tt9/app/validate-languages.gradle

static def validateDictionaryWord(String word, int lineNumber, String validCharacters, String errorMsgPrefix) {
	int errorCount = 0
	def errors = ''

	if (word.matches("(\\d.+?|.+?\\d|\\d)")) {
		errorCount++
		errors += "${errorMsgPrefix}. Found numbers on line ${lineNumber}. Remove all numbers.\n"
	}

	if (word.matches("^\\P{L}+\$")) {
		errorCount++
		errors += "${errorMsgPrefix}. Found a garbage word: '${word}' on line ${lineNumber}.\n"
	}

	if (word.matches("^.\$")) {
		errorCount++
		errors += "${errorMsgPrefix}. Found a single letter: '${word}' on line ${lineNumber}. Only uppercase single letters are allowed. The rest of the alphabet will be added automatically.\n"
	}

	if (errorCount == 0 && !word.matches(validCharacters)) {
		errorCount++
		errors += "${errorMsgPrefix}. Word '${word}' on line ${lineNumber} contains characters outside of the defined alphabet: $validCharacters.\n"
	}

	return [errorCount, errors]
}


static def validateDictionaryLine(String line, int lineNumber) {
	if (line == "") {
		return "There is no word on line ${lineNumber}. Remove all empty lines."
	} else if (line.contains(" ")) {
		return "Found space on line ${lineNumber}. Make sure each word is on a new line. Phrases are not allowed."
	}

	return ''
}


static def extractAlphabetCharsFromLine(String line) {
	if (line.contains('PUNCTUATION') || line.contains('SPECIAL') || !line.matches('\\s+- \\[.+?\\].*')) {
		return ''
	}

	return line.replaceFirst('^\\s+- \\[', '').replaceFirst('\\].*', '').replace(',', '').replace(' ', '')
}


static def parseLanguageFile(File languageFile, String dictionariesDir) {
	String alphabet = languageFile.name.contains("Hebrew") ? '"' : ''
	File dictionaryFile
	int errorCount = 0
	String errorMsg = ""

	boolean hasLayout = false
	boolean isLocaleValid = false
	String localeString = ""
	String dictionaryFileName = ""

	for (String line : languageFile.readLines()) {
		if (
			line.matches("^[a-zA-Z].*")
			&& !line.startsWith("abcString")
			&& !line.startsWith("dictionaryFile")
			&& !line.startsWith("hasUpperCase")
			&& !line.startsWith("layout")
			&& !line.startsWith("locale")
			&& !line.startsWith("name")
		) {
			def parts = line.split(":")
			def property = parts.length > 0 ? parts[0] : line

			errorCount++
			errorMsg += "Language '${languageFile.name}' is invalid. Found unknown property: '${property}'.\n"
		}

		if (line.startsWith("hasUpperCase") && !line.endsWith("yes") && !line.endsWith("no")) {
			def invalidVal = line.replace("hasUpperCase:", "").trim()
			errorCount++
			errorMsg += "Language '${languageFile.name}' is invalid. Unrecognized 'hasUpperCase' value: '${invalidVal}'. Only 'yes' and 'no' are allowed.\n"
		}

		if (line.startsWith("layout")) {
			hasLayout = true
		}

		if (line.startsWith("locale")) {
			localeString = line.replace("locale:", "").trim()
			isLocaleValid = line.matches("^locale:\\s*[a-z]{2}(?:-[A-Z]{2})?")
		}

		if (line.startsWith("dictionaryFile")) {
			dictionaryFileName = line.replace("dictionaryFile:", "").trim()
		}

		def lineCharacters = extractAlphabetCharsFromLine(line)
		alphabet += lineCharacters
	}

	if (!hasLayout) {
		errorCount++
		errorMsg += "Language '${languageFile.name}' is invalid. Missing 'layout' property.\n"
	}

	if (alphabet.isEmpty()) {
		errorCount++
		errorMsg += "Language '${languageFile.name}' is invalid. No language characters found. Make sure 'layout' contains series of characters per each key in the format: ' - [a, b, c]' and so on\n"
	}

	if (!isLocaleValid) {
		errorCount++
		def msg = localeString.isEmpty() ? "Missing 'locale' property." : "Unrecognized locale format: '${localeString}'"
		errorMsg += "Language '${languageFile.name}' is invalid. ${msg}\n"
	}

	dictionaryFile = new File("$dictionariesDir/${dictionaryFileName}")
	if (dictionaryFileName.isEmpty() || !dictionaryFile.exists()) {
		errorCount++
		errorMsg += "Could not find dictionary file: '${dictionaryFileName}' in: '${dictionariesDir}'. Make sure 'dictionaryFile' is set correctly in: '${languageFile.name}'.\n"
	}

	return [alphabet, dictionaryFile, errorCount, errorMsg]
}


static def parseDictionaryFile(String alphabet, File dictionaryFile, int MAX_ERRORS, String CSV_DELIMITER, int MAX_WORD_FREQUENCY) {
	final GEOGRAPHICAL_NAME = ~"[A-Z]\\w+-[^\\n]+"
	final VALID_CHARS = alphabet.toUpperCase() == alphabet ? "^[${alphabet}\\-\\.']+\$" : "^[${alphabet}${alphabet.toUpperCase()}\\-\\.']+\$"
	final int MAX_SORTING_ERRORS = Math.ceil(MAX_ERRORS / 10)

	def uniqueWords = [:]

	int errorCount = 0
	int sortingErrorCount = 0
	String errorMsg = ""

	def fileContents = dictionaryFile.readLines()
	for (int lineNumber = 1, previousWordLength = 0; lineNumber <= fileContents.size() && errorCount < MAX_ERRORS; lineNumber++) {
		String line = fileContents.get(lineNumber - 1)

		String error = validateDictionaryLine(line, lineNumber)
		if (!error.isEmpty()) {
			errorCount++
			errorMsg += "Dictionary '${dictionaryFile.name}' is invalid. ${error}\n"
			break
		}

		String[] parts = line.split(CSV_DELIMITER, 2)
		String word = parts[0]
		int frequency
		try {
			frequency = (parts.length > 1 ? parts[1] : "0") as int
		} catch (Exception ignored) {
			frequency = -1
		}

		if (frequency < 0 || frequency > MAX_WORD_FREQUENCY) {
			errorCount++
			errorMsg += "Dictionary '${dictionaryFile.name}' is invalid. Found out-of-range word frequency: '${parts[1]}' on line ${lineNumber}. Frequency must be an integer between 0 and ${MAX_WORD_FREQUENCY}.\n"
		}

		if (sortingErrorCount < MAX_SORTING_ERRORS && word.length() < previousWordLength) {
			sortingErrorCount++
			errorCount++

			if (sortingErrorCount == MAX_SORTING_ERRORS) {
				errorMsg += "Too many sorting errors in '${dictionaryFile.name}'. Disabling sorting check until the end of the file.\n"
			} else {
				errorMsg += "Dictionary '${dictionaryFile.name}' is not sorted. Word: '${word}' on line ${lineNumber} is shorter than the previous one. Ensure all words are sorted by length and sequence.\n"
			}
		}
		previousWordLength = word.length()

		def (wordErrorCount, wordErrors) = validateDictionaryWord(word, lineNumber, VALID_CHARS, "Dictionary '${dictionaryFile.name}' is invalid")
		errorCount += wordErrorCount
		errorMsg += wordErrors

		String uniqueWordKey = word ==~ GEOGRAPHICAL_NAME ? word : word.toLowerCase()
		if (uniqueWords[uniqueWordKey] != null && uniqueWords[uniqueWordKey] == true) {
			errorCount++
			errorMsg += "Dictionary '${dictionaryFile.name}' is invalid. Found a repeating word: '${word}' on line ${lineNumber}. Ensure all words appear only once.\n"
		} else {
			uniqueWords[uniqueWordKey] = true
		}
	}

	return [errorMsg, errorCount]
}


ext.validateLanguageFiles = { definitionsDir, dictionariesDir, outputFile ->
	int errorCount = 0

	outputFile.text = ""

	def errorStream = fileTree(definitionsDir).getFiles().parallelStream().map { File languageFile ->
		if (errorCount >= MAX_ERRORS) {
			return "Too many errors! Skipping: ${languageFile}\n"
		}

		def (alphabet, dictionaryFile, langFileErrorCount, langFileErrorMsg) = parseLanguageFile(languageFile, dictionariesDir)
		errorCount += langFileErrorCount
		if (!langFileErrorMsg.isEmpty()) {
			outputFile.text += "${languageFile.name} INVALID \n"
			return langFileErrorMsg
		}

		def (dictionaryErrorMsg, dictionaryErrorCount) = parseDictionaryFile(alphabet, dictionaryFile, MAX_ERRORS, CSV_DELIMITER, MAX_WORD_FREQUENCY)
		errorCount += dictionaryErrorCount
		if (!dictionaryErrorMsg.isEmpty()) {
			outputFile.text += "${languageFile.name} INVALID \n"
			return dictionaryErrorMsg
		}

		outputFile.text += "${languageFile.name} OK\n"
		return ""
	}

	String errorsMsg = errorStream.reduce("", String::concat)
	if (errorsMsg) {
		throw new GradleException(errorsMsg)
	}
}