ssml-builder-js 2.10.0 → 2.12.0

package/dist/elements.mjs

@@ -36,7 +36,7 @@ import {
   resolveExpressAsStyles,
   updateEditableText,
   validateAzureSsml
-} from "./chunk-BYIZQL2W.mjs";
+} from "./chunk-FHDFRM2F.mjs";
 import "./chunk-6S5ODO6A.mjs";
 
 // packages/ssml-editor-react/src/ssmlInsertions.ts

package/dist/index.d.mts

@@ -1,4 +1,4 @@
-export { AudioElement, AzureDiagnosticCode, AzureLanguageNormalizationOptions, AzureSsmlValidationOptions, AzureValidationOptions, AzureVoiceCatalogMetadata, AzureVoiceDefinition, AzureVoiceMetadata, BookmarkElement, BreakElement, BuildPartialSsmlOptions, CustomElement, EmphasisElement, ExpressAsElement, ExtractSsmlTranslatableTextOptions, FromPlainTextToSsmlOptions, LangElement, LexiconElement, MapSsmlTextNodesOptions, MarkElement, MsttsAudioDurationElement, MsttsEmbeddingElement, MsttsSilenceElement, MsttsTtsEmbeddingElement, MsttsVisemeElement, MsttsVoiceConversionElement, NamedElement, ParagraphElement, PhonemeElement, ProsodyElement, SayAsElement, SentenceElement, SsmlAttributeValue, SsmlAttributes, SsmlBackgroundAudioNode, SsmlBreakElement, SsmlDiagnostic, SsmlDiagnosticSeverity, SsmlDiagnosticSource, SsmlDialogNode, SsmlDocument, SsmlElement, SsmlElementBase, SsmlEmbeddingNode, SsmlExpressAsElement, SsmlNode, SsmlPartialContext, SsmlPartialProsody, SsmlPartialVoice, SsmlPhonemeElement, SsmlProsodyElement, SsmlSayAsElement, SsmlStructureIntegrityResult, SsmlStructureMismatch, SsmlText, SsmlTextNodeContext, SsmlTtsEmbeddingNode, SsmlTurnNode, SsmlValidationError, SsmlVoiceConversionNode, SsmlVoiceElement, SubElement, VoiceElement, WordElement, areAzureLanguagesEquivalent, buildPartialSsml, buildSsml, extractSsmlText, extractSsmlTranslatableText, fromPlainTextToSsml, getAzureVoiceCatalogMetadata, getBuiltInVoiceCatalogMetadata, isValidAzureAudioDuration, mapSsmlTextNodes, normalizeAzureLanguage, parseSsml, validateAzureSsml, validateSsml, validateSsmlStructureIntegrity } from './core.mjs';
+export { AudioElement, AzureDiagnosticCode, AzureLanguageNormalizationOptions, AzureSsmlValidationOptions, AzureValidationOptions, AzureVoiceCatalogMetadata, AzureVoiceDefinition, AzureVoiceMetadata, BookmarkElement, BreakElement, BuildPartialSsmlOptions, CustomElement, EmphasisElement, ExpressAsElement, ExtractSsmlTranslatableTextOptions, FromPlainTextToSsmlOptions, LangElement, LexiconElement, MapSsmlTextNodesOptions, MarkElement, MsttsAudioDurationElement, MsttsEmbeddingElement, MsttsSilenceElement, MsttsTtsEmbeddingElement, MsttsVisemeElement, MsttsVoiceConversionElement, NamedElement, ParagraphElement, PhonemeElement, ProsodyElement, SayAsElement, SentenceElement, SsmlAttributeValue, SsmlAttributes, SsmlBackgroundAudioNode, SsmlBreakElement, SsmlDiagnostic, SsmlDiagnosticSeverity, SsmlDiagnosticSource, SsmlDialogNode, SsmlDocument, SsmlElement, SsmlElementBase, SsmlEmbeddingNode, SsmlExpressAsElement, SsmlNode, SsmlPartialContext, SsmlPartialProsody, SsmlPartialVoice, SsmlPhonemeElement, SsmlProsodyElement, SsmlSayAsElement, SsmlStructureIntegrityResult, SsmlStructureMismatch, SsmlText, SsmlTextNodeContext, SsmlTtsEmbeddingNode, SsmlTurnNode, SsmlValidationError, SsmlVoiceConversionNode, SsmlVoiceElement, SubElement, VoiceElement, WordElement, areAzureLanguagesEquivalent, buildPartialSsml, buildSsml, extractSsmlText, extractSsmlTranslatableText, fromPlainTextToSsml, getAzureVoiceCatalogMetadata, getBuiltInVoiceCatalogMetadata, isValidAzureAudioDuration, mapSsmlTextNodes, normalizeAzureLanguage, parseSsml, splitSsmlDocument, validateAzureSsml, validateSsml, validateSsmlStructureIntegrity } from './core.mjs';
 
 interface TtsConfig {
     signal?: AbortSignal;
@@ -8,6 +8,31 @@ interface TtsConfig {
     region: string;
     outputFormat?: string;
 }
+interface SsmlSynthesisBoundary {
+    text: string;
+    audioOffsetMs: number;
+    durationMs: number;
+}
+interface SsmlSynthesisViseme {
+    visemeId: number;
+    audioOffsetMs: number;
+}
+interface SsmlSynthesisBookmark {
+    name: string;
+    audioOffsetMs: number;
+}
+/** Audio and Azure Speech synchronization events emitted for one SSML request. */
+interface SsmlSynthesisResult {
+    audioData: ArrayBuffer;
+    durationMs: number;
+    boundaries?: SsmlSynthesisBoundary[];
+    /** Alias matching the Azure Speech event name. */
+    wordBoundary?: SsmlSynthesisBoundary[];
+    /** Alias for consumers that use Azure's word-boundary terminology. */
+    wordBoundaries?: SsmlSynthesisBoundary[];
+    visemes?: SsmlSynthesisViseme[];
+    bookmarks?: SsmlSynthesisBookmark[];
+}
 interface AzureTtsLogger {
     debug?: (...args: unknown[]) => void;
     info?: (...args: unknown[]) => void;
@@ -40,8 +65,36 @@ declare class AzureTtsClient {
     #private;
     constructor(options: AzureTtsClientOptions);
     synthesize(ssml: string): Promise<ArrayBuffer>;
+    synthesizeSsml(ssml: string): Promise<SsmlSynthesisResult>;
 }
 
+declare function synthesizeSsml(ssml: string, config: TtsConfig): Promise<SsmlSynthesisResult>;
+/** Backward-compatible audio-only synthesis helper. */
 declare function synthesizeSpeech(ssml: string, config: TtsConfig): Promise<ArrayBuffer>;
 
-export { AzureTtsClient, type AzureTtsClientOptions, AzureTtsError, type AzureTtsLogger, AzureTtsSdkError, type TtsConfig, synthesizeSpeech };
+interface FetchAzureVoiceCatalogOptions {
+    apiKey: string;
+    region: string | string[];
+}
+interface AzureVoiceCatalogVoice {
+    name: string;
+    locale: string;
+    secondaryLocales?: readonly string[];
+    styles?: readonly string[];
+    regions: readonly string[];
+    status?: "ga" | "preview" | "deprecated";
+}
+interface FetchedAzureVoiceCatalogMetadata {
+    voiceCount: number;
+    generatedAt: string;
+    apiVersion: string;
+    regions: readonly string[];
+}
+interface AzureVoiceCatalog {
+    voices: readonly AzureVoiceCatalogVoice[];
+    metadata: FetchedAzureVoiceCatalogMetadata;
+}
+/** Fetches and deduplicates the current Azure Speech voice catalog for one or more regions. */
+declare function fetchAzureVoiceCatalog(options: FetchAzureVoiceCatalogOptions): Promise<AzureVoiceCatalog>;
+
+export { AzureTtsClient, type AzureTtsClientOptions, AzureTtsError, type AzureTtsLogger, AzureTtsSdkError, type AzureVoiceCatalog, type AzureVoiceCatalogVoice, type FetchAzureVoiceCatalogOptions, type FetchedAzureVoiceCatalogMetadata, type SsmlSynthesisBookmark, type SsmlSynthesisBoundary, type SsmlSynthesisResult, type SsmlSynthesisViseme, type TtsConfig, fetchAzureVoiceCatalog, synthesizeSpeech, synthesizeSsml };

package/dist/index.d.ts

@@ -1,4 +1,4 @@
-export { AudioElement, AzureDiagnosticCode, AzureLanguageNormalizationOptions, AzureSsmlValidationOptions, AzureValidationOptions, AzureVoiceCatalogMetadata, AzureVoiceDefinition, AzureVoiceMetadata, BookmarkElement, BreakElement, BuildPartialSsmlOptions, CustomElement, EmphasisElement, ExpressAsElement, ExtractSsmlTranslatableTextOptions, FromPlainTextToSsmlOptions, LangElement, LexiconElement, MapSsmlTextNodesOptions, MarkElement, MsttsAudioDurationElement, MsttsEmbeddingElement, MsttsSilenceElement, MsttsTtsEmbeddingElement, MsttsVisemeElement, MsttsVoiceConversionElement, NamedElement, ParagraphElement, PhonemeElement, ProsodyElement, SayAsElement, SentenceElement, SsmlAttributeValue, SsmlAttributes, SsmlBackgroundAudioNode, SsmlBreakElement, SsmlDiagnostic, SsmlDiagnosticSeverity, SsmlDiagnosticSource, SsmlDialogNode, SsmlDocument, SsmlElement, SsmlElementBase, SsmlEmbeddingNode, SsmlExpressAsElement, SsmlNode, SsmlPartialContext, SsmlPartialProsody, SsmlPartialVoice, SsmlPhonemeElement, SsmlProsodyElement, SsmlSayAsElement, SsmlStructureIntegrityResult, SsmlStructureMismatch, SsmlText, SsmlTextNodeContext, SsmlTtsEmbeddingNode, SsmlTurnNode, SsmlValidationError, SsmlVoiceConversionNode, SsmlVoiceElement, SubElement, VoiceElement, WordElement, areAzureLanguagesEquivalent, buildPartialSsml, buildSsml, extractSsmlText, extractSsmlTranslatableText, fromPlainTextToSsml, getAzureVoiceCatalogMetadata, getBuiltInVoiceCatalogMetadata, isValidAzureAudioDuration, mapSsmlTextNodes, normalizeAzureLanguage, parseSsml, validateAzureSsml, validateSsml, validateSsmlStructureIntegrity } from './core.js';
+export { AudioElement, AzureDiagnosticCode, AzureLanguageNormalizationOptions, AzureSsmlValidationOptions, AzureValidationOptions, AzureVoiceCatalogMetadata, AzureVoiceDefinition, AzureVoiceMetadata, BookmarkElement, BreakElement, BuildPartialSsmlOptions, CustomElement, EmphasisElement, ExpressAsElement, ExtractSsmlTranslatableTextOptions, FromPlainTextToSsmlOptions, LangElement, LexiconElement, MapSsmlTextNodesOptions, MarkElement, MsttsAudioDurationElement, MsttsEmbeddingElement, MsttsSilenceElement, MsttsTtsEmbeddingElement, MsttsVisemeElement, MsttsVoiceConversionElement, NamedElement, ParagraphElement, PhonemeElement, ProsodyElement, SayAsElement, SentenceElement, SsmlAttributeValue, SsmlAttributes, SsmlBackgroundAudioNode, SsmlBreakElement, SsmlDiagnostic, SsmlDiagnosticSeverity, SsmlDiagnosticSource, SsmlDialogNode, SsmlDocument, SsmlElement, SsmlElementBase, SsmlEmbeddingNode, SsmlExpressAsElement, SsmlNode, SsmlPartialContext, SsmlPartialProsody, SsmlPartialVoice, SsmlPhonemeElement, SsmlProsodyElement, SsmlSayAsElement, SsmlStructureIntegrityResult, SsmlStructureMismatch, SsmlText, SsmlTextNodeContext, SsmlTtsEmbeddingNode, SsmlTurnNode, SsmlValidationError, SsmlVoiceConversionNode, SsmlVoiceElement, SubElement, VoiceElement, WordElement, areAzureLanguagesEquivalent, buildPartialSsml, buildSsml, extractSsmlText, extractSsmlTranslatableText, fromPlainTextToSsml, getAzureVoiceCatalogMetadata, getBuiltInVoiceCatalogMetadata, isValidAzureAudioDuration, mapSsmlTextNodes, normalizeAzureLanguage, parseSsml, splitSsmlDocument, validateAzureSsml, validateSsml, validateSsmlStructureIntegrity } from './core.js';
 
 interface TtsConfig {
     signal?: AbortSignal;
@@ -8,6 +8,31 @@ interface TtsConfig {
     region: string;
     outputFormat?: string;
 }
+interface SsmlSynthesisBoundary {
+    text: string;
+    audioOffsetMs: number;
+    durationMs: number;
+}
+interface SsmlSynthesisViseme {
+    visemeId: number;
+    audioOffsetMs: number;
+}
+interface SsmlSynthesisBookmark {
+    name: string;
+    audioOffsetMs: number;
+}
+/** Audio and Azure Speech synchronization events emitted for one SSML request. */
+interface SsmlSynthesisResult {
+    audioData: ArrayBuffer;
+    durationMs: number;
+    boundaries?: SsmlSynthesisBoundary[];
+    /** Alias matching the Azure Speech event name. */
+    wordBoundary?: SsmlSynthesisBoundary[];
+    /** Alias for consumers that use Azure's word-boundary terminology. */
+    wordBoundaries?: SsmlSynthesisBoundary[];
+    visemes?: SsmlSynthesisViseme[];
+    bookmarks?: SsmlSynthesisBookmark[];
+}
 interface AzureTtsLogger {
     debug?: (...args: unknown[]) => void;
     info?: (...args: unknown[]) => void;
@@ -40,8 +65,36 @@ declare class AzureTtsClient {
     #private;
     constructor(options: AzureTtsClientOptions);
     synthesize(ssml: string): Promise<ArrayBuffer>;
+    synthesizeSsml(ssml: string): Promise<SsmlSynthesisResult>;
 }
 
+declare function synthesizeSsml(ssml: string, config: TtsConfig): Promise<SsmlSynthesisResult>;
+/** Backward-compatible audio-only synthesis helper. */
 declare function synthesizeSpeech(ssml: string, config: TtsConfig): Promise<ArrayBuffer>;
 
-export { AzureTtsClient, type AzureTtsClientOptions, AzureTtsError, type AzureTtsLogger, AzureTtsSdkError, type TtsConfig, synthesizeSpeech };
+interface FetchAzureVoiceCatalogOptions {
+    apiKey: string;
+    region: string | string[];
+}
+interface AzureVoiceCatalogVoice {
+    name: string;
+    locale: string;
+    secondaryLocales?: readonly string[];
+    styles?: readonly string[];
+    regions: readonly string[];
+    status?: "ga" | "preview" | "deprecated";
+}
+interface FetchedAzureVoiceCatalogMetadata {
+    voiceCount: number;
+    generatedAt: string;
+    apiVersion: string;
+    regions: readonly string[];
+}
+interface AzureVoiceCatalog {
+    voices: readonly AzureVoiceCatalogVoice[];
+    metadata: FetchedAzureVoiceCatalogMetadata;
+}
+/** Fetches and deduplicates the current Azure Speech voice catalog for one or more regions. */
+declare function fetchAzureVoiceCatalog(options: FetchAzureVoiceCatalogOptions): Promise<AzureVoiceCatalog>;
+
+export { AzureTtsClient, type AzureTtsClientOptions, AzureTtsError, type AzureTtsLogger, AzureTtsSdkError, type AzureVoiceCatalog, type AzureVoiceCatalogVoice, type FetchAzureVoiceCatalogOptions, type FetchedAzureVoiceCatalogMetadata, type SsmlSynthesisBookmark, type SsmlSynthesisBoundary, type SsmlSynthesisResult, type SsmlSynthesisViseme, type TtsConfig, fetchAzureVoiceCatalog, synthesizeSpeech, synthesizeSsml };

package/dist/index.js

@@ -45,6 +45,7 @@ __export(src_exports, {
   buildSsml: () => buildSsml,
   extractSsmlText: () => extractSsmlText,
   extractSsmlTranslatableText: () => extractSsmlTranslatableText,
+  fetchAzureVoiceCatalog: () => fetchAzureVoiceCatalog,
   fromPlainTextToSsml: () => fromPlainTextToSsml,
   getAzureVoiceCatalogMetadata: () => getAzureVoiceCatalogMetadata,
   getBuiltInVoiceCatalogMetadata: () => getBuiltInVoiceCatalogMetadata,
@@ -52,7 +53,9 @@ __export(src_exports, {
   mapSsmlTextNodes: () => mapSsmlTextNodes,
   normalizeAzureLanguage: () => normalizeAzureLanguage,
   parseSsml: () => parseSsml,
+  splitSsmlDocument: () => splitSsmlDocument,
   synthesizeSpeech: () => synthesizeSpeech,
+  synthesizeSsml: () => synthesizeSsml,
   validateAzureSsml: () => validateAzureSsml,
   validateSsml: () => validateSsml,
   validateSsmlStructureIntegrity: () => validateSsmlStructureIntegrity
@@ -986,6 +989,104 @@ function buildPartialSsml(textOrOptions, context) {
   return serializePartialSsml(textOrOptions.text, textOrOptions);
 }
 
+// packages/ssml-core/src/split.ts
+var DEFAULT_MAX_LENGTH = 1e4;
+function cloneElement(element, children) {
+  return { ...element, children };
+}
+function documentWithChildren(document, children) {
+  return buildSsml({ ...document, children });
+}
+function wrapWithContext(node, context) {
+  return context.reduceRight((current, parent) => cloneElement(parent, [current]), node);
+}
+function documentWithNode(document, node, context) {
+  return documentWithChildren(document, [wrapWithContext(node, context)]);
+}
+function splitTextNode(text, document, context, maxLength) {
+  if (!text) return [text];
+  const parts = [];
+  let start = 0;
+  while (start < text.length) {
+    let end = start + 1;
+    let bestEnd = end;
+    while (end <= text.length) {
+      if (documentWithNode(document, text.slice(start, end), context).length > maxLength) break;
+      bestEnd = end;
+      end += 1;
+    }
+    if (bestEnd === start) {
+      throw new RangeError("maxLength is too small to contain the SSML document wrapper");
+    }
+    const segment = text.slice(start, bestEnd);
+    const boundary = Math.max(segment.lastIndexOf(" "), segment.lastIndexOf("\n"), segment.lastIndexOf("	"));
+    const splitEnd = boundary > 0 ? start + boundary + 1 : bestEnd;
+    parts.push(text.slice(start, splitEnd));
+    start = splitEnd;
+  }
+  return parts;
+}
+function splitNode(document, node, maxLength, context = []) {
+  if (typeof node === "string" || node.type === "text") {
+    const value = typeof node === "string" ? node : node.value;
+    if (documentWithNode(document, node, context).length <= maxLength) return [node];
+    return splitTextNode(value, document, context, maxLength);
+  }
+  if (documentWithNode(document, node, context).length <= maxLength) return [node];
+  const children = node.children ?? [];
+  if (children.length === 0) {
+    throw new RangeError(`maxLength is too small to contain <${node.type}>`);
+  }
+  const splitChildren = children.flatMap((child) => splitNode(document, child, maxLength, [...context, node]));
+  const parts = [];
+  let group = [];
+  const flush = () => {
+    if (group.length > 0) {
+      parts.push(cloneElement(node, group));
+      group = [];
+    }
+  };
+  for (const child of splitChildren) {
+    const candidate = cloneElement(node, [...group, child]);
+    if (documentWithNode(document, candidate, context).length <= maxLength) {
+      group.push(child);
+      continue;
+    }
+    flush();
+    if (documentWithNode(document, cloneElement(node, [child]), context).length > maxLength) {
+      throw new RangeError(`maxLength is too small to contain <${node.type}>`);
+    }
+    group.push(child);
+  }
+  flush();
+  return parts;
+}
+function splitSsmlDocument(ssml, maxLength = DEFAULT_MAX_LENGTH) {
+  if (!Number.isInteger(maxLength) || maxLength <= 0) {
+    throw new RangeError("maxLength must be a positive integer");
+  }
+  const document = parseSsml(ssml);
+  if (ssml.length <= maxLength) return [ssml];
+  const children = document.children ?? [];
+  const splitChildren = children.flatMap((child) => splitNode(document, child, maxLength));
+  const chunks = [];
+  let group = [];
+  for (const child of splitChildren) {
+    const candidate = [...group, child];
+    if (documentWithChildren(document, candidate).length <= maxLength) {
+      group = candidate;
+      continue;
+    }
+    if (group.length > 0) chunks.push(group);
+    group = [child];
+    if (documentWithChildren(document, group).length > maxLength) {
+      throw new RangeError("maxLength is too small to contain the SSML document wrapper");
+    }
+  }
+  if (group.length > 0) chunks.push(group);
+  return chunks.map((chunk) => documentWithChildren(document, chunk));
+}
+
 // packages/ssml-core/src/validation.ts
 var PARSER_POSITION_SUFFIX = / at position (\d+)$/;
 function validateSsml(xmlString) {
@@ -1501,6 +1602,9 @@ var AZURE_VOICE_DEFINITIONS = [
   },
   { name: "es-ES-ElviraNeural", locale: "es-ES" },
   { name: "fil-PH-AngeloNeural", locale: "fil-PH" },
+  { name: "fil-PH-Angelo:DragonHDLatestNeural", locale: "fil-PH" },
+  { name: "fil-PH-BlessicaNeural", locale: "fil-PH" },
+  { name: "fil-PH-Blessica:DragonHDLatestNeural", locale: "fil-PH" },
   { name: "fr-FR-DeniseNeural", locale: "fr-FR", styles: ["cheerful", "sad"] },
   { name: "fr-FR-HenriNeural", locale: "fr-FR", styles: ["cheerful", "sad"] },
   { name: "id-ID-GadisNeural", locale: "id-ID" },
@@ -1597,6 +1701,18 @@ var ALLOWED_SILENCE_TYPES = /* @__PURE__ */ new Set([
   "Enumerationcomma"
 ]);
 var ALLOWED_VISEME_TYPES = /* @__PURE__ */ new Set(["redlips_front", "FacialExpression"]);
+var DEFAULT_PREVIEW_TAGS = /* @__PURE__ */ new Set(["mstts:voiceconversion"]);
+function featureStatusForTag(name, options) {
+  const tagName = canonicalTagName(name);
+  const configured = Object.entries(options.tagStatuses ?? {}).find(
+    ([candidate]) => canonicalTagName(candidate) === tagName
+  )?.[1];
+  if (configured) return configured;
+  if ((options.previewTags ?? [...DEFAULT_PREVIEW_TAGS]).some((candidate) => canonicalTagName(candidate) === tagName))
+    return "preview";
+  if ((options.deprecatedTags ?? []).some((candidate) => canonicalTagName(candidate) === tagName)) return "deprecated";
+  return void 0;
+}
 function decodeAttribute(value) {
   return value.replace(
     /&(?:amp|apos|gt|lt|quot);/gi,
@@ -1665,6 +1781,7 @@ function tokenizeElements(source) {
       attributes,
       childElementIndex,
       end,
+      depth: openElements.length + 1,
       name: tokenName,
       parentName: parent?.name,
       parentVoiceName,
@@ -1713,9 +1830,9 @@ function isValidAzureAudioDuration(value) {
   return Number(clock[1]) > 0 || Number(clock[2]) > 0 || Number(clock[3]) > 0 || Number(clock[4] ?? 0) > 0;
 }
 function isValidAzureBackgroundAudioDuration(value) {
-  const match = /^(\d+(?:\.\d+)?)(ms|s)?$/i.exec(value.trim());
+  const match = /^(\d+)$/.exec(value.trim());
   if (!match) return false;
-  const milliseconds = Number(match[1]) * (match[2]?.toLowerCase() === "s" ? 1e3 : 1);
+  const milliseconds = Number(match[1]);
   return Number.isFinite(milliseconds) && milliseconds >= 0 && milliseconds <= 1e4;
 }
 function attr(token, name) {
@@ -1855,13 +1972,18 @@ function validateAudioSource(token, source, diagnostics, options, elementName2)
     addDiagnostic(diagnostics, source, token.start, `<${elementName2} src> must be an absolute HTTP(S) URL.`);
     return;
   }
+  if (parsed.username || parsed.password)
+    addDiagnostic(diagnostics, source, token.start, `<${elementName2} src> must not contain URL credentials.`);
   if (parsed.protocol !== "https:" && !(options.allowHttpAudio && parsed.protocol === "http:"))
     addDiagnostic(diagnostics, source, token.start, `<${elementName2} src> must use HTTPS.`);
   const isAllowedOrigin = options.allowedAudioOrigins?.some((allowedOrigin) => {
     try {
-      return new URL(allowedOrigin).origin === parsed.origin;
+      const configured = new URL(allowedOrigin);
+      if (configured.protocol !== "https:" && configured.protocol !== "http:" || configured.username || configured.password || configured.pathname !== "/" || configured.search || configured.hash)
+        return false;
+      return configured.origin === parsed.origin;
     } catch {
-      return allowedOrigin === parsed.origin;
+      return false;
     }
   }) ?? false;
   if (options.allowedAudioOrigins && !isAllowedOrigin)
@@ -1876,6 +1998,25 @@ function validateAudioSource(token, source, diagnostics, options, elementName2)
 }
 function validateElement(token, source, diagnostics, voiceName, options, voiceCatalog) {
   const name = token.name.toLowerCase();
+  const tagStatus = featureStatusForTag(token.name, options);
+  if (tagStatus === "preview")
+    addDiagnostic(
+      diagnostics,
+      source,
+      token.start,
+      `<${token.name}> is an Azure Speech preview feature and may change or require preview access.`,
+      "warning",
+      "azure-preview-tag"
+    );
+  if (tagStatus === "deprecated")
+    addDiagnostic(
+      diagnostics,
+      source,
+      token.start,
+      `<${token.name}> is deprecated by Azure Speech; migrate to a supported alternative.`,
+      "info",
+      "azure-deprecated-tag"
+    );
   if (name === "voice" && !attr(token, "name")?.trim())
     addDiagnostic(diagnostics, source, token.start, '<voice> requires a non-empty "name" attribute.');
   if (name === "break") {
@@ -2050,6 +2191,9 @@ function validateAzureSsml(ssml, options = {}) {
   const maxLength = options.maxLength ?? 1e4;
   if (ssml.length > maxLength)
     addDiagnostic(diagnostics, ssml, maxLength, `SSML exceeds the maximum length of ${maxLength} characters.`);
+  if (options.maxXmlDepth !== void 0 && (!Number.isInteger(options.maxXmlDepth) || options.maxXmlDepth <= 0)) {
+    addDiagnostic(diagnostics, ssml, 0, "maxXmlDepth must be a positive integer.");
+  }
   try {
     parseSsml(ssml);
   } catch (error) {
@@ -2059,6 +2203,18 @@ function validateAzureSsml(ssml, options = {}) {
     return diagnostics;
   }
   const tokens = tokenizeElements(ssml);
+  if (options.maxXmlDepth !== void 0) {
+    for (const token of tokens) {
+      if (token.depth > options.maxXmlDepth) {
+        addDiagnostic(
+          diagnostics,
+          ssml,
+          token.start,
+          `XML nesting depth ${token.depth} exceeds the configured maximum of ${options.maxXmlDepth}.`
+        );
+      }
+    }
+  }
   const speak = tokens.find((token) => token.name.toLowerCase() === "speak");
   const voices = tokens.filter((token) => token.name.toLowerCase() === "voice");
   const backgroundAudioTokens = tokens.filter((token) => token.name.toLowerCase() === "mstts:backgroundaudio");
@@ -2087,6 +2243,24 @@ function validateAzureSsml(ssml, options = {}) {
     const name = attr(token, "name")?.trim();
     const language = attr(token, "xml:lang")?.trim() || (speak ? attr(speak, "xml:lang")?.trim() : void 0);
     const definition = name ? voiceCatalog.get(name.toLowerCase()) : void 0;
+    if (name && definition?.status === "preview")
+      addDiagnostic(
+        diagnostics,
+        ssml,
+        token.start,
+        `Voice "${name}" is an Azure Speech preview voice and may change or require preview access.`,
+        "warning",
+        "azure-preview-voice"
+      );
+    if (name && definition?.status === "deprecated")
+      addDiagnostic(
+        diagnostics,
+        ssml,
+        token.start,
+        `Voice "${name}" is deprecated by Azure Speech; migrate to a supported voice.`,
+        "info",
+        "azure-deprecated-voice"
+      );
     if (name && !definition && policySeverity)
       addDiagnostic(
         diagnostics,
@@ -2249,7 +2423,8 @@ function closeSpeechResources(speechConfig, synthesizer) {
   } catch {
   }
 }
-async function synthesizeSpeech(ssml, config) {
+var ticksToMilliseconds = (ticks) => Math.max(0, ticks) / 1e4;
+async function synthesizeSsml(ssml, config) {
   if (config.signal?.aborted) {
     throw createSpeechSdkError("Speech synthesis was cancelled.");
   }
@@ -2276,6 +2451,22 @@ async function synthesizeSpeech(ssml, config) {
       closeResources();
       reject(createSpeechSdkError(error));
     };
+    const boundaries = [];
+    const visemes = [];
+    const bookmarks = [];
+    synthesizer.wordBoundary = (_sender, event) => {
+      boundaries.push({
+        text: event.text,
+        audioOffsetMs: ticksToMilliseconds(event.audioOffset),
+        durationMs: ticksToMilliseconds(event.duration)
+      });
+    };
+    synthesizer.visemeReceived = (_sender, event) => {
+      visemes.push({ visemeId: event.visemeId, audioOffsetMs: ticksToMilliseconds(event.audioOffset) });
+    };
+    synthesizer.bookmarkReached = (_sender, event) => {
+      bookmarks.push({ name: event.text, audioOffsetMs: ticksToMilliseconds(event.audioOffset) });
+    };
     const cb = (result) => {
       if (settled) return;
       const { reason, errorDetails } = result;
@@ -2287,7 +2478,20 @@ async function synthesizeSpeech(ssml, config) {
       settled = true;
       cleanup();
       closeResources();
-      resolve(result.audioData);
+      const eventDurationMs = Math.max(
+        0,
+        ...(boundaries ?? []).map((boundary) => boundary.audioOffsetMs + boundary.durationMs),
+        ...(visemes ?? []).map((viseme) => viseme.audioOffsetMs),
+        ...(bookmarks ?? []).map((bookmark) => bookmark.audioOffsetMs)
+      );
+      const durationMs = result.audioDuration ? ticksToMilliseconds(result.audioDuration) : eventDurationMs;
+      resolve({
+        audioData: result.audioData,
+        durationMs,
+        ...boundaries.length > 0 ? { boundaries, wordBoundary: boundaries, wordBoundaries: boundaries } : {},
+        ...visemes.length > 0 ? { visemes } : {},
+        ...bookmarks.length > 0 ? { bookmarks } : {}
+      });
     };
     try {
       if (config.signal) {
@@ -2306,6 +2510,9 @@ async function synthesizeSpeech(ssml, config) {
     }
   });
 }
+async function synthesizeSpeech(ssml, config) {
+  return (await synthesizeSsml(ssml, config)).audioData;
+}
 
 // packages/azure-tts-client/src/client.ts
 var ENDPOINT_TEMPLATE = "https://{region}.tts.speech.microsoft.com/cognitiveservices/v1";
@@ -2322,8 +2529,92 @@ var AzureTtsClient = class {
     const config = { endpoint, region, subscriptionKey, outputFormat, signal, timeoutMs };
     return synthesizeSpeech(ssml, config);
   }
+  async synthesizeSsml(ssml) {
+    const { region, subscriptionKey, outputFormat, signal, timeoutMs } = __privateGet(this, _options);
+    const endpoint = __privateGet(this, _options).endpoint?.trim() || ENDPOINT_TEMPLATE.replace("{region}", region);
+    __privateGet(this, _options).logger?.debug?.("Using Azure TTS endpoint:", endpoint);
+    return synthesizeSsml(ssml, { endpoint, region, subscriptionKey, outputFormat, signal, timeoutMs });
+  }
 };
 _options = new WeakMap();
+
+// packages/azure-tts-client/src/voiceCatalog.ts
+var AZURE_VOICE_API_VERSION = "2025-10-01";
+function stringValue(value) {
+  return typeof value === "string" && value.trim() ? value.trim() : void 0;
+}
+function stringList(value) {
+  if (!Array.isArray(value)) return [];
+  return [...new Set(value.map(stringValue).filter((item) => item !== void 0))];
+}
+function normalizeStatus(value) {
+  const status = stringValue(value)?.toLowerCase();
+  if (status === "preview" || status === "deprecated" || status === "ga") return status;
+  return void 0;
+}
+function normalizeRegions(region) {
+  const regions = Array.isArray(region) ? region : [region];
+  const result = [...new Set(regions.map((item) => item.trim()).filter(Boolean))];
+  if (result.length === 0) throw new TypeError("At least one Azure Speech region is required.");
+  return result;
+}
+async function fetchRegionVoices(region, apiKey) {
+  const endpoint = `https://${encodeURIComponent(region)}.tts.speech.microsoft.com/cognitiveservices/voices/list`;
+  const response = await fetch(endpoint, {
+    headers: {
+      Accept: "application/json",
+      "Ocp-Apim-Subscription-Key": apiKey
+    }
+  });
+  if (!response.ok) {
+    throw new Error(`Azure List Voices API request failed for region "${region}" with HTTP ${response.status}.`);
+  }
+  const payload = await response.json();
+  if (!Array.isArray(payload)) throw new Error(`Azure List Voices API returned an invalid response for "${region}".`);
+  return payload.filter((item) => Boolean(item && typeof item === "object"));
+}
+async function fetchAzureVoiceCatalog(options) {
+  if (!options || typeof options.apiKey !== "string" || !options.apiKey.trim())
+    throw new TypeError("An Azure Speech API key is required.");
+  const regions = normalizeRegions(options.region);
+  const payloads = await Promise.all(regions.map((region) => fetchRegionVoices(region, options.apiKey)));
+  const voices = /* @__PURE__ */ new Map();
+  for (let regionIndex = 0; regionIndex < payloads.length; regionIndex += 1) {
+    const region = regions[regionIndex];
+    for (const record of payloads[regionIndex]) {
+      const name = stringValue(record.ShortName) ?? stringValue(record.Name);
+      const locale = stringValue(record.Locale);
+      if (!name || !locale) continue;
+      const key = name.toLowerCase();
+      const existing = voices.get(key);
+      const secondaryLocales = stringList(record.SecondaryLocaleList);
+      const styles = stringList(record.StyleList);
+      const status = normalizeStatus(record.Status);
+      const merged = {
+        name: existing?.name ?? name,
+        locale: existing?.locale ?? locale,
+        regions: [.../* @__PURE__ */ new Set([...existing?.regions ?? [], region])]
+      };
+      const mergedSecondaryLocales = [.../* @__PURE__ */ new Set([...existing?.secondaryLocales ?? [], ...secondaryLocales])];
+      if (mergedSecondaryLocales.length > 0) merged.secondaryLocales = mergedSecondaryLocales;
+      const mergedStyles = [.../* @__PURE__ */ new Set([...existing?.styles ?? [], ...styles])];
+      if (mergedStyles.length > 0) merged.styles = mergedStyles;
+      if (status) merged.status = status;
+      else if (existing?.status) merged.status = existing.status;
+      voices.set(key, merged);
+    }
+  }
+  const sortedVoices = [...voices.values()].sort((first, second) => first.name.localeCompare(second.name));
+  return {
+    voices: sortedVoices,
+    metadata: {
+      voiceCount: sortedVoices.length,
+      generatedAt: (/* @__PURE__ */ new Date()).toISOString(),
+      apiVersion: AZURE_VOICE_API_VERSION,
+      regions
+    }
+  };
+}
 // Annotate the CommonJS export names for ESM import in node:
 0 && (module.exports = {
   AzureTtsClient,
@@ -2334,6 +2625,7 @@ _options = new WeakMap();
   buildSsml,
   extractSsmlText,
   extractSsmlTranslatableText,
+  fetchAzureVoiceCatalog,
   fromPlainTextToSsml,
   getAzureVoiceCatalogMetadata,
   getBuiltInVoiceCatalogMetadata,
@@ -2341,7 +2633,9 @@ _options = new WeakMap();
   mapSsmlTextNodes,
   normalizeAzureLanguage,
   parseSsml,
+  splitSsmlDocument,
   synthesizeSpeech,
+  synthesizeSsml,
   validateAzureSsml,
   validateSsml,
   validateSsmlStructureIntegrity