@mux/ai 0.23.0 → 0.24.0

package/dist/{index-ePt2EMpB.d.ts → index-C5zhbQHK.d.ts}

@@ -834,12 +834,14 @@ declare function censorVttContent(rawVtt: string, profanity: string[], mode: Cen
  */
 declare function applyOverrideLists(detected: string[], alwaysCensor: string[], neverCensor: string[]): string[];
 /**
- * Applies static find/replace pairs to cue text lines in raw VTT content
- * using word-boundary regex. Defaults to case-sensitive matching since static
- * replacements typically target specific known strings; each entry may opt
- * into case-insensitive matching via `caseSensitive: false`. Headers,
- * timestamps, and cue identifiers are untouched. Returns the edited VTT and
- * the total number of replacements.
+ * Applies static find/replace pairs to cue text blocks in raw VTT content.
+ * Whitespace in the find string can match VTT line wrapping, and word-character
+ * edges are constrained to avoid replacing substrings inside words.
+ * Defaults to case-sensitive matching since static replacements typically
+ * target specific known strings; each entry may opt into case-insensitive
+ * matching via `caseSensitive: false`. Headers, timestamps, and cue
+ * identifiers are untouched. Returns the edited VTT and the total number of
+ * replacements.
  */
 declare function applyReplacements(rawVtt: string, replacements: CaptionReplacement[]): {
     editedVtt: string;

package/dist/index.d.ts

@@ -1,4 +1,4 @@
-export { I as ISO639_1, a as ISO639_3, L as LanguageCodePair, S as SupportedISO639_1, b as SupportedISO639_3, g as getLanguageCodePair, c as getLanguageName, i as isUndeterminedLanguageCode, d as isValidISO639_1, e as isValidISO639_3, t as toISO639_1, f as toISO639_3, h as workflows } from './index-ePt2EMpB.js';
+export { I as ISO639_1, a as ISO639_3, L as LanguageCodePair, S as SupportedISO639_1, b as SupportedISO639_3, g as getLanguageCodePair, c as getLanguageName, i as isUndeterminedLanguageCode, d as isValidISO639_1, e as isValidISO639_3, t as toISO639_1, f as toISO639_3, h as workflows } from './index-C5zhbQHK.js';
 import { W as WorkflowCredentials, S as StoragePutObjectInput, a as StoragePresignGetObjectInput } from './shots-mGJfYGiO.js';
 export { A as AssetTextTrack, C as ChunkEmbedding, b as ChunkingStrategy, E as Encrypted, c as EncryptedPayload, I as ImageSubmissionMode, M as MuxAIOptions, d as MuxAsset, P as PlaybackAsset, e as PlaybackPolicy, f as StorageAdapter, T as TextChunk, g as TokenChunkingConfig, h as TokenUsage, i as ToneType, U as UsageMetadata, V as VTTChunkingConfig, j as VideoEmbeddingsResult, k as WorkflowCredentialsInput, l as WorkflowMuxClient, m as decryptFromWorkflow, n as encryptForWorkflow } from './shots-mGJfYGiO.js';
 import { WORKFLOW_SERIALIZE, WORKFLOW_DESERIALIZE } from '@workflow/serde';
@@ -9,7 +9,7 @@ import '@ai-sdk/google';
 import '@ai-sdk/openai';
 import '@mux/mux-node';
 
-var version = "0.23.0";
+var version = "0.24.0";
 
 type MuxAiErrorType = "validation_error" | "processing_error" | "timeout_error";
 /**

package/dist/index.js

@@ -5,7 +5,7 @@ var __export = (target, all) => {
 };
 
 // package.json
-var version = "0.23.0";
+var version = "0.24.0";
 
 // src/lib/language-codes.ts
 var ISO639_1_TO_3 = {
@@ -2180,7 +2180,8 @@ function getReliableLanguageCode(track, confidenceThreshold = LOW_CONFIDENCE_THR
     return void 0;
   if (isUndeterminedLanguageCode(track.language_code))
     return void 0;
-  if (track.auto_language_confidence !== void 0 && track.auto_language_confidence < confidenceThreshold) {
+  const { auto_language_confidence: autoLanguageConfidence } = track;
+  if (autoLanguageConfidence !== void 0 && autoLanguageConfidence < confidenceThreshold) {
     return void 0;
   }
   return track.language_code;
@@ -4128,12 +4129,64 @@ function transformCueText(rawVtt, transform) {
   });
   return transformed.join("\n");
 }
+function transformCueTextBlocks(rawVtt, transform) {
+  const lines = rawVtt.split("\n");
+  const transformed = [];
+  let cueTextLines;
+  let currentCueStartTime = 0;
+  const flushCueText = () => {
+    if (!cueTextLines)
+      return;
+    if (cueTextLines.length > 0) {
+      transformed.push(...transform(cueTextLines.join("\n"), currentCueStartTime).split("\n"));
+    }
+    cueTextLines = void 0;
+  };
+  for (const line of lines) {
+    if (line.includes("-->")) {
+      flushCueText();
+      const startTimestamp = line.split("-->")[0].trim();
+      currentCueStartTime = vttTimestampToSeconds(startTimestamp);
+      transformed.push(line);
+      cueTextLines = [];
+      continue;
+    }
+    if (cueTextLines) {
+      if (line.trim() === "") {
+        flushCueText();
+        transformed.push(line);
+      } else {
+        cueTextLines.push(line);
+      }
+      continue;
+    }
+    transformed.push(line);
+  }
+  flushCueText();
+  return transformed.join("\n");
+}
+function escapeRegex(text) {
+  return text.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
+}
+function buildStaticReplacementRegex(find, caseSensitive) {
+  const pattern = find.split(/(\s+)/).map((part) => /\s+/.test(part) ? "\\s+" : escapeRegex(part)).join("");
+  const firstNonWhitespaceIndex = find.search(/\S/);
+  const firstNonWhitespace = firstNonWhitespaceIndex >= 0 ? find[firstNonWhitespaceIndex] : void 0;
+  const lastNonWhitespaceMatch = /\S(?=\s*$)/.exec(find);
+  const lastNonWhitespace = lastNonWhitespaceMatch?.[0];
+  const startsAtTextBoundary = firstNonWhitespaceIndex === 0 && !!firstNonWhitespace && /\w/.test(firstNonWhitespace);
+  const endsAtTextBoundary = !!lastNonWhitespace && find.trimEnd().length === find.length && /\w/.test(lastNonWhitespace);
+  const prefix = startsAtTextBoundary ? "(?<!\\w)" : "";
+  const suffix = endsAtTextBoundary ? "(?!\\w)" : "";
+  const flags = caseSensitive === false ? "gi" : "g";
+  return new RegExp(`${prefix}${pattern}${suffix}`, flags);
+}
 function buildReplacementRegex(words) {
   const filtered = words.filter((w) => w.length > 0);
   if (filtered.length === 0)
     return null;
   filtered.sort((a, b) => b.length - a.length);
-  const escaped = filtered.map((w) => w.replace(/[.*+?^${}()|[\]\\]/g, "\\$&"));
+  const escaped = filtered.map(escapeRegex);
   const pattern = escaped.join("|");
   return new RegExp(`\\b(?:${pattern})\\b`, "gi");
 }
@@ -4185,12 +4238,10 @@ function applyReplacements(rawVtt, replacements) {
     return { editedVtt: rawVtt, replacements: [] };
   }
   const records = [];
-  const editedVtt = transformCueText(rawVtt, (line, cueStartTime) => {
-    let result = line;
+  const editedVtt = transformCueTextBlocks(rawVtt, (cueText, cueStartTime) => {
+    let result = cueText;
     for (const { find, replace, caseSensitive } of filtered) {
-      const escaped = find.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
-      const flags = caseSensitive === false ? "gi" : "g";
-      const regex = new RegExp(`\\b${escaped}\\b`, flags);
+      const regex = buildStaticReplacementRegex(find, caseSensitive);
       result = result.replace(regex, (match) => {
         records.push({ cueStartTime, before: match, after: replace });
         return replace;