@storyteller-platform/ghost-story 0.1.4 → 0.1.6

package/dist/{AudioFormat--LEggByS.d.ts → AudioFormat-BJuNZJlT.d.ts}

@@ -93,6 +93,10 @@ interface WhisperServerOptions {
     temperature?: number;
     apiKey?: string;
     inputFormat?: AudioFormat;
+    /**
+     * timeout in milliseconds waiting for a response from the whisper server
+     */
+    timeout?: number;
 }
 interface RecognitionResult$1 {
     transcript: string;

package/dist/{AudioFormat-Cq6sRKDj.d.cts → AudioFormat-CKxAQAj2.d.cts}

@@ -93,6 +93,10 @@ interface WhisperServerOptions {
     temperature?: number;
     apiKey?: string;
     inputFormat?: AudioFormat;
+    /**
+     * timeout in milliseconds waiting for a response from the whisper server
+     */
+    timeout?: number;
 }
 interface RecognitionResult$1 {
     transcript: string;

package/dist/api/APIOptions.d.cts

@@ -1,4 +1,4 @@
-import { a as RecognitionOptions } from '../AudioFormat-Cq6sRKDj.cjs';
+import { a as RecognitionOptions } from '../AudioFormat-CKxAQAj2.cjs';
 import 'node:fs';
 import 'node:stream';
 import '../config.cjs';

package/dist/api/APIOptions.d.ts

@@ -1,4 +1,4 @@
-import { a as RecognitionOptions } from '../AudioFormat--LEggByS.js';
+import { a as RecognitionOptions } from '../AudioFormat-BJuNZJlT.js';
 import 'node:fs';
 import 'node:stream';
 import '../config.js';

package/dist/api/Recognition.d.cts

@@ -1,6 +1,6 @@
 import 'node:fs';
 import 'node:stream';
-export { Y as Audio, A as AudioInput, R as RecognitionEngine, a as RecognitionOptions, b as RecognitionResult, r as recognitionEngines, c as recognize } from '../AudioFormat-Cq6sRKDj.cjs';
+export { Y as Audio, A as AudioInput, R as RecognitionEngine, a as RecognitionOptions, b as RecognitionResult, r as recognitionEngines, c as recognize } from '../AudioFormat-CKxAQAj2.cjs';
 import '../config.cjs';
 import '../utilities/Timeline.cjs';
 import '../utilities/Timing.cjs';

package/dist/api/Recognition.d.ts

@@ -1,6 +1,6 @@
 import 'node:fs';
 import 'node:stream';
-export { Y as Audio, A as AudioInput, R as RecognitionEngine, a as RecognitionOptions, b as RecognitionResult, r as recognitionEngines, c as recognize } from '../AudioFormat--LEggByS.js';
+export { Y as Audio, A as AudioInput, R as RecognitionEngine, a as RecognitionOptions, b as RecognitionResult, r as recognitionEngines, c as recognize } from '../AudioFormat-BJuNZJlT.js';
 import '../config.js';
 import '../utilities/Timeline.js';
 import '../utilities/Timing.js';

package/dist/audio/AudioConverter.d.cts

@@ -1,4 +1,4 @@
-import { d as AudioFormat, f as AudioSource, R as RecognitionEngine } from '../AudioFormat-Cq6sRKDj.cjs';
+import { d as AudioFormat, f as AudioSource, R as RecognitionEngine } from '../AudioFormat-CKxAQAj2.cjs';
 import { Readable } from 'node:stream';
 import { ConversionMode } from '../config.cjs';
 import 'node:fs';

package/dist/audio/AudioConverter.d.ts

@@ -1,4 +1,4 @@
-import { d as AudioFormat, f as AudioSource, R as RecognitionEngine } from '../AudioFormat--LEggByS.js';
+import { d as AudioFormat, f as AudioSource, R as RecognitionEngine } from '../AudioFormat-BJuNZJlT.js';
 import { Readable } from 'node:stream';
 import { ConversionMode } from '../config.js';
 import 'node:fs';

package/dist/audio/AudioFormat.d.cts

@@ -1,4 +1,4 @@
-export { P as AudioEncoding, d as AudioFormat, e as AudioFormatInfo, S as ServiceCapabilities, n as formatFromExtension, o as formatToExtension, s as getTargetFormat, u as needsConversion, w as serviceCapabilities } from '../AudioFormat-Cq6sRKDj.cjs';
+export { P as AudioEncoding, d as AudioFormat, e as AudioFormatInfo, S as ServiceCapabilities, n as formatFromExtension, o as formatToExtension, s as getTargetFormat, u as needsConversion, w as serviceCapabilities } from '../AudioFormat-CKxAQAj2.cjs';
 import 'node:fs';
 import 'node:stream';
 import '../config.cjs';

package/dist/audio/AudioFormat.d.ts

@@ -1,4 +1,4 @@
-export { P as AudioEncoding, d as AudioFormat, e as AudioFormatInfo, S as ServiceCapabilities, n as formatFromExtension, o as formatToExtension, s as getTargetFormat, u as needsConversion, w as serviceCapabilities } from '../AudioFormat--LEggByS.js';
+export { P as AudioEncoding, d as AudioFormat, e as AudioFormatInfo, S as ServiceCapabilities, n as formatFromExtension, o as formatToExtension, s as getTargetFormat, u as needsConversion, w as serviceCapabilities } from '../AudioFormat-BJuNZJlT.js';
 import 'node:fs';
 import 'node:stream';
 import '../config.js';

package/dist/audio/AudioSource.d.cts

@@ -1,6 +1,6 @@
 import 'node:fs';
 import 'node:stream';
-export { f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, j as RawAudioInput, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, p as getFormat, q as getFormatInfo, t as isAudioSource, v as normalizeToAudioSource, x as toBuffer, y as toFilePath, z as toReadStream } from '../AudioFormat-Cq6sRKDj.cjs';
+export { f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, j as RawAudioInput, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, p as getFormat, q as getFormatInfo, t as isAudioSource, v as normalizeToAudioSource, x as toBuffer, y as toFilePath, z as toReadStream } from '../AudioFormat-CKxAQAj2.cjs';
 import '../config.cjs';
 import '../utilities/Timeline.cjs';
 import '../utilities/Timing.cjs';

package/dist/audio/AudioSource.d.ts

@@ -1,6 +1,6 @@
 import 'node:fs';
 import 'node:stream';
-export { f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, j as RawAudioInput, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, p as getFormat, q as getFormatInfo, t as isAudioSource, v as normalizeToAudioSource, x as toBuffer, y as toFilePath, z as toReadStream } from '../AudioFormat--LEggByS.js';
+export { f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, j as RawAudioInput, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, p as getFormat, q as getFormatInfo, t as isAudioSource, v as normalizeToAudioSource, x as toBuffer, y as toFilePath, z as toReadStream } from '../AudioFormat-BJuNZJlT.js';
 import '../config.js';
 import '../utilities/Timeline.js';
 import '../utilities/Timing.js';

package/dist/audio/index.d.cts

@@ -1,4 +1,4 @@
-export { P as AudioEncoding, d as AudioFormat, e as AudioFormatInfo, f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, j as RawAudioInput, S as ServiceCapabilities, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, n as formatFromExtension, o as formatToExtension, p as getFormat, q as getFormatInfo, s as getTargetFormat, t as isAudioSource, u as needsConversion, v as normalizeToAudioSource, w as serviceCapabilities, x as toBuffer, y as toFilePath, z as toReadStream } from '../AudioFormat-Cq6sRKDj.cjs';
+export { P as AudioEncoding, d as AudioFormat, e as AudioFormatInfo, f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, j as RawAudioInput, S as ServiceCapabilities, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, n as formatFromExtension, o as formatToExtension, p as getFormat, q as getFormatInfo, s as getTargetFormat, t as isAudioSource, u as needsConversion, v as normalizeToAudioSource, w as serviceCapabilities, x as toBuffer, y as toFilePath, z as toReadStream } from '../AudioFormat-CKxAQAj2.cjs';
 export { ConversionOptions, PrepareForServiceOptions, PreparedAudio, StreamForUploadOptions, StreamForUploadResult, convertToBuffer, convertToFile, convertToStream, createStreamForUpload, createStreamingConversion, getAudioDuration, prepareForService, prepareWavForService } from './AudioConverter.cjs';
 import 'node:fs';
 import 'node:stream';

package/dist/audio/index.d.ts

@@ -1,4 +1,4 @@
-export { P as AudioEncoding, d as AudioFormat, e as AudioFormatInfo, f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, j as RawAudioInput, S as ServiceCapabilities, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, n as formatFromExtension, o as formatToExtension, p as getFormat, q as getFormatInfo, s as getTargetFormat, t as isAudioSource, u as needsConversion, v as normalizeToAudioSource, w as serviceCapabilities, x as toBuffer, y as toFilePath, z as toReadStream } from '../AudioFormat--LEggByS.js';
+export { P as AudioEncoding, d as AudioFormat, e as AudioFormatInfo, f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, j as RawAudioInput, S as ServiceCapabilities, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, n as formatFromExtension, o as formatToExtension, p as getFormat, q as getFormatInfo, s as getTargetFormat, t as isAudioSource, u as needsConversion, v as normalizeToAudioSource, w as serviceCapabilities, x as toBuffer, y as toFilePath, z as toReadStream } from '../AudioFormat-BJuNZJlT.js';
 export { ConversionOptions, PrepareForServiceOptions, PreparedAudio, StreamForUploadOptions, StreamForUploadResult, convertToBuffer, convertToFile, convertToStream, createStreamForUpload, createStreamingConversion, getAudioDuration, prepareForService, prepareWavForService } from './AudioConverter.js';
 import 'node:fs';
 import 'node:stream';

package/dist/cli/bin.cjs

@@ -24,6 +24,7 @@ var __toESM = (mod, isNodeMode, target) => (target = mod != null ? __create(__ge
 ));
 var import_zli = require("@robingenz/zli");
 var import_cli_progress = require("cli-progress");
+var import_fs_extra = require("fs-extra");
 var import_zod = require("zod");
 var import_config = require("./config.cjs");
 var import_install = require("./install.cjs");
@@ -194,7 +195,16 @@ const transcribeCommand = (0, import_zli.defineCommand)({
       import_cli_progress.Presets.shades_classic
     );
     try {
-      const [inputPath] = args;
+      let [inputPath] = args;
+      const path = await import("node:path");
+      inputPath = path.resolve(process.cwd(), inputPath);
+      let outputPath = options.output ? path.resolve(process.cwd(), options.output) : void 0;
+      if (outputPath && !path.extname(outputPath)) {
+        outputPath = `${outputPath}.json`;
+      }
+      if (outputPath) {
+        (0, import_fs_extra.ensureDirSync)(path.dirname(outputPath));
+      }
       const result = await recognize(inputPath, {
         engine: options.engine,
         language: options.language,
@@ -219,14 +229,14 @@ const transcribeCommand = (0, import_zli.defineCommand)({
           }
         }
       });
-      if (!options.output) {
+      if (!outputPath) {
         console.log(JSON.stringify(result, null, 2));
         return;
       }
       const { writeFile } = await import("node:fs/promises");
-      await writeFile(options.output, JSON.stringify(result, null, 2));
+      await writeFile(outputPath, JSON.stringify(result, null, 2));
       console.log(`
-Transcription written to ${options.output}`);
+Transcription written to ${outputPath}`);
     } finally {
       if (!options.noProgress) {
         progressBar.stop();

package/dist/cli/bin.js

@@ -6,6 +6,7 @@ import {
   processConfig
 } from "@robingenz/zli";
 import { Presets, SingleBar } from "cli-progress";
+import { ensureDirSync } from "fs-extra";
 import { z } from "zod";
 import {
   BUILD_VARIANTS,
@@ -193,7 +194,16 @@ const transcribeCommand = defineCommand({
       Presets.shades_classic
     );
     try {
-      const [inputPath] = args;
+      let [inputPath] = args;
+      const path = await import("node:path");
+      inputPath = path.resolve(process.cwd(), inputPath);
+      let outputPath = options.output ? path.resolve(process.cwd(), options.output) : void 0;
+      if (outputPath && !path.extname(outputPath)) {
+        outputPath = `${outputPath}.json`;
+      }
+      if (outputPath) {
+        ensureDirSync(path.dirname(outputPath));
+      }
       const result = await recognize(inputPath, {
         engine: options.engine,
         language: options.language,
@@ -218,14 +228,14 @@ const transcribeCommand = defineCommand({
           }
         }
       });
-      if (!options.output) {
+      if (!outputPath) {
         console.log(JSON.stringify(result, null, 2));
         return;
       }
       const { writeFile } = await import("node:fs/promises");
-      await writeFile(options.output, JSON.stringify(result, null, 2));
+      await writeFile(outputPath, JSON.stringify(result, null, 2));
       console.log(`
-Transcription written to ${options.output}`);
+Transcription written to ${outputPath}`);
     } finally {
       if (!options.noProgress) {
         progressBar.stop();

package/dist/cli/install.cjs

@@ -43,6 +43,8 @@ var import_node_os = __toESM(require("node:os"), 1);
 var import_node_path = __toESM(require("node:path"), 1);
 var import_promises2 = require("node:stream/promises");
 var import_node_zlib = require("node:zlib");
+var import_chalk = __toESM(require("chalk"), 1);
+var import_cli_progress = require("cli-progress");
 var import_tar = require("tar");
 var import_config = require("./config.cjs");
 async function downloadFile(options) {
@@ -68,9 +70,27 @@ async function downloadFile(options) {
   }
   await (0, import_promises.mkdir)(import_node_path.default.dirname(destPath), { recursive: true });
   const fileStream = (0, import_node_fs.createWriteStream)(destPath);
+  const progressBar = new import_cli_progress.SingleBar(
+    {
+      etaBuffer: 2,
+      hideCursor: null,
+      noTTYOutput: !process.stderr.isTTY,
+      autopadding: true,
+      format: `${import_chalk.default.yellow("{bar}")} | {percentage}% | {current}/{sum} MB | ({speed} MB/s)`
+    },
+    import_cli_progress.Presets.shades_classic
+  );
   const reader = response.body.getReader();
   let downloaded = 0;
-  let lastPrinted = 0;
+  let lastTime = Date.now();
+  let lastDownloaded = 0;
+  if (printOutput) {
+    progressBar.start(totalSize, 0, {
+      current: "0 MB",
+      sum: `${(totalSize / 1024 / 1024).toFixed(2)} MB`,
+      speed: 0
+    });
+  }
   try {
     while (true) {
       const { done, value } = await reader.read();
@@ -81,18 +101,25 @@ async function downloadFile(options) {
       fileStream.write(value);
       onProgress == null ? void 0 : onProgress(downloaded, totalSize);
       if (printOutput && totalSize > 0) {
-        const shouldPrint = downloaded - lastPrinted > 10 * 1024 * 1024 || downloaded === totalSize;
+        const deltaTime = Date.now() - lastTime;
+        const shouldPrint = deltaTime > 1e3 || downloaded === totalSize;
         if (shouldPrint) {
-          const percent = (downloaded / totalSize * 100).toFixed(1);
-          console.log(
-            `  ${(downloaded / 1024 / 1024).toFixed(1)} MB / ${(totalSize / 1024 / 1024).toFixed(1)} MB (${percent}%)`
-          );
-          lastPrinted = downloaded;
+          const speed = (downloaded - lastDownloaded) / 1024 / 1024 / (deltaTime / 1e3);
+          lastTime = Date.now();
+          lastDownloaded = downloaded;
+          progressBar.update(downloaded, {
+            current: (downloaded / 1024 / 1024).toFixed(2),
+            sum: (totalSize / 1024 / 1024).toFixed(2),
+            speed: speed.toFixed(2)
+          });
         }
       }
     }
   } finally {
     fileStream.end();
+    if (printOutput) {
+      progressBar.stop();
+    }
   }
   await new Promise((resolve, reject) => {
     fileStream.on("finish", resolve);

package/dist/cli/install.js

@@ -9,6 +9,8 @@ import os from "node:os";
 import path from "node:path";
 import { pipeline } from "node:stream/promises";
 import { createGunzip } from "node:zlib";
+import chalk from "chalk";
+import { Presets, SingleBar } from "cli-progress";
 import { extract } from "tar";
 import {
   MODEL_SIZES,
@@ -52,9 +54,27 @@ async function downloadFile(options) {
   }
   await mkdir(path.dirname(destPath), { recursive: true });
   const fileStream = createWriteStream(destPath);
+  const progressBar = new SingleBar(
+    {
+      etaBuffer: 2,
+      hideCursor: null,
+      noTTYOutput: !process.stderr.isTTY,
+      autopadding: true,
+      format: `${chalk.yellow("{bar}")} | {percentage}% | {current}/{sum} MB | ({speed} MB/s)`
+    },
+    Presets.shades_classic
+  );
   const reader = response.body.getReader();
   let downloaded = 0;
-  let lastPrinted = 0;
+  let lastTime = Date.now();
+  let lastDownloaded = 0;
+  if (printOutput) {
+    progressBar.start(totalSize, 0, {
+      current: "0 MB",
+      sum: `${(totalSize / 1024 / 1024).toFixed(2)} MB`,
+      speed: 0
+    });
+  }
   try {
     while (true) {
       const { done, value } = await reader.read();
@@ -65,18 +85,25 @@ async function downloadFile(options) {
       fileStream.write(value);
       onProgress == null ? void 0 : onProgress(downloaded, totalSize);
       if (printOutput && totalSize > 0) {
-        const shouldPrint = downloaded - lastPrinted > 10 * 1024 * 1024 || downloaded === totalSize;
+        const deltaTime = Date.now() - lastTime;
+        const shouldPrint = deltaTime > 1e3 || downloaded === totalSize;
         if (shouldPrint) {
-          const percent = (downloaded / totalSize * 100).toFixed(1);
-          console.log(
-            `  ${(downloaded / 1024 / 1024).toFixed(1)} MB / ${(totalSize / 1024 / 1024).toFixed(1)} MB (${percent}%)`
-          );
-          lastPrinted = downloaded;
+          const speed = (downloaded - lastDownloaded) / 1024 / 1024 / (deltaTime / 1e3);
+          lastTime = Date.now();
+          lastDownloaded = downloaded;
+          progressBar.update(downloaded, {
+            current: (downloaded / 1024 / 1024).toFixed(2),
+            sum: (totalSize / 1024 / 1024).toFixed(2),
+            speed: speed.toFixed(2)
+          });
         }
       }
     }
   } finally {
     fileStream.end();
+    if (printOutput) {
+      progressBar.stop();
+    }
   }
   await new Promise((resolve, reject) => {
     fileStream.on("finish", resolve);

package/dist/convert.d.cts

@@ -1,4 +1,4 @@
-import { d as AudioFormat } from './AudioFormat-Cq6sRKDj.cjs';
+import { d as AudioFormat } from './AudioFormat-CKxAQAj2.cjs';
 import 'node:fs';
 import 'node:stream';
 import './config.cjs';

package/dist/convert.d.ts

@@ -1,4 +1,4 @@
-import { d as AudioFormat } from './AudioFormat--LEggByS.js';
+import { d as AudioFormat } from './AudioFormat-BJuNZJlT.js';
 import 'node:fs';
 import 'node:stream';
 import './config.js';

package/dist/fetch.cjs

@@ -0,0 +1,35 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var fetch_exports = {};
+__export(fetch_exports, {
+  createTimeoutAgent: () => createTimeoutAgent
+});
+module.exports = __toCommonJS(fetch_exports);
+var import_undici = require("undici");
+const createTimeoutAgent = (timeout) => new import_undici.Agent({
+  // 1 hour timeout lmao
+  headersTimeout: timeout,
+  bodyTimeout: timeout,
+  // connection shouldnt be that long
+  connectTimeout: 3e4
+});
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  createTimeoutAgent
+});

package/dist/fetch.d.cts

@@ -0,0 +1,9 @@
+import { Agent } from 'undici';
+
+/**
+ * necessary bc there's a default timeout of 300 seconds for all `fetch` requests, which is
+ * often shorter than the time it takes for the whisper server to respond
+ */
+declare const createTimeoutAgent: (timeout: number) => Agent;
+
+export { createTimeoutAgent };

package/dist/fetch.d.ts

@@ -0,0 +1,9 @@
+import { Agent } from 'undici';
+
+/**
+ * necessary bc there's a default timeout of 300 seconds for all `fetch` requests, which is
+ * often shorter than the time it takes for the whisper server to respond
+ */
+declare const createTimeoutAgent: (timeout: number) => Agent;
+
+export { createTimeoutAgent };

package/dist/fetch.js

@@ -0,0 +1,11 @@
+import { Agent } from "undici";
+const createTimeoutAgent = (timeout) => new Agent({
+  // 1 hour timeout lmao
+  headersTimeout: timeout,
+  bodyTimeout: timeout,
+  // connection shouldnt be that long
+  connectTimeout: 3e4
+});
+export {
+  createTimeoutAgent
+};

package/dist/index.d.cts

@@ -1,4 +1,4 @@
-export { d as AudioFormat, e as AudioFormatInfo, A as AudioInput, f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, L as Language, E as Languages, O as OpenAICloudSTTOptions, B as OpenAIResult, j as RawAudioInput, R as RecognitionEngine, a as RecognitionOptions, b as RecognitionResult, S as ServiceCapabilities, W as WhisperCppModelId, G as WhisperCppOptions, F as WhisperCppResult, K as WhisperServerOptions, J as WhisperServerResult, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, n as formatFromExtension, o as formatToExtension, p as getFormat, q as getFormatInfo, s as getTargetFormat, t as isAudioSource, u as needsConversion, v as normalizeToAudioSource, C as openaiInputPreference, r as recognitionEngines, c as recognize, D as recognizeOpenAI, I as recognizeWhisperCpp, N as recognizeWhisperServer, w as serviceCapabilities, x as toBuffer, y as toFilePath, z as toReadStream, H as whisperCppInputPreference, M as whisperServerInputPreference } from './AudioFormat-Cq6sRKDj.cjs';
+export { d as AudioFormat, e as AudioFormatInfo, A as AudioInput, f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, L as Language, E as Languages, O as OpenAICloudSTTOptions, B as OpenAIResult, j as RawAudioInput, R as RecognitionEngine, a as RecognitionOptions, b as RecognitionResult, S as ServiceCapabilities, W as WhisperCppModelId, G as WhisperCppOptions, F as WhisperCppResult, K as WhisperServerOptions, J as WhisperServerResult, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, n as formatFromExtension, o as formatToExtension, p as getFormat, q as getFormatInfo, s as getTargetFormat, t as isAudioSource, u as needsConversion, v as normalizeToAudioSource, C as openaiInputPreference, r as recognitionEngines, c as recognize, D as recognizeOpenAI, I as recognizeWhisperCpp, N as recognizeWhisperServer, w as serviceCapabilities, x as toBuffer, y as toFilePath, z as toReadStream, H as whisperCppInputPreference, M as whisperServerInputPreference } from './AudioFormat-CKxAQAj2.cjs';
 export { ConversionOptions, PreparedAudio, convertToBuffer, convertToFile, createStreamingConversion, prepareForService, prepareWavForService } from './audio/AudioConverter.cjs';
 export { BUILD_VARIANTS, BuildVariant, RECOGNITION_ENGINES, WHISPER_CPP_VERSION, WHISPER_MODELS, WhisperModel, applyLegacyCpuFallback, detectPlatform, getConfiguredVariant, getInstallDir, getModelDir, getModelPath, getVadModelPath, getWhisperExecutablePath, getWhisperServerExecutablePath, isValidModel, resolveVariant } from './cli/config.cjs';
 export { InstallBinaryOptions, InstallModelOptions, InstallVadModelOptions, ensureWhisperInstalled, installBinary, installModel, installVadModel } from './cli/install.cjs';

package/dist/index.d.ts

@@ -1,4 +1,4 @@
-export { d as AudioFormat, e as AudioFormatInfo, A as AudioInput, f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, L as Language, E as Languages, O as OpenAICloudSTTOptions, B as OpenAIResult, j as RawAudioInput, R as RecognitionEngine, a as RecognitionOptions, b as RecognitionResult, S as ServiceCapabilities, W as WhisperCppModelId, G as WhisperCppOptions, F as WhisperCppResult, K as WhisperServerOptions, J as WhisperServerResult, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, n as formatFromExtension, o as formatToExtension, p as getFormat, q as getFormatInfo, s as getTargetFormat, t as isAudioSource, u as needsConversion, v as normalizeToAudioSource, C as openaiInputPreference, r as recognitionEngines, c as recognize, D as recognizeOpenAI, I as recognizeWhisperCpp, N as recognizeWhisperServer, w as serviceCapabilities, x as toBuffer, y as toFilePath, z as toReadStream, H as whisperCppInputPreference, M as whisperServerInputPreference } from './AudioFormat--LEggByS.js';
+export { d as AudioFormat, e as AudioFormatInfo, A as AudioInput, f as AudioSource, g as AudioSourceFromBuffer, h as AudioSourceFromFile, i as AudioSourceFromStream, L as Language, E as Languages, O as OpenAICloudSTTOptions, B as OpenAIResult, j as RawAudioInput, R as RecognitionEngine, a as RecognitionOptions, b as RecognitionResult, S as ServiceCapabilities, W as WhisperCppModelId, G as WhisperCppOptions, F as WhisperCppResult, K as WhisperServerOptions, J as WhisperServerResult, k as audioSourceFromBuffer, l as audioSourceFromFile, m as audioSourceFromStream, n as formatFromExtension, o as formatToExtension, p as getFormat, q as getFormatInfo, s as getTargetFormat, t as isAudioSource, u as needsConversion, v as normalizeToAudioSource, C as openaiInputPreference, r as recognitionEngines, c as recognize, D as recognizeOpenAI, I as recognizeWhisperCpp, N as recognizeWhisperServer, w as serviceCapabilities, x as toBuffer, y as toFilePath, z as toReadStream, H as whisperCppInputPreference, M as whisperServerInputPreference } from './AudioFormat-BJuNZJlT.js';
 export { ConversionOptions, PreparedAudio, convertToBuffer, convertToFile, createStreamingConversion, prepareForService, prepareWavForService } from './audio/AudioConverter.js';
 export { BUILD_VARIANTS, BuildVariant, RECOGNITION_ENGINES, WHISPER_CPP_VERSION, WHISPER_MODELS, WhisperModel, applyLegacyCpuFallback, detectPlatform, getConfiguredVariant, getInstallDir, getModelDir, getModelPath, getVadModelPath, getWhisperExecutablePath, getWhisperServerExecutablePath, isValidModel, resolveVariant } from './cli/config.js';
 export { InstallBinaryOptions, InstallModelOptions, InstallVadModelOptions, ensureWhisperInstalled, installBinary, installModel, installVadModel } from './cli/install.js';

package/dist/recognition/AmazonTranscribeSTT.d.cts

@@ -1,4 +1,4 @@
-import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat-Cq6sRKDj.cjs';
+import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat-CKxAQAj2.cjs';
 import { Timeline } from '../utilities/Timeline.cjs';
 import { Timing } from '../utilities/Timing.cjs';
 import 'node:fs';

package/dist/recognition/AmazonTranscribeSTT.d.ts

@@ -1,4 +1,4 @@
-import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat--LEggByS.js';
+import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat-BJuNZJlT.js';
 import { Timeline } from '../utilities/Timeline.js';
 import { Timing } from '../utilities/Timing.js';
 import 'node:fs';

package/dist/recognition/AzureCognitiveServicesSTT.d.cts

@@ -1,5 +1,5 @@
 import * as SpeechSDK from 'microsoft-cognitiveservices-speech-sdk';
-import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat-Cq6sRKDj.cjs';
+import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat-CKxAQAj2.cjs';
 import { Timeline } from '../utilities/Timeline.cjs';
 import { Timing } from '../utilities/Timing.cjs';
 import 'node:fs';

package/dist/recognition/AzureCognitiveServicesSTT.d.ts

@@ -1,5 +1,5 @@
 import * as SpeechSDK from 'microsoft-cognitiveservices-speech-sdk';
-import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat--LEggByS.js';
+import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat-BJuNZJlT.js';
 import { Timeline } from '../utilities/Timeline.js';
 import { Timing } from '../utilities/Timing.js';
 import 'node:fs';

package/dist/recognition/DeepgramSTT.cjs

@@ -24,6 +24,7 @@ __export(DeepgramSTT_exports, {
 module.exports = __toCommonJS(DeepgramSTT_exports);
 var import_audio = require("../audio/index.cjs");
 var import_config = require("../config.cjs");
+var import_fetch = require("../fetch.cjs");
 var import_ObjectUtilities = require("../utilities/ObjectUtilities.cjs");
 const SERVICE_ID = "deepgram";
 function formatToContentType(format) {
@@ -93,7 +94,8 @@ async function recognize(input, languageCode, timing, options, signal) {
         // necessary because otherwise there's some excruciating type error when running `yarn check`
         // from the root
         body: uploadResult.stream,
-        signal: signal ?? null
+        signal: signal ?? null,
+        dispatcher: (0, import_fetch.createTimeoutAgent)(opts.timeout)
       });
       const conversionPromise = (_a2 = uploadResult.start) == null ? void 0 : _a2.call(uploadResult);
       const resp = await fetchPromise;
@@ -187,7 +189,11 @@ function applyPunctuationToTimeline(timeline, transcript) {
 const defaultDeepgramSTTOptions = {
   apiKey: "",
   model: "nova-3",
-  punctuate: true
+  punctuate: true,
+  timeout: 30 * 60 * 1e3,
+  // 30 minutes
+  inputFormat: void 0,
+  conversionMode: void 0
 };
 // Annotate the CommonJS export names for ESM import in node:
 0 && (module.exports = {

package/dist/recognition/DeepgramSTT.d.cts

@@ -1,4 +1,4 @@
-import { j as RawAudioInput, f as AudioSource, d as AudioFormat } from '../AudioFormat-Cq6sRKDj.cjs';
+import { j as RawAudioInput, f as AudioSource, d as AudioFormat } from '../AudioFormat-CKxAQAj2.cjs';
 import { ConversionMode } from '../config.cjs';
 import { TimelineEntry } from '../utilities/Timeline.cjs';
 import { Timing } from '../utilities/Timing.cjs';
@@ -17,7 +17,15 @@ interface DeepgramSTTOptions {
     punctuate: boolean;
     inputFormat?: AudioFormat | undefined;
     conversionMode?: ConversionMode | undefined;
+    timeout?: number;
 }
-declare const defaultDeepgramSTTOptions: DeepgramSTTOptions;
+declare const defaultDeepgramSTTOptions: {
+    apiKey: string;
+    model: string;
+    punctuate: boolean;
+    timeout: number;
+    inputFormat: AudioFormat | undefined;
+    conversionMode: ConversionMode | undefined;
+};
 
 export { type DeepgramSTTOptions, defaultDeepgramSTTOptions, recognize };

package/dist/recognition/DeepgramSTT.d.ts

@@ -1,4 +1,4 @@
-import { j as RawAudioInput, f as AudioSource, d as AudioFormat } from '../AudioFormat--LEggByS.js';
+import { j as RawAudioInput, f as AudioSource, d as AudioFormat } from '../AudioFormat-BJuNZJlT.js';
 import { ConversionMode } from '../config.js';
 import { TimelineEntry } from '../utilities/Timeline.js';
 import { Timing } from '../utilities/Timing.js';
@@ -17,7 +17,15 @@ interface DeepgramSTTOptions {
     punctuate: boolean;
     inputFormat?: AudioFormat | undefined;
     conversionMode?: ConversionMode | undefined;
+    timeout?: number;
 }
-declare const defaultDeepgramSTTOptions: DeepgramSTTOptions;
+declare const defaultDeepgramSTTOptions: {
+    apiKey: string;
+    model: string;
+    punctuate: boolean;
+    timeout: number;
+    inputFormat: AudioFormat | undefined;
+    conversionMode: ConversionMode | undefined;
+};
 
 export { type DeepgramSTTOptions, defaultDeepgramSTTOptions, recognize };

package/dist/recognition/DeepgramSTT.js

@@ -6,6 +6,7 @@ import {
   serviceCapabilities
 } from "../audio/index.js";
 import { getConversionMode } from "../config.js";
+import { createTimeoutAgent } from "../fetch.js";
 import { extendDeep } from "../utilities/ObjectUtilities.js";
 const SERVICE_ID = "deepgram";
 function formatToContentType(format) {
@@ -75,7 +76,8 @@ async function recognize(input, languageCode, timing, options, signal) {
         // necessary because otherwise there's some excruciating type error when running `yarn check`
         // from the root
         body: uploadResult.stream,
-        signal: signal ?? null
+        signal: signal ?? null,
+        dispatcher: createTimeoutAgent(opts.timeout)
       });
       const conversionPromise = (_a2 = uploadResult.start) == null ? void 0 : _a2.call(uploadResult);
       const resp = await fetchPromise;
@@ -169,7 +171,11 @@ function applyPunctuationToTimeline(timeline, transcript) {
 const defaultDeepgramSTTOptions = {
   apiKey: "",
   model: "nova-3",
-  punctuate: true
+  punctuate: true,
+  timeout: 30 * 60 * 1e3,
+  // 30 minutes
+  inputFormat: void 0,
+  conversionMode: void 0
 };
 export {
   defaultDeepgramSTTOptions,

package/dist/recognition/GoogleCloudSTT.d.cts

@@ -1,4 +1,4 @@
-import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat-Cq6sRKDj.cjs';
+import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat-CKxAQAj2.cjs';
 import { Timeline } from '../utilities/Timeline.cjs';
 import { Timing } from '../utilities/Timing.cjs';
 import 'node:fs';

package/dist/recognition/GoogleCloudSTT.d.ts

@@ -1,4 +1,4 @@
-import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat--LEggByS.js';
+import { d as AudioFormat, j as RawAudioInput, f as AudioSource } from '../AudioFormat-BJuNZJlT.js';
 import { Timeline } from '../utilities/Timeline.js';
 import { Timing } from '../utilities/Timing.js';
 import 'node:fs';

package/dist/recognition/OpenAICloudSTT.d.cts

@@ -1,4 +1,4 @@
-export { Q as InputPreference, O as OpenAICloudSTTOptions, B as RecognitionResult, C as inputPreference, D as recognize } from '../AudioFormat-Cq6sRKDj.cjs';
+export { Q as InputPreference, O as OpenAICloudSTTOptions, B as RecognitionResult, C as inputPreference, D as recognize } from '../AudioFormat-CKxAQAj2.cjs';
 import '../utilities/Timeline.cjs';
 import '../utilities/Timing.cjs';
 import 'node:fs';

package/dist/recognition/OpenAICloudSTT.d.ts

@@ -1,4 +1,4 @@
-export { Q as InputPreference, O as OpenAICloudSTTOptions, B as RecognitionResult, C as inputPreference, D as recognize } from '../AudioFormat--LEggByS.js';
+export { Q as InputPreference, O as OpenAICloudSTTOptions, B as RecognitionResult, C as inputPreference, D as recognize } from '../AudioFormat-BJuNZJlT.js';
 import '../utilities/Timeline.js';
 import '../utilities/Timing.js';
 import 'node:fs';

package/dist/recognition/WhisperCppSTT.cjs

@@ -319,7 +319,7 @@ function buildTranscribeArgs(options) {
     "--print-progress",
     options.language ? ["--language", options.language.toLowerCase()] : null,
     options.flashAttention ? ["--flash-attn"] : null,
-    options.suppressNonSpeechTokens ? ["--suppress-nst", "--no-prints"] : null,
+    options.suppressNonSpeechTokens ? ["--suppress-nst"] : null,
     ["--processors", String(options.processors)],
     ["--threads", String(options.threads)]
   ];

package/dist/recognition/WhisperCppSTT.d.cts

@@ -1,4 +1,4 @@
-export { T as InputPreference, L as Language, E as Languages, F as RecognitionResult, W as WhisperCppModelId, G as WhisperCppOptions, V as ensureModelDownloaded, U as ensureWhisperCppInstalled, H as inputPreference, I as recognize } from '../AudioFormat-Cq6sRKDj.cjs';
+export { T as InputPreference, L as Language, E as Languages, F as RecognitionResult, W as WhisperCppModelId, G as WhisperCppOptions, V as ensureModelDownloaded, U as ensureWhisperCppInstalled, H as inputPreference, I as recognize } from '../AudioFormat-CKxAQAj2.cjs';
 import '../utilities/Timeline.cjs';
 import '../utilities/Timing.cjs';
 export { WhisperModel } from '../cli/config.cjs';

package/dist/recognition/WhisperCppSTT.d.ts

@@ -1,4 +1,4 @@
-export { T as InputPreference, L as Language, E as Languages, F as RecognitionResult, W as WhisperCppModelId, G as WhisperCppOptions, V as ensureModelDownloaded, U as ensureWhisperCppInstalled, H as inputPreference, I as recognize } from '../AudioFormat--LEggByS.js';
+export { T as InputPreference, L as Language, E as Languages, F as RecognitionResult, W as WhisperCppModelId, G as WhisperCppOptions, V as ensureModelDownloaded, U as ensureWhisperCppInstalled, H as inputPreference, I as recognize } from '../AudioFormat-BJuNZJlT.js';
 import '../utilities/Timeline.js';
 import '../utilities/Timing.js';
 export { WhisperModel } from '../cli/config.js';

package/dist/recognition/WhisperCppSTT.js

@@ -298,7 +298,7 @@ function buildTranscribeArgs(options) {
     "--print-progress",
     options.language ? ["--language", options.language.toLowerCase()] : null,
     options.flashAttention ? ["--flash-attn"] : null,
-    options.suppressNonSpeechTokens ? ["--suppress-nst", "--no-prints"] : null,
+    options.suppressNonSpeechTokens ? ["--suppress-nst"] : null,
     ["--processors", String(options.processors)],
     ["--threads", String(options.threads)]
   ];

package/dist/recognition/WhisperServerSTT.cjs

@@ -25,12 +25,15 @@ module.exports = __toCommonJS(WhisperServerSTT_exports);
 var import_node_fs = require("node:fs");
 var import_node_path = require("node:path");
 var import_audio = require("../audio/index.cjs");
+var import_fetch = require("../fetch.cjs");
 var import_WhisperTimeline = require("../utilities/WhisperTimeline.cjs");
 const inputPreference = "file";
 const defaultOptions = {
   baseURL: "http://localhost:8080",
   inferencePath: "/audio/transcriptions",
-  temperature: 0
+  temperature: 0,
+  timeout: 30 * 60 * 1e3
+  // 30 minutes
 };
 async function recognize(input, languageCode, timing, options) {
   const opts = { ...defaultOptions, ...options };
@@ -61,8 +64,16 @@ async function recognize(input, languageCode, timing, options) {
     if (opts.apiKey) {
       headers["Authorization"] = `Bearer ${opts.apiKey}`;
     }
-    const doUpload = () => fetch(url, { method: "POST", body: form, headers });
-    const response = await timing.timeAsync("upload", doUpload);
+    const response = await timing.timeAsync(
+      "upload",
+      async () => fetch(url, {
+        method: "POST",
+        body: form,
+        headers,
+        dispatcher: (0, import_fetch.createTimeoutAgent)(opts.timeout)
+        // this is necessary because  `tsc` from `/web` will cry and shit its pants
+      })
+    );
     if (!response.ok) {
       const text = await response.text();
       throw new Error(`Whisper server error: ${response.status} ${text}`);

package/dist/recognition/WhisperServerSTT.d.cts

@@ -1,4 +1,4 @@
-export { X as InputPreference, J as RecognitionResult, K as WhisperServerOptions, M as inputPreference, N as recognize } from '../AudioFormat-Cq6sRKDj.cjs';
+export { X as InputPreference, J as RecognitionResult, K as WhisperServerOptions, M as inputPreference, N as recognize } from '../AudioFormat-CKxAQAj2.cjs';
 import '../utilities/Timeline.cjs';
 import '../utilities/Timing.cjs';
 import 'node:fs';

package/dist/recognition/WhisperServerSTT.d.ts

@@ -1,4 +1,4 @@
-export { X as InputPreference, J as RecognitionResult, K as WhisperServerOptions, M as inputPreference, N as recognize } from '../AudioFormat--LEggByS.js';
+export { X as InputPreference, J as RecognitionResult, K as WhisperServerOptions, M as inputPreference, N as recognize } from '../AudioFormat-BJuNZJlT.js';
 import '../utilities/Timeline.js';
 import '../utilities/Timing.js';
 import 'node:fs';

package/dist/recognition/WhisperServerSTT.js

@@ -7,6 +7,7 @@ import {
   prepareWavForService,
   toFilePath
 } from "../audio/index.js";
+import { createTimeoutAgent } from "../fetch.js";
 import {
   calculateWhisperSplits,
   countProcessorBoundaries,
@@ -17,7 +18,9 @@ const inputPreference = "file";
 const defaultOptions = {
   baseURL: "http://localhost:8080",
   inferencePath: "/audio/transcriptions",
-  temperature: 0
+  temperature: 0,
+  timeout: 30 * 60 * 1e3
+  // 30 minutes
 };
 async function recognize(input, languageCode, timing, options) {
   const opts = { ...defaultOptions, ...options };
@@ -48,8 +51,16 @@ async function recognize(input, languageCode, timing, options) {
     if (opts.apiKey) {
       headers["Authorization"] = `Bearer ${opts.apiKey}`;
     }
-    const doUpload = () => fetch(url, { method: "POST", body: form, headers });
-    const response = await timing.timeAsync("upload", doUpload);
+    const response = await timing.timeAsync(
+      "upload",
+      async () => fetch(url, {
+        method: "POST",
+        body: form,
+        headers,
+        dispatcher: createTimeoutAgent(opts.timeout)
+        // this is necessary because  `tsc` from `/web` will cry and shit its pants
+      })
+    );
     if (!response.ok) {
       const text = await response.text();
       throw new Error(`Whisper server error: ${response.status} ${text}`);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@storyteller-platform/ghost-story",
-  "version": "0.1.4",
+  "version": "0.1.6",
   "description": "An easy-to-use speech toolset. Fork of the original echogarden project.",
   "author": "Thomas F. K. Jorna",
   "license": "GPL-3.0",
@@ -76,6 +76,7 @@
     "microsoft-cognitiveservices-speech-sdk": "~1.44.0",
     "openai": "~4.103.0",
     "tar": "^7.4.3",
+    "undici": "7.14.0",
     "zod": "^3.25.0"
   },
   "devDependencies": {