echogarden 2.0.7 → 2.0.9

package/src/cli/CLI.ts

@@ -22,7 +22,7 @@ import { removePackage } from '../utilities/PackageManager.js'
 import { appName } from '../api/Common.js'
 import { ServerOptions, startServer } from '../server/Server.js'
 import { OpenPromise } from '../utilities/OpenPromise.js'
-import { getLowercaseFileExtension, joinPath, parsePath, resolveToModuleRootDir } from '../utilities/PathUtilities.js'
+import { getDirName, getFileNameWithoutExtension, getLowercaseFileExtension, joinPath, parsePath, resolveToModuleRootDir } from '../utilities/PathUtilities.js'
 import { CLIOptions, CLIOptionsKeys } from './CLIOptions.js'
 import { convertHtmlToText, formatIntegerWithLeadingZeros, formatListWithQuotedElements } from '../utilities/StringUtilities.js'
 
@@ -517,9 +517,7 @@ export async function speak(operationData: CLIOperationData) {
 	}
 
 	for (const outputFilename of outputFilenames) {
-		const placeholderPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-		if (placeholderPatternMatch) {
+		if (isPlaceholderFilePath(outputFilename)) {
 			continue
 		}
 
@@ -562,9 +560,7 @@ export async function transcribe(operationData: CLIOperationData) {
 	}
 
 	for (const outputFilename of outputFilenames) {
-		const partPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-		if (partPatternMatch) {
+		if (isPlaceholderFilePath(outputFilename)) {
 			continue
 		}
 
@@ -658,9 +654,7 @@ export async function align(operationData: CLIOperationData) {
 	}
 
 	for (const outputFilename of outputFilenames) {
-		const partPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-		if (partPatternMatch) {
+		if (isPlaceholderFilePath(outputFilename)) {
 			continue
 		}
 
@@ -763,9 +757,7 @@ export async function alignTranslation(operationData: CLIOperationData) {
 	}
 
 	for (const outputFilename of outputFilenames) {
-		const partPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-		if (partPatternMatch) {
+		if (isPlaceholderFilePath(outputFilename)) {
 			continue
 		}
 
@@ -892,9 +884,7 @@ export async function alignTranscriptAndTranslation(operationData: CLIOperationD
 	}
 
 	for (const outputFilename of outputFilenames) {
-		const partPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-		if (partPatternMatch) {
+		if (isPlaceholderFilePath(outputFilename)) {
 			continue
 		}
 
@@ -994,9 +984,7 @@ export async function alignTimelineTranslation(operationData: CLIOperationData)
 	const allowOverwrite = getWithDefault(cliOptions.overwrite, overwriteByDefault)
 
 	for (const outputFilename of outputFilenames) {
-		const partPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-		if (partPatternMatch) {
+		if (isPlaceholderFilePath(outputFilename)) {
 			continue
 		}
 
@@ -1071,9 +1059,7 @@ export async function translateText(operationData: CLIOperationData) {
 		logger.start('\nWrite output files')
 
 		for (const outputFilename of outputFilenames) {
-			const partPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-			if (partPatternMatch) {
+			if (isPlaceholderFilePath(outputFilename)) {
 				continue
 			}
 
@@ -1134,9 +1120,7 @@ export async function translateSpeech(operationData: CLIOperationData) {
 	}
 
 	for (const outputFilename of outputFilenames) {
-		const partPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-		if (partPatternMatch) {
+		if (isPlaceholderFilePath(outputFilename)) {
 			continue
 		}
 
@@ -1298,9 +1282,7 @@ export async function detectVoiceActivity(operationData: CLIOperationData) {
 	}
 
 	for (const outputFilename of outputFilenames) {
-		const partPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-		if (partPatternMatch) {
+		if (isPlaceholderFilePath(outputFilename)) {
 			continue
 		}
 
@@ -1879,13 +1861,7 @@ async function checkOutputFilenames(outputFilenames: string[], acceptMediaOutput
 			throw new Error(errorText)
 		}
 
-		const placeholderPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-		if (placeholderPatternMatch) {
-			if (placeholderPatternMatch[1] != 'segment') {
-				throw new Error(`Invalid placeholder pattern: '${placeholderPatternMatch[1]}'. Placeholder output filename pattern currently only supports 'segment'. For example: '/out/[segment].wav'`)
-			}
-
+		if (isPlaceholderFilePath(outputFilename)) {
 			includesPlaceholderPattern = true
 		}
 	}
@@ -1911,13 +1887,13 @@ async function writeOutputFilesForSegment(outputFilenames: string[], index: numb
 	}
 
 	for (const outputFilename of outputFilenames) {
-		const placeholderPatternMatch = outputFilename.match(filenamePlaceholderPattern)
-
-		if (!placeholderPatternMatch) {
+		if (!isPlaceholderFilePath(outputFilename)) {
 			continue
 		}
 
-		const segmentFilename = outputFilename.replace(filenamePlaceholderPattern, `${formatIntegerWithLeadingZeros(index + 1, digitCount)} - ${initialText}.$2`)
+		const fileDir = getDirName(outputFilename)
+		const fileExtension = getLowercaseFileExtension(outputFilename)
+		const segmentFilename = joinPath(fileDir, `${formatIntegerWithLeadingZeros(index + 1, digitCount)} - ${initialText}.${fileExtension}`)
 
 		const fileSaver = getFileSaver(segmentFilename, allowOverwrite)
 		await fileSaver(audio, timeline, text)
@@ -2016,12 +1992,16 @@ function getFileSaver(outputFilePath: string, allowOverwrite: boolean): FileSave
 	return fileSaver
 }
 
+function isPlaceholderFilePath(filePath: string) {
+	const filenameWithoutExtension = getFileNameWithoutExtension(filePath)
+
+	return filenameWithoutExtension === '[segment]'
+}
+
 const supportedMetadataFileExtensions = ['txt', 'json']
 const supportedSubtitleFileExtensions = ['srt', 'vtt']
 const supportedOutputMediaFileExtensions = ['wav', 'mp3', 'opus', 'm4a', 'ogg', 'flac']
 
-const filenamePlaceholderPattern = /\[(.*)\]\.(.*)$/
-
 const overwriteByDefault = false
 
 startIfInWorkerThread()

package/src/nlp/ChineseSegmentation.ts

@@ -61,7 +61,7 @@ let JiebaWasmInstance: typeof import('jieba-wasm')
 
 async function getJiebaWasmInstance() {
 	if (!JiebaWasmInstance) {
-		const { default: JiebaWasm } = await import('jieba-wasm')
+		const JiebaWasm = await import('jieba-wasm')
 
 		JiebaWasmInstance = JiebaWasm
 	}

package/src/synthesis/VitsTTS.ts

@@ -149,7 +149,7 @@ export class VitsTTS {
 
 		await logger.startAsync('Generate audio using synthesis model')
 
-		const { default: Onnx } = await import('onnxruntime-node')
+		const Onnx = await import('onnxruntime-node')
 
 		const inputTensor = new Onnx.Tensor('int64', bigIntIds, [1, bigIntIds.length])
 		const inputLengthsTensor = new Onnx.Tensor('int64', idLengths, [1])
@@ -187,7 +187,7 @@ export class VitsTTS {
 		const logger = new Logger()
 		await logger.startAsync('Initialize VITS ONNX synthesis model')
 
-		const { default: Onnx } = await import('onnxruntime-node')
+		const Onnx = await import('onnxruntime-node')
 
 		const filesInModelPath = await readdir(this.modelPath)
 		const onnxModelFilename = filesInModelPath.find(filename => filename.endsWith('.onnx'))

package/src/text-translation/GoogleTranslateTextTranslation.ts

@@ -1,8 +1,8 @@
 import { request } from "gaxios"
 import { Logger } from "../utilities/Logger.js"
-import { Fragment, splitToFragments, splitToLines } from "../nlp/Segmentation.js"
+import { Fragment, splitToFragments } from "../nlp/Segmentation.js"
 import { TranslationPair } from "../api/TextTranslation.js"
-import { getChromeOnWindowsHeaders, getChromeOnAndroidHeaders } from "../utilities/BrowserRequestHeaders.js"
+import { getChromeOnAndroidHeaders } from "../utilities/BrowserRequestHeaders.js"
 import { logToStderr } from "../utilities/Utilities.js"
 import { getShortLanguageCode } from "../utilities/Locale.js"
 import { PlainTextOptions } from "../api/Common.js"
@@ -90,7 +90,7 @@ export async function translateText(
 
 		const translatedTextForPart = fragmentTranslationPair[0].translatedText
 
-		logger.logTitledMessage(`Translated part`, `"${translatedTextForPart.replaceAll(' | ', '\n') }"`)
+		logger.logTitledMessage(`Translated part`, `"${translatedTextForPart.replaceAll(' | ', '\n')}"`)
 
 		const splitTranslation = translatedTextForPart.split(`|`)
 
@@ -211,241 +211,6 @@ export async function translateText_MobileWeb(
 	}]
 }
 
-export async function translateText_WebAPI(text: string, sourceLanguage: string, targetLanguage: string, tld = 'com') {
-	const logger = new Logger()
-
-	logger.start(`Prepare request`)
-
-	const bodyFormParameters = [[text, sourceLanguage, targetLanguage, 1], []]
-
-	const requestForm = [[['MkEWBc', JSON.stringify(bodyFormParameters), null, 'generic']]]
-	const stringifiedForm = JSON.stringify(requestForm)
-	const requestBody =
-		`f.req=${encodeURIComponent(stringifiedForm)}&`// +
-
-	// Example:
-	//
-	// f.req: [[["MkEWBc","[[\"Hey there!\\n\\nHow are you doing today?\",\"en\",\"es\",1],[]]",null,"generic"]]]
-	// at: AFS6Qyg7vBwAdNvYB3ncQi_YvmYH:1717146157247
-
-	logger.start(`Request translation from Google Translate`)
-
-	const response = await request<string>({
-		method: 'POST',
-
-		url: `https://translate.google.${tld}/_/TranslateWebserverUi/data/batchexecute`,
-
-		params: {
-			'rpcids': 'MkEWBc',
-			/*'source-path': '/',
-			'f.sid': '4165351081036910547',
-			'bl': 'boq_translate-webserver_20240528.08_p1',
-			'hl': 'en-US',
-			'soc-app': '1',
-			'soc-platform': '1',
-			'soc-device': '1',
-			'_reqid': '435214',
-			*/
-			'rt': 'c',
-		},
-
-		headers: {
-			...getChromeOnWindowsHeaders({
-				origin: `https://translate.google.${tld}`,
-				referrer: `https://translate.google.${tld}/`
-			}),
-
-			'Content-Type': 'application/x-www-form-urlencoded',
-		},
-
-		body: requestBody,
-
-		//proxy: 'http://localhost:8080',
-
-		responseType: 'text'
-	})
-
-	logger.start('Parse response')
-
-	const responseLines = splitToLines(response.data).filter(line => line.trim().length > 0)
-
-	const dataLine = responseLines[2]
-
-	const outerArray = JSON.parse(dataLine)
-	const innerArray = JSON.parse(outerArray[0][2])
-	const sentenceEntries: any[] = innerArray[1][0][0][5]
-
-	const translationsPairs = sentenceEntries.map<TranslationPair>((element: any) => {
-		const translationPair: TranslationPair = {
-			sourceText: element[6],
-			translatedText: element[0],
-		}
-
-		return translationPair
-	})
-
-	logger.end()
-
-	return translationsPairs
-}
-
-export async function translateText_OldWebAPI(text: string, sourceLanguage: string, targetLanguage: string, tld = 'com') {
-	const logger = new Logger()
-
-	logger.start(`Prepare request`)
-
-	const token = getToken(text)
-
-	logger.start(`Request translation from Google Translate`)
-
-	const response = await request<string>({
-		method: 'GET',
-
-		url: `https://translate.google.${tld}/translate_a/single`,
-
-		params: {
-			client: 't',
-			sl: normalizeLanguageCodeForGoogleTranslate(sourceLanguage),
-			tl: normalizeLanguageCodeForGoogleTranslate(targetLanguage),
-			hl: normalizeLanguageCodeForGoogleTranslate(targetLanguage),
-			dt: ['at', 'bd', 'ex', 'ld', 'md', 'qca', 'rw', 'rm', 'ss', 't'],
-			ie: 'UTF-8',
-			oe: 'UTF-8',
-			otf: 1,
-			ssel: 0,
-			tsel: 0,
-			kc: 7,
-			q: text,
-			tk: token,
-		},
-
-		headers: getChromeOnWindowsHeaders({
-			origin: `https://translate.google.${tld}`,
-			referrer: `https://translate.google.${tld}/`
-		}),
-
-		responseType: 'json'
-	})
-
-	const translationsPairs: TranslationPair[] = []
-
-	for (const element of response.data[0] as any as any[]) {
-		if (element == null || element[0] == null || element[1] == null) {
-			continue
-		}
-
-		translationsPairs.push({
-			sourceText: element[1],
-			translatedText: element[0],
-		})
-	}
-
-	logger.end()
-
-	return translationsPairs
-}
-
-function getToken(query: any) {
-	function shiftLeftOrRightThenSumOrXor(num: number, optString: any) {
-		for (let i = 0; i < optString.length - 2; i += 3) {
-			let acc = optString.charAt(i + 2)
-
-			if ('a' <= acc) {
-				acc = acc.charCodeAt(0) - 87
-			} else {
-				acc = Number(acc)
-			}
-
-			if (optString.charAt(i + 1) == '+') {
-				acc = num >>> acc
-			} else {
-				acc = num << acc
-			}
-
-			if (optString.charAt(i) == '+') {
-				num += acc & 4294967295
-			} else {
-				num ^= acc
-			}
-		}
-
-		return num
-	}
-
-	function transformQuery(query: any) {
-		const bytesArray: number[] = []
-
-		let idx: number = 0
-
-		for (let i = 0; i < query.length; i++) {
-			let charCode = query.charCodeAt(i)
-
-			if (128 > charCode) {
-				bytesArray[idx++] = charCode
-			} else {
-				if (2048 > charCode) {
-					bytesArray[idx++] = (charCode >> 6) | 192
-				} else {
-					if (
-						55296 == (charCode & 64512) &&
-						i + 1 < query.length &&
-						56320 == (query.charCodeAt(i + 1) & 64512)
-					) {
-						charCode =
-							65536 +
-							((charCode & 1023) << 10) +
-							(query.charCodeAt(++i) & 1023)
-
-						bytesArray[idx++] = (charCode >> 18) | 240
-						bytesArray[idx++] = ((charCode >> 12) & 63) | 128
-					} else {
-						bytesArray[idx++] = (charCode >> 12) | 224
-					}
-
-					bytesArray[idx++] = ((charCode >> 6) & 63) | 128
-				}
-
-				bytesArray[idx++] = (charCode & 63) | 128
-			}
-		}
-
-		return bytesArray
-	}
-
-	function calcHash(query: any, windowTkk: any) {
-		const tkkSplited = windowTkk.split('.')
-		const tkkIndex = Number(tkkSplited[0]) || 0
-		const tkkKey = Number(tkkSplited[1]) || 0
-
-		const bytesArray = transformQuery(query)
-
-		let encondingRound = tkkIndex
-
-		for (let i = 0; i < bytesArray.length; i++) {
-			encondingRound += bytesArray[i]
-
-			encondingRound = shiftLeftOrRightThenSumOrXor(encondingRound, '+-a^+6')
-		}
-
-		encondingRound = shiftLeftOrRightThenSumOrXor(encondingRound, '+-3^+b+-f')
-
-		encondingRound ^= tkkKey
-
-		if (encondingRound <= 0) {
-			encondingRound = (encondingRound & 2147483647) + 2147483648
-		}
-
-		const normalizedResult = encondingRound % 1000000
-
-		return normalizedResult.toString() + '.' + (normalizedResult ^ tkkIndex)
-	}
-
-	// TKK value from https://github.com/FilipePS/Traduzir-paginas-web/blob/f3a4956a1aa96b7a9124864158a5200827694521/background/translationService.js
-	const googleTranslateTKK = '448487.932609646'
-
-	return calcHash(query, googleTranslateTKK)
-}
-
 function normalizeLanguageCodeForGoogleTranslate(languageCode: string) {
 	switch (languageCode) {
 		case 'zh': {
@@ -555,6 +320,8 @@ export const supportedLanguageCodes = [
 	'ha', // Hausa
 	'haw', // Hawaiian
 	'he', // Hebrew
+	'iu', // Inuktitut (Syllabics)
+	'iu-Latn', // Inuktitut (Latin script)
 	'iw', // Hebrew
 	'hil', // Hiligaynon
 	'hi', // Hindi
@@ -657,7 +424,7 @@ export const supportedLanguageCodes = [
 	'sm', // Samoan
 	'sg', // Sango
 	'sa', // Sanskrit
-	'sat-Latn', // Santali
+	'sat-Latn', // Santali (Latin script)
 	'gd', // Scots Gaelic
 	'nso', // Sepedi
 	'sr', // Serbian
@@ -680,7 +447,7 @@ export const supportedLanguageCodes = [
 	'sv', // Swedish
 	'ty', // Tahitian
 	'tg', // Tajik
-	'ber-Latn', // Tamazight
+	'ber-Latn', // Tamazight (Latin script)
 	'ber', // Tamazight (Tifinagh)
 	'ta', // Tamil
 	'tt', // Tatar

package/src/typings/Fillers.d.ts

@@ -21,7 +21,6 @@ declare module 'cldr-segmentation' {
 	}
 }
 
-declare module 'sam-js'
 declare module 'winax'
 
 declare module 'tinyld' {

package/src/utilities/Utilities.ts

@@ -145,7 +145,7 @@ export function printMatrix(matrix: Float32Array[]) {
 
 export async function parseJson(jsonText: string, useJson5 = false) {
 	if (useJson5) {
-		const { default: JSON5 } = await import('json5')
+		const JSON5 = await import('json5')
 
 		return JSON5.parse(jsonText)
 	} else {
@@ -157,7 +157,7 @@ export async function stringifyAndFormatJson(obj: any, useJson5 = false) {
 	let textContent: string
 
 	if (useJson5) {
-		const { default: JSON5 } = await import('json5')
+		const JSON5 = await import('json5')
 
 		textContent = JSON5.stringify(obj, undefined, 4)
 	} else {