echogarden 1.4.4 → 1.6.0

package/src/text-translation/GoogleTranslateTextTranslation.ts

@@ -0,0 +1,667 @@
+import { request } from "gaxios"
+import { Logger } from "../utilities/Logger.js"
+import { Fragment, splitToFragments, splitToLines, splitToParagraphs, splitToSentences } from "../nlp/Segmentation.js"
+import { TranslationPair } from "../api/TextTranslation.js"
+import { getChromeOnWindowsHeaders, getChromeOnAndroidHeaders } from "../utilities/BrowserRequestHeaders.js"
+import { logToStderr } from "../utilities/Utilities.js"
+import { getShortLanguageCode } from "../utilities/Locale.js"
+
+const log = logToStderr
+
+export async function translateText(text: string, sourceLanguage: string, targetLanguage: string, tld = 'com') {
+	const logger = new Logger()
+
+	const maxTextLengthPerRequest = 2000
+
+	const paragraphs = splitToParagraphs(text, 'double', 'preserve')
+
+	const fragmentsForParagraph: Fragment[][] = []
+
+	for (const paragraph of paragraphs) {
+		const fragments = await splitToFragments(paragraph, maxTextLengthPerRequest, getShortLanguageCode(sourceLanguage))
+
+		fragmentsForParagraph.push(fragments)
+	}
+
+	const fragmentBundles: Fragment[][] = [[]]
+
+	function totalLengthOfLastBundle() {
+		let total = 0
+
+		for (const fragment of fragmentBundles[fragmentBundles.length - 1]) {
+			total += fragment.length
+		}
+
+		return total
+	}
+
+	const fragmentToParagraphIndex = new Map<Fragment, number>()
+
+	for (let paragraphIndex = 0; paragraphIndex < paragraphs.length; paragraphIndex++) {
+		const fragments = fragmentsForParagraph[paragraphIndex]
+
+		for (const fragment of fragments) {
+			let lastBundle = fragmentBundles[fragmentBundles.length - 1]
+
+			if (totalLengthOfLastBundle() + fragment.length > maxTextLengthPerRequest) {
+				lastBundle = []
+				fragmentBundles.push(lastBundle)
+			}
+
+			lastBundle.push(fragment)
+
+			fragmentToParagraphIndex.set(fragment, paragraphIndex)
+		}
+	}
+
+	const translatedFragmentsForParagraphs = paragraphs.map(_ => [] as string[])
+
+	for (let fragmentBundleIndex = 0; fragmentBundleIndex < fragmentBundles.length; fragmentBundleIndex++) {
+		const bundle = fragmentBundles[fragmentBundleIndex]
+
+		const joinedBundle = bundle.map(x => x.text.replaceAll('|', ' ')).join(` | `)
+
+		logger.logTitledMessage(`\nTranslate part ${fragmentBundleIndex + 1} of ${fragmentBundles.length}`, joinedBundle)
+
+		logger.start(`Request translation from Google Translate`)
+		const fragmentTranslationPair = await translateText_MobileWeb(joinedBundle, sourceLanguage, targetLanguage)
+		logger.end()
+
+		const translatedTextForBundle = fragmentTranslationPair[0].translatedText
+
+		logger.logTitledMessage(`Translated part`, `"${translatedTextForBundle}"`)
+
+		const splitTranslation = translatedTextForBundle.split(`|`)
+
+		for (let fragmentIndex = 0; fragmentIndex < bundle.length; fragmentIndex++) {
+			const fragment = bundle[fragmentIndex]
+			const translatedFragment = splitTranslation[fragmentIndex].trim()
+
+			const paragraphIndex = fragmentToParagraphIndex.get(fragment)!
+
+			translatedFragmentsForParagraphs[paragraphIndex].push(translatedFragment)
+		}
+	}
+
+	const translationPairs: TranslationPair[] = []
+
+	for (let paragraphIndex = 0; paragraphIndex < paragraphs.length; paragraphIndex++) {
+		translationPairs.push({
+			sourceText: paragraphs[paragraphIndex],
+			translatedText: translatedFragmentsForParagraphs[paragraphIndex].join(' ')
+		})
+	}
+
+	return translationPairs
+}
+
+export async function translateText_MobileWeb(text: string, sourceLanguage: string, targetLanguage: string, tld = 'com') {
+	const logger = new Logger()
+
+	logger.start(`Request translation from Google Translate`)
+
+	const response = await request<string>({
+		method: 'GET',
+
+		url: `https://translate.google.${tld}/m`,
+
+		//proxy: 'http://localhost:8080',
+
+		params: {
+			'sl': sourceLanguage,
+			'tl': targetLanguage,
+			'q': text,
+		},
+
+		headers: {
+			...getChromeOnAndroidHeaders({
+				origin: `https://translate.google.${tld}`,
+				referrer: `https://translate.google.${tld}/m`
+			}),
+		},
+
+		responseType: 'text',
+	})
+
+	logger.start('Parse response')
+
+	const html = response.data
+
+	let translatedText: string
+
+	const useJsdom = false
+
+	if (useJsdom) {
+		const { JSDOM } = await import('jsdom')
+
+		const doc = new JSDOM(html, {})
+
+		const resultElement = doc.window.document.querySelector('.result-container')
+
+		translatedText = resultElement?.textContent || ''
+	} else {
+		const regex = /<div\s+class="result-container".*?<\/div>/is
+		const match = html.match(regex)
+
+		if (match) {
+			const { htmlToText } = await import('html-to-text')
+
+			translatedText = htmlToText(match[0], { wordwrap: Infinity }).trim()
+		} else {
+			translatedText = ''
+		}
+	}
+
+	logger.end()
+
+	return [{
+		sourceText: text,
+		translatedText: translatedText
+	}]
+}
+
+export async function translateText_WebAPI(text: string, sourceLanguage: string, targetLanguage: string, tld = 'com') {
+	const logger = new Logger()
+
+	logger.start(`Prepare request`)
+
+	const bodyFormParameters = [[text, sourceLanguage, targetLanguage, 1], []]
+
+	const requestForm = [[['MkEWBc', JSON.stringify(bodyFormParameters), null, 'generic']]]
+	const stringifiedForm = JSON.stringify(requestForm)
+	const requestBody =
+		`f.req=${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': {
+			return 'zh-CN'
+		}
+
+		case 'he': {
+			return 'iw'
+		}
+
+		case 'jv': {
+			return 'jw'
+		}
+
+		default: {
+			return languageCode
+		}
+	}
+}
+
+export function supportsLanguage(langCode: string) {
+	return supportedLanguages.includes(normalizeLanguageCodeForGoogleTranslate(langCode))
+}
+
+// 243 Languages supported
+export const supportedLanguages = [
+	'ab', // Abkhaz
+	'ace', // Acehnese
+	'ach', // Acholi
+	'aa', // Afar
+	'af', // Afrikaans
+	'sq', // Albanian
+	'alz', // Alur
+	'am', // Amharic
+	'ar', // Arabic
+	'hy', // Armenian
+	'as', // Assamese
+	'av', // Avar
+	'awa', // Awadhi
+	'ay', // Aymara
+	'az', // Azerbaijani
+	'ban', // Balinese
+	'bal', // Baluchi
+	'bm', // Bambara
+	'bci', // Baoulé
+	'ba', // Bashkir
+	'eu', // Basque
+	'btx', // Batak Karo
+	'bts', // Batak Simalungun
+	'bbc', // Batak Toba
+	'be', // Belarusian
+	'bem', // Bemba
+	'bn', // Bengali
+	'bew', // Betawi
+	'bho', // Bhojpuri
+	'bik', // Bikol
+	'bs', // Bosnian
+	'br', // Breton
+	'bg', // Bulgarian
+	'bua', // Buryat
+	'yue', // Cantonese
+	'ca', // Catalan
+	'ceb', // Cebuano
+	'ch', // Chamorro
+	'ce', // Chechen
+	'ny', // Chichewa
+	'zh-CN', // Chinese (Simplified)
+	'zh-TW', // Chinese (Traditional)
+	'chk', // Chuukese
+	'cv', // Chuvash
+	'co', // Corsican
+	'crh', // Crimean Tatar
+	'hr', // Croatian
+	'cs', // Czech
+	'da', // Danish
+	'fa-AF', // Dari
+	'dv', // Dhivehi
+	'din', // Dinka
+	'doi', // Dogri
+	'dov', // Dombe
+	'nl', // Dutch
+	'dyu', // Dyula
+	'dz', // Dzongkha
+	'en', // English
+	'eo', // Esperanto
+	'et', // Estonian
+	'ee', // Ewe
+	'fo', // Faroese
+	'fj', // Fijian
+	'tl', // Filipino
+	'fi', // Finnish
+	'fon', // Fon
+	'fr', // French
+	'fy', // Frisian
+	'fur', // Friulian
+	'ff', // Fulani
+	'gaa', // Ga
+	'gl', // Galician
+	'ka', // Georgian
+	'de', // German
+	'el', // Greek
+	'gn', // Guarani
+	'gu', // Gujarati
+	'ht', // Haitian Creole
+	'cnh', // Hakha Chin
+	'ha', // Hausa
+	'haw', // Hawaiian
+	'iw', // Hebrew
+	'hil', // Hiligaynon
+	'hi', // Hindi
+	'hmn', // Hmong
+	'hu', // Hungarian
+	'hrx', // Hunsrik
+	'iba', // Iban
+	'is', // Icelandic
+	'ig', // Igbo
+	'ilo', // Ilocano
+	'id', // Indonesian
+	'ga', // Irish
+	'it', // Italian
+	'jam', // Jamaican Patois
+	'ja', // Japanese
+	'jw', // Javanese
+	'kac', // Jingpo
+	'kl', // Kalaallisut
+	'kn', // Kannada
+	'kr', // Kanuri
+	'pam', // Kapampangan
+	'kk', // Kazakh
+	'kha', // Khasi
+	'km', // Khmer
+	'cgg', // Kiga
+	'kg', // Kikongo
+	'rw', // Kinyarwanda
+	'ktu', // Kituba
+	'trp', // Kokborok
+	'kv', // Komi
+	'gom', // Konkani
+	'ko', // Korean
+	'kri', // Krio
+	'ku', // Kurdish (Kurmanji)
+	'ckb', // Kurdish (Sorani)
+	'ky', // Kyrgyz
+	'lo', // Lao
+	'ltg', // Latgalian
+	'la', // Latin
+	'lv', // Latvian
+	'lij', // Ligurian
+	'li', // Limburgish
+	'ln', // Lingala
+	'lt', // Lithuanian
+	'lmo', // Lombard
+	'lg', // Luganda
+	'luo', // Luo
+	'lb', // Luxembourgish
+	'mk', // Macedonian
+	'mad', // Madurese
+	'mai', // Maithili
+	'mak', // Makassar
+	'mg', // Malagasy
+	'ms', // Malay
+	'ms-Arab', // Malay (Jawi)
+	'ml', // Malayalam
+	'mt', // Maltese
+	'mam', // Mam
+	'gv', // Manx
+	'mi', // Maori
+	'mr', // Marathi
+	'mh', // Marshallese
+	'mwr', // Marwadi
+	'mfe', // Mauritian Creole
+	'chm', // Meadow Mari
+	'mni-Mtei', // Meiteilon (Manipuri)
+	'min', // Minang
+	'lus', // Mizo
+	'mn', // Mongolian
+	'my', // Myanmar (Burmese)
+	'nhe', // Nahuatl (Eastern Huasteca)
+	'ndc-ZW', // Ndau
+	'nr', // Ndebele (South)
+	'new', // Nepalbhasa (Newari)
+	'ne', // Nepali
+	'bm-Nkoo', // NKo
+	'no', // Norwegian
+	'nus', // Nuer
+	'oc', // Occitan
+	'or', // Odia (Oriya)
+	'om', // Oromo
+	'os', // Ossetian
+	'pag', // Pangasinan
+	'pap', // Papiamento
+	'ps', // Pashto
+	'fa', // Persian
+	'pl', // Polish
+	'pt', // Portuguese (Brazil)
+	'pt-PT', // Portuguese (Portugal)
+	'pa', // Punjabi (Gurmukhi)
+	'pa-Arab', // Punjabi (Shahmukhi)
+	'qu', // Quechua
+	'kek', // Qʼeqchiʼ
+	'rom', // Romani
+	'ro', // Romanian
+	'rn', // Rundi
+	'ru', // Russian
+	'se', // Sami (North)
+	'sm', // Samoan
+	'sg', // Sango
+	'sa', // Sanskrit
+	'sat-Latn', // Santali
+	'gd', // Scots Gaelic
+	'nso', // Sepedi
+	'sr', // Serbian
+	'st', // Sesotho
+	'crs', // Seychellois Creole
+	'shn', // Shan
+	'sn', // Shona
+	'scn', // Sicilian
+	'szl', // Silesian
+	'sd', // Sindhi
+	'si', // Sinhala
+	'sk', // Slovak
+	'sl', // Slovenian
+	'so', // Somali
+	'es', // Spanish
+	'su', // Sundanese
+	'sus', // Susu
+	'sw', // Swahili
+	'ss', // Swati
+	'sv', // Swedish
+	'ty', // Tahitian
+	'tg', // Tajik
+	'ber-Latn', // Tamazight
+	'ber', // Tamazight (Tifinagh)
+	'ta', // Tamil
+	'tt', // Tatar
+	'te', // Telugu
+	'tet', // Tetum
+	'th', // Thai
+	'bo', // Tibetan
+	'ti', // Tigrinya
+	'tiv', // Tiv
+	'tpi', // Tok Pisin
+	'to', // Tongan
+	'ts', // Tsonga
+	'tn', // Tswana
+	'tcy', // Tulu
+	'tum', // Tumbuka
+	'tr', // Turkish
+	'tk', // Turkmen
+	'tyv', // Tuvan
+	'ak', // Twi
+	'udm', // Udmurt
+	'uk', // Ukrainian
+	'ur', // Urdu
+	'ug', // Uyghur
+	'uz', // Uzbek
+	've', // Venda
+	'vec', // Venetian
+	'vi', // Vietnamese
+	'war', // Waray
+	'cy', // Welsh
+	'wo', // Wolof
+	'xh', // Xhosa
+	'sah', // Yakut
+	'yi', // Yiddish
+	'yo', // Yoruba
+	'yua', // Yucatec Maya
+	'zap', // Zapotec
+	'zu', // Zulu
+]