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

package/dist/PollyServerProvider.d.ts

@@ -46,19 +46,6 @@ export interface PollyProviderConfig extends TTSServerConfig {
      */
     defaultVoice?: string;
 }
-/**
- * AWS Polly Server Provider
- *
- * Provides high-quality neural text-to-speech with precise word-level timing
- * through AWS Polly speech marks.
- *
- * Features:
- * - Native speech marks support (millisecond precision)
- * - Neural and standard voices
- * - 25+ languages
- * - Full SSML support
- * - Parallel audio + speech marks requests
- */
 export declare class PollyServerProvider extends BaseTTSProvider {
     readonly providerId = "aws-polly";
     readonly providerName = "AWS Polly";
@@ -100,6 +87,14 @@ export declare class PollyServerProvider extends BaseTTSProvider {
      * Clearly documents what features are W3C-standard vs AWS-specific.
      */
     getCapabilities(): ServerProviderCapabilities;
+    /**
+     * Map an AWS SDK error to a TTSError, matching the granularity
+     * GoogleCloudTTSProvider gives its own vendor errors: a bare
+     * `PROVIDER_ERROR` for everything hides cases callers (Host A maps
+     * `RATE_LIMIT_EXCEEDED` / `AUTHENTICATION_ERROR` / `INVALID_REQUEST` /
+     * `TEXT_TOO_LONG` onto HTTP status codes) already know how to handle.
+     */
+    private mapPollyErrorToTTSError;
     /**
      * Clean up AWS Polly client
      */

package/dist/PollyServerProvider.js

@@ -17,6 +17,34 @@ import { BaseTTSProvider, TTSError, TTSErrorCode, } from "@pie-players/tts-serve
  * - Full SSML support
  * - Parallel audio + speech marks requests
  */
+/**
+ * Polly's own SSML vocabulary, on top of the standard elements the base provider
+ * knows. `<amazon:effect>` and the `<aws-*>` family are Polly-only.
+ */
+const POLLY_SSML_TAGS = ["<amazon:", "<aws-"];
+/**
+ * AWS SDK exception names (`error.name`) that mean the request itself was
+ * invalid or unsupported, as opposed to a transient/provider-side failure.
+ * @see https://docs.aws.amazon.com/polly/latest/dg/API_Reference.html
+ */
+const POLLY_INVALID_REQUEST_EXCEPTIONS = new Set([
+    "InvalidSampleRateException",
+    "InvalidSsmlException",
+    "LanguageNotSupportedException",
+    "ValidationException",
+    "EngineNotSupportedException",
+    "MarksNotSupportedForFormatException",
+    "SsmlMarksNotSupportedForTextTypeException",
+    "UnsupportedPlsAlphabetException",
+    "UnsupportedPlsLanguageException",
+    "InvalidS3BucketException",
+    "InvalidS3KeyException",
+    "InvalidSnsTopicArnException",
+]);
+const POLLY_RATE_LIMIT_EXCEPTIONS = new Set([
+    "ThrottlingException",
+    "ServiceQuotaExceededException",
+]);
 export class PollyServerProvider extends BaseTTSProvider {
     providerId = "aws-polly";
     providerName = "AWS Polly";
@@ -120,29 +148,22 @@ export class PollyServerProvider extends BaseTTSProvider {
             };
         }
         catch (error) {
-            throw new TTSError(TTSErrorCode.PROVIDER_ERROR, `AWS Polly synthesis failed: ${error instanceof Error ? error.message : String(error)}`, { error, request }, this.providerId);
+            throw this.mapPollyErrorToTTSError(error, { request });
         }
     }
     /**
      * Synthesize audio stream
      */
     async synthesizeAudio(request, voice) {
-        // Detect if text contains SSML tags
-        const isSsml = request.text.includes("<speak") ||
-            request.text.includes("<emphasis") ||
-            request.text.includes("<break") ||
-            request.text.includes("<prosody") ||
-            request.text.includes("<phoneme") ||
-            request.text.includes("<amazon:") ||
-            request.text.includes("<aws-");
-        const textType = isSsml ? "ssml" : "text";
-        if (isSsml) {
+        if (this.detectSSML(request.text, POLLY_SSML_TAGS)) {
             console.log("[PollyServerProvider] Detected SSML content, using TextType: ssml");
         }
+        const { text, isSsml } = this.applyProsody(request.text, request, POLLY_SSML_TAGS);
+        const textType = isSsml ? "ssml" : "text";
         const command = new SynthesizeSpeechCommand({
             Engine: this.engine === "neural" ? Engine.NEURAL : Engine.STANDARD,
             OutputFormat: this.resolveOutputFormat(request),
-            Text: request.text,
+            Text: text,
             TextType: textType,
             VoiceId: voice,
             SampleRate: String(request.sampleRate || 24000),
@@ -172,19 +193,12 @@ export class PollyServerProvider extends BaseTTSProvider {
      * Synthesize speech marks
      */
     async synthesizeSpeechMarks(request, voice) {
-        // Detect if text contains SSML tags (same logic as audio synthesis)
-        const isSsml = request.text.includes("<speak") ||
-            request.text.includes("<emphasis") ||
-            request.text.includes("<break") ||
-            request.text.includes("<prosody") ||
-            request.text.includes("<phoneme") ||
-            request.text.includes("<amazon:") ||
-            request.text.includes("<aws-");
+        const { text, isSsml } = this.applyProsody(request.text, request, POLLY_SSML_TAGS);
         const textType = isSsml ? "ssml" : "text";
         const command = new SynthesizeSpeechCommand({
             Engine: this.engine === "neural" ? Engine.NEURAL : Engine.STANDARD,
             OutputFormat: "json",
-            Text: request.text,
+            Text: text,
             TextType: textType,
             VoiceId: voice,
             SpeechMarkTypes: this.resolveSpeechMarkTypes(request),
@@ -273,7 +287,7 @@ export class PollyServerProvider extends BaseTTSProvider {
             });
         }
         catch (error) {
-            throw new TTSError(TTSErrorCode.PROVIDER_ERROR, `Failed to get voices: ${error instanceof Error ? error.message : String(error)}`, { error }, this.providerId);
+            throw this.mapPollyErrorToTTSError(error);
         }
     }
     /**
@@ -310,6 +324,33 @@ export class PollyServerProvider extends BaseTTSProvider {
             },
         };
     }
+    /**
+     * Map an AWS SDK error to a TTSError, matching the granularity
+     * GoogleCloudTTSProvider gives its own vendor errors: a bare
+     * `PROVIDER_ERROR` for everything hides cases callers (Host A maps
+     * `RATE_LIMIT_EXCEEDED` / `AUTHENTICATION_ERROR` / `INVALID_REQUEST` /
+     * `TEXT_TOO_LONG` onto HTTP status codes) already know how to handle.
+     */
+    mapPollyErrorToTTSError(error, extraDetails) {
+        const err = error;
+        const message = err?.message || String(error);
+        const statusCode = err?.$metadata?.httpStatusCode;
+        const details = { error, ...extraDetails };
+        if (err?.name === "TextLengthExceededException") {
+            return new TTSError(TTSErrorCode.TEXT_TOO_LONG, `AWS Polly text length exceeded: ${message}`, details, this.providerId);
+        }
+        if (err?.name && POLLY_INVALID_REQUEST_EXCEPTIONS.has(err.name)) {
+            return new TTSError(TTSErrorCode.INVALID_REQUEST, `Invalid request to AWS Polly: ${message}`, details, this.providerId);
+        }
+        if ((err?.name && POLLY_RATE_LIMIT_EXCEPTIONS.has(err.name)) ||
+            statusCode === 429) {
+            return new TTSError(TTSErrorCode.RATE_LIMIT_EXCEEDED, `AWS Polly rate limit exceeded: ${message}`, details, this.providerId);
+        }
+        if (statusCode === 401 || statusCode === 403) {
+            return new TTSError(TTSErrorCode.AUTHENTICATION_ERROR, `AWS Polly authentication failed: ${message}`, details, this.providerId);
+        }
+        return new TTSError(TTSErrorCode.PROVIDER_ERROR, `AWS Polly error: ${message}`, details, this.providerId);
+    }
     /**
      * Clean up AWS Polly client
      */

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@pie-players/tts-server-polly",
-	"version": "0.3.66",
+	"version": "0.3.68",
 	"description": "AWS Polly provider for server-side TTS with speech marks support",
 	"type": "module",
 	"main": "./dist/index.js",
@@ -39,7 +39,7 @@
 	},
 	"dependencies": {
 		"@aws-sdk/client-polly": "^3.1033.0",
-		"@pie-players/tts-server-core": "0.3.66"
+		"@pie-players/tts-server-core": "0.3.68"
 	},
 	"devDependencies": {
 		"typescript": "^5.9.3",