@pie-players/tts-server-google 0.3.66 → 0.3.68

package/dist/GoogleCloudTTSProvider.d.ts

@@ -119,10 +119,6 @@ export declare class GoogleCloudTTSProvider extends BaseTTSProvider {
      * Extract words from existing SSML (simplified version for v1)
      */
     private extractWordsFromSSML;
-    /**
-     * Escape special XML characters for SSML
-     */
-    private escapeSSML;
     /**
      * Extract speech marks from Google's timepoints
      */
@@ -130,7 +126,6 @@ export declare class GoogleCloudTTSProvider extends BaseTTSProvider {
     /**
      * Detect if text contains SSML markup
      */
-    private detectSSML;
     /**
      * Get available voices from Google Cloud TTS
      */

package/dist/GoogleCloudTTSProvider.js

@@ -164,11 +164,10 @@ export class GoogleCloudTTSProvider extends BaseTTSProvider {
      * Synthesize audio stream only (no speech marks)
      */
     async synthesizeAudio(request, voice) {
-        // Detect if text contains SSML tags
-        const isSsml = this.detectSSML(request.text);
-        if (isSsml && this.enableLogging) {
+        if (this.detectSSML(request.text) && this.enableLogging) {
             console.log("[GoogleCloudTTS] Detected SSML content");
         }
+        const { text, isSsml } = this.applyProsody(request.text, request);
         // Parse voice name to extract language code
         const languageCode = voice.split("-").slice(0, 2).join("-"); // e.g., "en-US" from "en-US-Wavenet-A"
         // Map our audio encoding to Google's enum
@@ -178,7 +177,7 @@ export class GoogleCloudTTSProvider extends BaseTTSProvider {
             OGG_OPUS: "OGG_OPUS",
         };
         const [response] = await this.client.synthesizeSpeech({
-            input: isSsml ? { ssml: request.text } : { text: request.text },
+            input: isSsml ? { ssml: text } : { text },
             voice: {
                 languageCode,
                 name: voice,
@@ -210,10 +209,12 @@ export class GoogleCloudTTSProvider extends BaseTTSProvider {
         // Check if the text is already SSML
         const isUserSSML = this.detectSSML(request.text);
         // If user provided SSML, we need to inject marks within the existing SSML
-        // For simplicity in v1, we'll inject marks for plain text only
+        // For simplicity in v1, we'll inject marks for plain text only. Already-SSML
+        // input skips prosody wrapping too: injecting <prosody> into markup the
+        // caller authored themselves would require parsing it.
         const { ssml, wordMap } = isUserSSML
             ? this.extractWordsFromSSML(request.text)
-            : this.injectSSMLMarks(request.text);
+            : this.injectSSMLMarks(request.text, this.buildProsodyAttrs(request));
         if (this.enableLogging) {
             console.log(`[GoogleCloudTTS] Injected ${wordMap.length} SSML marks`);
         }
@@ -266,7 +267,7 @@ export class GoogleCloudTTSProvider extends BaseTTSProvider {
     /**
      * Inject SSML marks before each word in plain text
      */
-    injectSSMLMarks(text) {
+    injectSSMLMarks(text, prosodyAttrs = "") {
         const words = [];
         const wordRegex = /\b[\w']+\b/g;
         let match;
@@ -279,7 +280,7 @@ export class GoogleCloudTTSProvider extends BaseTTSProvider {
             words.push({ word, start, end, markName });
         }
         // Build SSML with marks
-        let ssml = "<speak>";
+        let ssml = prosodyAttrs ? `<speak><prosody ${prosodyAttrs}>` : "<speak>";
         let lastEnd = 0;
         for (const { word, start, end, markName } of words) {
             // Add text before word (including whitespace and punctuation)
@@ -289,7 +290,10 @@ export class GoogleCloudTTSProvider extends BaseTTSProvider {
             lastEnd = end;
         }
         // Add remaining text
-        ssml += this.escapeSSML(text.slice(lastEnd)) + "</speak>";
+        ssml +=
+            this.escapeSSML(text.slice(lastEnd)) +
+                (prosodyAttrs ? "</prosody>" : "") +
+                "</speak>";
         return { ssml, wordMap: words };
     }
     /**
@@ -304,17 +308,6 @@ export class GoogleCloudTTSProvider extends BaseTTSProvider {
             .trim();
         return this.injectSSMLMarks(plainText);
     }
-    /**
-     * Escape special XML characters for SSML
-     */
-    escapeSSML(text) {
-        return text
-            .replace(/&/g, "&amp;")
-            .replace(/</g, "&lt;")
-            .replace(/>/g, "&gt;")
-            .replace(/"/g, "&quot;")
-            .replace(/'/g, "&apos;");
-    }
     /**
      * Extract speech marks from Google's timepoints
      */
@@ -344,15 +337,6 @@ export class GoogleCloudTTSProvider extends BaseTTSProvider {
     /**
      * Detect if text contains SSML markup
      */
-    detectSSML(text) {
-        return (text.includes("<speak") ||
-            text.includes("<prosody") ||
-            text.includes("<emphasis") ||
-            text.includes("<break") ||
-            text.includes("<phoneme") ||
-            text.includes("<say-as") ||
-            text.includes("<mark"));
-    }
     /**
      * Get available voices from Google Cloud TTS
      */

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@pie-players/tts-server-google",
-	"version": "0.3.66",
+	"version": "0.3.68",
 	"description": "Google Cloud Text-to-Speech provider for server-side TTS with speech marks support",
 	"type": "module",
 	"main": "./dist/index.js",
@@ -40,7 +40,7 @@
 	},
 	"dependencies": {
 		"@google-cloud/text-to-speech": "^6.4.0",
-		"@pie-players/tts-server-core": "0.3.66"
+		"@pie-players/tts-server-core": "0.3.68"
 	},
 	"devDependencies": {
 		"typescript": "^5.9.3",