tercul-backend/internal/jobs/linguistics/language_detector.go

package linguistics

import (
	"strings"
)

const (
	LangEN = "en"
	LangES = "es"
	LangFR = "fr"
)

// languageDetector detects the language of a text
type languageDetector struct{}

// NewLanguageDetector creates a new LanguageDetector
func NewLanguageDetector() *languageDetector {
	return &languageDetector{}
}

// Detect detects the language of a text and returns the language code, confidence, and error
func (d *languageDetector) DetectLanguage(text string) (string, error) {
	// This is a simplified implementation
	// In a real-world scenario, you would use a library like github.com/pemistahl/lingua-go
	// or call an external API for language detection

	// For demonstration purposes, we'll use a simple heuristic based on common words
	content := strings.ToLower(text)

	// Check for English
	englishWords := []string{"the", "and", "is", "in", "to", "of", "that", "for"}
	englishCount := countWords(content, englishWords)

	// Check for Spanish
	spanishWords := []string{"el", "la", "es", "en", "de", "que", "por", "para"}
	spanishCount := countWords(content, spanishWords)

	// Check for French
	frenchWords := []string{"le", "la", "est", "en", "de", "que", "pour", "dans"}
	frenchCount := countWords(content, frenchWords)

	// Determine the most likely language
	if englishCount > spanishCount && englishCount > frenchCount {
		return LangEN, nil
	} else if spanishCount > englishCount && spanishCount > frenchCount {
		return LangES, nil
	} else if frenchCount > englishCount && frenchCount > spanishCount {
		return LangFR, nil
	}

	// Default to English if we can't determine the language
	return LangEN, nil
}

// countWords counts the occurrences of words in a text
func countWords(text string, words []string) int {
	count := 0
	for _, word := range words {
		count += strings.Count(text, " "+word+" ")
	}
	return count
}