, "Auto", refFile, "", null, 32, 2, true, speed, null, true, true];
-        console.log("[OmniVoice] 提交克隆请求:", { text: cleanText, length: cleanText.length });
-        const eventId = await submitCloneJob(data, signal);
-        const outputs = await waitCloneResult(eventId, signal);
-        const audioFile = outputs == null ? void 0 : outputs[0];
-        const statusText = outputs == null ? void 0 : outputs[1];
-        console.log("[OmniVoice] 合成结果:", { audioFile, statusText });
+        console.log("[TTS] 提交合成请求:", { text: cleanText, length: cleanText.length });
+        const audioBlob = await requestTTSAudio(cleanText, signal);
         if (signal.aborted || currentTtsRequestId.value !== requestId) return;
-        if (!audioFile || !audioFile.path && !audioFile.url) {
-          throw new Error(statusText || "OmniVoice 未返回合成音频");
+        if (!audioBlob || audioBlob.size === 0) {
+          throw new Error("TTS 返回空音频");
         }
-        const audioUrl = audioFile.url && audioFile.url.startsWith("http") ? audioFile.url : `${baseTTSUrl}/gradio_api/file=${audioFile.path}`;
-        await playAudioFromUrl(audioUrl, requestId, signal);
+        await playAudioBlob(audioBlob, requestId, signal);
       } catch (e) {
         if ((e == null ? void 0 : e.name) === "AbortError") {
-          console.log("[OmniVoice] TTS已取消");
+          console.log("[TTS] 已取消");
           return;
         }
-        console.error("[OmniVoice] TTS失败:", e);
+        console.error("[TTS] 失败:", e);
         standby();
       }
     };
@@ -864,15 +827,28 @@ const _sfc_main = {
     const audioToText = async (blob) => {
       const formData = new FormData();
       formData.append("file", blob, "recorded_audio.wav");
+      const params = new URLSearchParams({
+        domain: "general",
+        sampleRate: "16000"
+      });
+      const url = `${ASR_URL}/web/voice/jd/speech-to-text?${params}`;
       try {
-        const res = await fetch(ASR_URL + "/transcribe", {
+        const res = await fetch(url, {
           method: "POST",
+          headers: {
+            clientid: VOICE_CLIENT_ID,
+            "tenant-id": VOICE_TENANT_ID
+          },
           body: formData
         });
+        if (!res.ok) {
+          throw new Error(`STT HTTP ${res.status}`);
+        }
         const resJson = await res.json();
-        const text = (resJson == null ? void 0 : resJson.text) || (resJson == null ? void 0 : resJson.text);
-        const message = (resJson == null ? void 0 : resJson.success) || (resJson == null ? void 0 : resJson.success);
-        return { status: message, text };
+        const data = resJson == null ? void 0 : resJson.data;
+        const text = (resJson == null ? void 0 : resJson.text) ?? (typeof data === "string" ? data : data == null ? void 0 : data.text) ?? (resJson == null ? void 0 : resJson.result) ?? "";
+        const status = (resJson == null ? void 0 : resJson.success) ?? (resJson == null ? void 0 : resJson.code) ?? res.ok;
+        return { status, text };
       } catch (e) {
         console.error("STT失败:", e);
         return { status: e, text: "" };
@@ -929,7 +905,7 @@ const _sfc_main = {
     };
   }
 };
-const DigitalHuman = /* @__PURE__ */ _export_sfc(_sfc_main, [["__scopeId", "data-v-9c25016c"]]);
+const DigitalHuman = /* @__PURE__ */ _export_sfc(_sfc_main, [["__scopeId", "data-v-9379171c"]]);
 DigitalHuman.install = (app) => app.component("DigitalHuman", DigitalHuman);
 const components = [DemoButton, DemoInput, DigitalHuman];
 const install = (app) => {

package/dist/ruiyun-human.es.js

@@ -383,8 +383,11 @@ var recorder = { exports: {} };
 var recorderExports = recorder.exports;
 var jsAudioRecorder = recorderExports;
 const Recorder = /* @__PURE__ */ getDefaultExportFromCjs(jsAudioRecorder);
-const baseTTSUrl = "http://192.168.3.182:8002";
-const ASR_URL = "http://192.168.3.182:9988";
+const VOICE_CLIENT_ID = "e5cd2b";
+const VOICE_TENANT_ID = "000000";
+const TTS_USER = "web-test-user";
+const TTS_PCM_SAMPLE_RATE = 16e3;
+const TTS_MAX_RETRIES = 5;
 const _sfc_main = {
   __name: "index",
   props: {
@@ -395,6 +398,8 @@ const _sfc_main = {
   },
   emits: ["loaded", "changed", "paused", "resumed", "audio-ready"],
   setup(__props, { expose: __expose, emit: __emit }) {
+    const TTS_URL = "http://localhost:6039";
+    const ASR_URL = "http://localhost:6039";
     const props = __props;
     const emit = __emit;
     const containerRef = ref(null);
@@ -647,112 +652,93 @@ const _sfc_main = {
         currentAudio.value = null;
       }
     }
-    const voiceFileCache = /* @__PURE__ */ new Map();
-    const FILEDATA_META = { _type: "gradio.FileData" };
-    const uploadReferenceAudio = async (blob, filename, signal) => {
-      const form = new FormData();
-      form.append("files", blob, filename);
-      const res = await fetch(`${baseTTSUrl}/gradio_api/upload`, {
-        method: "POST",
-        body: form,
-        signal
-      });
-      if (!res.ok) throw new Error(`参考音频上传失败 HTTP ${res.status}`);
-      const paths = await res.json();
-      const path = Array.isArray(paths) ? paths[0] : paths;
-      if (!path) throw new Error("参考音频上传失败：服务未返回文件路径");
-      return { path, orig_name: filename, meta: FILEDATA_META };
+    const base64ToBlob = (b64, mime = "audio/wav") => {
+      const raw = String(b64).replace(/^data:audio\/[\w+-]+;base64,/, "");
+      const binary = atob(raw);
+      const bytes = new Uint8Array(binary.length);
+      for (let i = 0; i < binary.length; i++) bytes[i] = binary.charCodeAt(i);
+      return new Blob([bytes], { type: mime });
     };
-    const resolveVoiceFileData = async (voiceSource, signal) => {
-      if (voiceSource && typeof voiceSource === "object" && !Array.isArray(voiceSource) && voiceSource.path) {
-        return {
-          path: voiceSource.path,
-          orig_name: voiceSource.orig_name || "reference.mp3",
-          meta: FILEDATA_META
-        };
-      }
-      if (voiceSource instanceof Blob) {
-        const name = voiceSource.name || `reference_${Date.now()}.mp3`;
-        return await uploadReferenceAudio(voiceSource, name, signal);
-      }
-      if (typeof voiceSource === "string" && voiceSource) {
-        if (voiceSource.startsWith(baseTTSUrl)) {
-          const match = voiceSource.match(/\/gradio_api\/file=(.+)$/);
-          if (match) {
-            const path = decodeURIComponent(match[1]);
-            const origName = path.split(/[\\/]/).pop() || "reference.mp3";
-            return { path, orig_name: origName, meta: FILEDATA_META };
-          }
-        }
-        const res = await fetch(voiceSource, { signal });
-        if (!res.ok) throw new Error(`参考音频下载失败 HTTP ${res.status}`);
-        const blob = await res.blob();
-        const fileName = decodeURIComponent(voiceSource.split("?")[0].split("/").pop()) || "reference.mp3";
-        return await uploadReferenceAudio(blob, fileName, signal);
-      }
-      throw new Error("不支持的参考音频来源：voiceUrl 应为音频 URL 字符串、File/Blob 或 { path } 对象");
+    const pcmToWavBlob = (pcmBuffer, sampleRate = TTS_PCM_SAMPLE_RATE) => {
+      const dataLength = pcmBuffer.byteLength;
+      const header = new ArrayBuffer(44);
+      const view = new DataView(header);
+      const writeStr = (offset, str) => {
+        for (let i = 0; i < str.length; i++) view.setUint8(offset + i, str.charCodeAt(i));
+      };
+      writeStr(0, "RIFF");
+      view.setUint32(4, 36 + dataLength, true);
+      writeStr(8, "WAVE");
+      writeStr(12, "fmt ");
+      view.setUint32(16, 16, true);
+      view.setUint16(20, 1, true);
+      view.setUint16(22, 1, true);
+      view.setUint32(24, sampleRate, true);
+      view.setUint32(28, sampleRate * 2, true);
+      view.setUint16(32, 2, true);
+      view.setUint16(34, 16, true);
+      writeStr(36, "data");
+      view.setUint32(40, dataLength, true);
+      return new Blob([header, pcmBuffer], { type: "audio/wav" });
     };
-    const submitCloneJob = async (data, signal) => {
-      const res = await fetch(`${baseTTSUrl}/gradio_api/call/_clone_fn`, {
-        method: "POST",
-        headers: { "Content-Type": "application/json" },
-        body: JSON.stringify({ data }),
-        signal
-      });
-      if (!res.ok) {
-        const text = await res.text().catch(() => "");
-        throw new Error(`克隆任务提交失败 HTTP ${res.status} ${text}`);
+    const ensurePlayableWavBlob = async (blob) => {
+      const buf = await blob.arrayBuffer();
+      if (buf.byteLength < 4) return blob;
+      const head = String.fromCharCode(...new Uint8Array(buf, 0, 4));
+      if (head === "RIFF") {
+        return new Blob([buf], { type: "audio/wav" });
       }
-      const json = await res.json();
-      if (!(json == null ? void 0 : json.event_id)) throw new Error("克隆任务提交失败：未返回 event_id");
-      return json.event_id;
+      console.log("[TTS] 检测到裸 PCM，补 WAV 头后播放", { size: buf.byteLength });
+      return pcmToWavBlob(buf);
     };
-    const waitCloneResult = async (eventId, signal) => {
-      const res = await fetch(`${baseTTSUrl}/gradio_api/call/_clone_fn/${eventId}`, {
-        signal,
-        headers: { Accept: "text/event-stream" }
-      });
-      if (!res.ok) throw new Error(`合成结果监听失败 HTTP ${res.status}`);
-      const reader = res.body.getReader();
-      const decoder = new TextDecoder("utf-8");
-      let buffer = "";
-      let eventName = "message";
-      while (true) {
-        const { done, value } = await reader.read();
-        if (done) break;
-        buffer += decoder.decode(value, { stream: true }).replace(/\r\n/g, "\n");
-        let sep;
-        while ((sep = buffer.indexOf("\n\n")) !== -1) {
-          const frame = buffer.slice(0, sep);
-          buffer = buffer.slice(sep + 2);
-          const dataLines = [];
-          for (const line of frame.split("\n")) {
-            if (line.startsWith("event:")) eventName = line.slice(6).trim();
-            else if (line.startsWith("data:")) dataLines.push(line.slice(5).replace(/^ /, ""));
-          }
-          if (dataLines.length === 0) continue;
-          const payload = dataLines.join("\n");
-          if (eventName === "complete") {
-            return JSON.parse(payload);
-          }
-          if (eventName === "error") {
-            let msg = payload;
-            try {
-              const parsed = JSON.parse(payload);
-              msg = typeof parsed === "string" ? parsed : (parsed == null ? void 0 : parsed.message) || payload;
-            } catch {
-            }
-            throw new Error(msg || "OmniVoice 合成失败");
+    const requestTTSAudio = async (text, signal) => {
+      let lastError = null;
+      for (let attempt = 1; attempt <= TTS_MAX_RETRIES; attempt++) {
+        if (signal == null ? void 0 : signal.aborted) throw new DOMException("Aborted", "AbortError");
+        const res = await fetch(`${TTS_URL}/web/voice/jd/text-to-speech`, {
+          method: "POST",
+          headers: {
+            "Content-Type": "application/json",
+            clientid: VOICE_CLIENT_ID,
+            "tenant-id": VOICE_TENANT_ID
+          },
+          body: JSON.stringify({ text, user: TTS_USER, timbre: 47 }),
+          signal
+        });
+        if (!res.ok) {
+          const errText = await res.text().catch(() => "");
+          lastError = new Error(`TTS HTTP ${res.status} ${errText}`);
+          if (res.status >= 500 && attempt < TTS_MAX_RETRIES) {
+            console.warn(`[TTS] ${res.status}，重试 ${attempt}/${TTS_MAX_RETRIES}`);
+            await new Promise((r) => setTimeout(r, 400 * attempt));
+            continue;
           }
-          eventName = "message";
+          throw lastError;
         }
+        const contentType = (res.headers.get("content-type") || "").toLowerCase();
+        if (contentType.includes("audio") || contentType.includes("octet-stream")) {
+          const raw = await res.blob();
+          console.log("[TTS] 收到音频", { size: raw.size, contentType });
+          return ensurePlayableWavBlob(raw);
+        }
+        const json = await res.json();
+        const data = json == null ? void 0 : json.data;
+        const audioUrl = typeof data === "object" && ((data == null ? void 0 : data.url) || (data == null ? void 0 : data.audioUrl)) || (json == null ? void 0 : json.url) || (json == null ? void 0 : json.audioUrl);
+        const b64 = (typeof data === "string" ? data : null) || (data == null ? void 0 : data.audio) || (data == null ? void 0 : data.base64) || (json == null ? void 0 : json.audio) || (json == null ? void 0 : json.base64);
+        if (audioUrl && typeof audioUrl === "string" && /^https?:\/\//.test(audioUrl)) {
+          const audioRes = await fetch(audioUrl, { signal });
+          if (!audioRes.ok) throw new Error(`TTS 音频下载失败 HTTP ${audioRes.status}`);
+          return ensurePlayableWavBlob(await audioRes.blob());
+        }
+        if (b64 && typeof b64 === "string") {
+          const mime = (data == null ? void 0 : data.format) ? `audio/${data.format}` : "audio/wav";
+          return ensurePlayableWavBlob(base64ToBlob(b64, mime));
+        }
+        throw new Error((json == null ? void 0 : json.msg) || (json == null ? void 0 : json.message) || "TTS 未返回可播放音频");
       }
-      throw new Error("OmniVoice 合成结果流提前中断");
+      throw lastError || new Error("TTS 请求失败");
     };
-    const playAudioFromUrl = async (audioUrl, requestId, signal) => {
-      const audioRes = await fetch(audioUrl, { signal });
-      if (!audioRes.ok) throw new Error(`合成音频下载失败 HTTP ${audioRes.status}`);
-      const audioBlob = await audioRes.blob();
+    const playAudioBlob = async (audioBlob, requestId, signal) => {
       const objectUrl = URL.createObjectURL(audioBlob);
       if (signal.aborted || currentTtsRequestId.value !== requestId) {
         URL.revokeObjectURL(objectUrl);
@@ -765,32 +751,29 @@ const _sfc_main = {
         if (currentAudio.value === audio) currentAudio.value = null;
       };
       audio.onplaying = () => {
-        console.log("[OmniVoice] 音频开始播放");
+        console.log("[TTS] 音频开始播放");
         speak();
       };
       audio.onended = () => {
-        console.log("[OmniVoice] 音频播放结束");
+        console.log("[TTS] 音频播放结束");
         cleanup();
         stopSpeaking();
       };
       audio.onerror = () => {
-        console.error("[OmniVoice] 音频播放失败");
+        console.error("[TTS] 音频播放失败", audio.error);
         cleanup();
         stopSpeaking();
       };
-      await audio.play();
+      try {
+        await audio.play();
+      } catch (e) {
+        console.error("[TTS] audio.play 被拒绝或失败:", e);
+        cleanup();
+        throw e;
+      }
     };
     const playTTS = async (text) => {
-      var _a, _b, _c, _d;
-      console.log(props.humans);
-      let voiceUrl = "";
-      let voiceUrl2 = "http://192.168.3.182:8002/gradio_api/file=C:\\Users\\Administrator\\AppData\\Local\\Temp\\gradio\\4eef6d9157de756de02bccca2f66389382e2ac4acb0c674981aa9fe13dcc294c\\李梓萌.mp3";
-      let voiceUrl1 = "http://192.168.3.182:8002/gradio_api/file=C:\\Users\\Administrator\\AppData\\Local\\Temp\\gradio\\ff2665f538a4b0e5efbcd79f9f699236f5f0e52cad7e17f0664ddf431169be30\\guozj.mp3";
-      if (props.humans[0].speakSet.sex == "Female / 女") {
-        voiceUrl = voiceUrl2;
-      } else {
-        voiceUrl = voiceUrl1;
-      }
+      var _a;
       if (!text || !((_a = currentHuman.value) == null ? void 0 : _a.isMuted)) return;
       abortTTS();
       ttsAbortController = new AbortController();
@@ -800,39 +783,19 @@ const _sfc_main = {
       try {
         const cleanText = text.replace(/```[\s\S]*?```/g, "").replace(/`[^`]+`/g, "").replace(/[#*_~]/g, "").replace(/\n+/g, "，").replace(/\s+/g, " ").replace(/[，。！？]{2,}/g, "，").trim();
         if (!cleanText) return;
-        const speed = Number((_c = (_b = currentHuman.value) == null ? void 0 : _b.speakSet) == null ? void 0 : _c.speed) || 1;
-        const voiceSource = voiceUrl || ((_d = currentHuman.value) == null ? void 0 : _d.voiceUrl);
-        if (!voiceSource) {
-          console.error("[OmniVoice] 未配置参考音频 voiceUrl，无法进行音色克隆");
-          standby();
-          return;
-        }
-        let refFile = voiceFileCache.get(voiceSource);
-        if (!refFile) {
-          console.log("[OmniVoice] 准备参考音频:", typeof voiceSource === "string" ? voiceSource : "Blob/File");
-          refFile = await resolveVoiceFileData(voiceSource, signal);
-          voiceFileCache.set(voiceSource, refFile);
-        }
-        console.log("[OmniVoice] 参考音频就绪:", refFile);
-        const data = [cleanText, "Auto", refFile, "", null, 32, 2, true, speed, null, true, true];
-        console.log("[OmniVoice] 提交克隆请求:", { text: cleanText, length: cleanText.length });
-        const eventId = await submitCloneJob(data, signal);
-        const outputs = await waitCloneResult(eventId, signal);
-        const audioFile = outputs == null ? void 0 : outputs[0];
-        const statusText = outputs == null ? void 0 : outputs[1];
-        console.log("[OmniVoice] 合成结果:", { audioFile, statusText });
+        console.log("[TTS] 提交合成请求:", { text: cleanText, length: cleanText.length });
+        const audioBlob = await requestTTSAudio(cleanText, signal);
         if (signal.aborted || currentTtsRequestId.value !== requestId) return;
-        if (!audioFile || !audioFile.path && !audioFile.url) {
-          throw new Error(statusText || "OmniVoice 未返回合成音频");
+        if (!audioBlob || audioBlob.size === 0) {
+          throw new Error("TTS 返回空音频");
         }
-        const audioUrl = audioFile.url && audioFile.url.startsWith("http") ? audioFile.url : `${baseTTSUrl}/gradio_api/file=${audioFile.path}`;
-        await playAudioFromUrl(audioUrl, requestId, signal);
+        await playAudioBlob(audioBlob, requestId, signal);
       } catch (e) {
         if ((e == null ? void 0 : e.name) === "AbortError") {
-          console.log("[OmniVoice] TTS已取消");
+          console.log("[TTS] 已取消");
           return;
         }
-        console.error("[OmniVoice] TTS失败:", e);
+        console.error("[TTS] 失败:", e);
         standby();
       }
     };
@@ -862,15 +825,28 @@ const _sfc_main = {
     const audioToText = async (blob) => {
       const formData = new FormData();
       formData.append("file", blob, "recorded_audio.wav");
+      const params = new URLSearchParams({
+        domain: "general",
+        sampleRate: "16000"
+      });
+      const url = `${ASR_URL}/web/voice/jd/speech-to-text?${params}`;
       try {
-        const res = await fetch(ASR_URL + "/transcribe", {
+        const res = await fetch(url, {
           method: "POST",
+          headers: {
+            clientid: VOICE_CLIENT_ID,
+            "tenant-id": VOICE_TENANT_ID
+          },
           body: formData
         });
+        if (!res.ok) {
+          throw new Error(`STT HTTP ${res.status}`);
+        }
         const resJson = await res.json();
-        const text = (resJson == null ? void 0 : resJson.text) || (resJson == null ? void 0 : resJson.text);
-        const message = (resJson == null ? void 0 : resJson.success) || (resJson == null ? void 0 : resJson.success);
-        return { status: message, text };
+        const data = resJson == null ? void 0 : resJson.data;
+        const text = (resJson == null ? void 0 : resJson.text) ?? (typeof data === "string" ? data : data == null ? void 0 : data.text) ?? (resJson == null ? void 0 : resJson.result) ?? "";
+        const status = (resJson == null ? void 0 : resJson.success) ?? (resJson == null ? void 0 : resJson.code) ?? res.ok;
+        return { status, text };
       } catch (e) {
         console.error("STT失败:", e);
         return { status: e, text: "" };
@@ -927,7 +903,7 @@ const _sfc_main = {
     };
   }
 };
-const DigitalHuman = /* @__PURE__ */ _export_sfc(_sfc_main, [["__scopeId", "data-v-9c25016c"]]);
+const DigitalHuman = /* @__PURE__ */ _export_sfc(_sfc_main, [["__scopeId", "data-v-9379171c"]]);
 DigitalHuman.install = (app) => app.component("DigitalHuman", DigitalHuman);
 const components = [DemoButton, DemoInput, DigitalHuman];
 const install = (app) => {

package/dist/style.css

@@ -1 +1 @@
-.demo-btn[data-v-9a48516a]{padding:6px 14px;border-radius:4px;border:none;cursor:pointer;font-size:14px}.primary[data-v-9a48516a]{background:#409eff;color:#fff}.success[data-v-9a48516a]{background:#67c23a;color:#fff}.warning[data-v-9a48516a]{background:#e6a23c;color:#fff}.demo-input-wrap[data-v-28a3493c]{display:inline-flex;align-items:center;border:1px solid #dcdcdc;border-radius:4px;padding:0 10px}.demo-input-wrap[data-v-28a3493c]:focus-within{border-color:#409eff}.demo-input[data-v-28a3493c]{border:none;outline:none;padding:8px 6px;flex:1}.prefix[data-v-28a3493c],.suffix[data-v-28a3493c]{color:#909399}.digital-human[data-v-9c25016c]{width:100%;height:100%;display:flex;align-items:center;justify-content:center}.digital-human canvas[data-v-9c25016c]{max-width:100%;max-height:100%}
+.demo-btn[data-v-9a48516a]{padding:6px 14px;border-radius:4px;border:none;cursor:pointer;font-size:14px}.primary[data-v-9a48516a]{background:#409eff;color:#fff}.success[data-v-9a48516a]{background:#67c23a;color:#fff}.warning[data-v-9a48516a]{background:#e6a23c;color:#fff}.demo-input-wrap[data-v-28a3493c]{display:inline-flex;align-items:center;border:1px solid #dcdcdc;border-radius:4px;padding:0 10px}.demo-input-wrap[data-v-28a3493c]:focus-within{border-color:#409eff}.demo-input[data-v-28a3493c]{border:none;outline:none;padding:8px 6px;flex:1}.prefix[data-v-28a3493c],.suffix[data-v-28a3493c]{color:#909399}.digital-human[data-v-9379171c]{width:100%;height:100%;display:flex;align-items:center;justify-content:center}.digital-human canvas[data-v-9379171c]{max-width:100%;max-height:100%}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ruiyun-human",
-  "version": "1.0.11",
+  "version": "1.0.13",
   "type": "module",
   "description": "数字人初步实现",
   "main": "./dist/ruiyun-human.cjs.js",