CodePress/engine/core/class/SearchEngine.php

<?php

class SearchEngine {
    private array $index = [];
    private CacheInterface $cache;

    public function __construct(?CacheInterface $cache = null) {
        $this->cache = $cache ?? new FileCache();
        $this->loadIndex();
    }

    public function indexContent(string $path, string $content, array $metadata = []): void {
        $words = $this->tokenize($content);
        $pathHash = md5($path);

        foreach ($words as $word) {
            if (!isset($this->index[$word])) {
                $this->index[$word] = [];
            }
            if (!in_array($pathHash, $this->index[$word])) {
                $this->index[$word][] = $pathHash;
            }
        }

        // Store metadata for this path
        $this->cache->set('search_meta_' . $pathHash, [
            'path' => $path,
            'title' => $metadata['title'] ?? basename($path),
            'snippet' => $this->generateSnippet($content),
            'last_modified' => $metadata['modified'] ?? time()
        ], 86400); // 24 hours

        $this->saveIndex();
    }

    public function search(string $query, int $limit = 20): array {
        $terms = $this->tokenize($query);
        $results = [];
        $pathScores = [];

        foreach ($terms as $term) {
            if (isset($this->index[$term])) {
                foreach ($this->index[$term] as $pathHash) {
                    if (!isset($pathScores[$pathHash])) {
                        $pathScores[$pathHash] = 0;
                    }
                    $pathScores[$pathHash]++;
                }
            }
        }

        // Sort by relevance (term frequency)
        arsort($pathScores);

        // Get top results
        $count = 0;
        foreach ($pathScores as $pathHash => $score) {
            if ($count >= $limit) break;

            $metadata = $this->cache->get('search_meta_' . $pathHash);
            if ($metadata) {
                $results[] = array_merge($metadata, ['score' => $score]);
                $count++;
            }
        }

        return $results;
    }

    public function removeFromIndex(string $path): void {
        $pathHash = md5($path);

        foreach ($this->index as $word => $paths) {
            $this->index[$word] = array_filter($paths, fn($hash) => $hash !== $pathHash);
            if (empty($this->index[$word])) {
                unset($this->index[$word]);
            }
        }

        $this->cache->delete('search_meta_' . $pathHash);
        $this->saveIndex();
    }

    public function clearIndex(): void {
        $this->index = [];
        $this->cache->clear();
        $this->saveIndex();
    }

    private function tokenize(string $text): array {
        // Convert to lowercase, remove punctuation, split into words
        $text = strtolower($text);
        $text = preg_replace('/[^\w\s]/u', ' ', $text);
        $words = preg_split('/\s+/u', $text, -1, PREG_SPLIT_NO_EMPTY);

        // Filter out common stop words and short words
        $stopWords = ['the', 'a', 'an', 'and', 'or', 'but', 'in', 'on', 'at', 'to', 'for', 'of', 'with', 'by', 'is', 'are', 'was', 'were', 'be', 'been', 'being', 'have', 'has', 'had', 'do', 'does', 'did', 'will', 'would', 'could', 'should', 'may', 'might', 'must', 'can'];
        $words = array_filter($words, function($word) use ($stopWords) {
            return strlen($word) > 2 && !in_array($word, $stopWords);
        });

        return array_unique($words);
    }

    private function generateSnippet(string $content, int $length = 150): string {
        // Remove HTML tags and extra whitespace
        $content = strip_tags($content);
        $content = preg_replace('/\s+/', ' ', $content);

        if (strlen($content) <= $length) {
            return $content;
        }

        return substr($content, 0, $length) . '...';
    }

    private function loadIndex(): void {
        $cached = $this->cache->get('search_index');
        if ($cached) {
            $this->index = $cached;
        }
    }

    private function saveIndex(): void {
        $this->cache->set('search_index', $this->index, 86400); // 24 hours
    }
}