專案定位

一個極簡的 PHP 單檔采集腳本，適用於靜態/偽靜態頁面、結構固定、無反爬、小規模一次性采集任務。核心流程：HTTP GET 抓取 HTML → 正則提取 <title> 與內容區塊 (指定開始/結束標記) → 簡易清洗 (去標籤、實體解碼) → 儲存靜態檔案。

完整程式碼 (重構增強版)

<?php
/**
 * 簡易文章采集器
 * 配置區塊請根據目標站點修改
 */
class SimpleArticleScraper {
    // ===== 配置 =====
    private $targetUrl = 'https://example.com/article/123.html';
    private $startMarker = '
'; // 內容區塊起始特徵
    private $endMarker   = '
';  // 內容區塊結束特徵
    private $outputDir   = __DIR__ . '/data/articles/';
    private $userAgent   = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36';
    private $timeout     = 30;

    public function run() {
        // 1. 抓取
        $html = $this->fetch($this->targetUrl);
        if (!$html) return $this->log("抓取失敗: {$this->targetUrl}");

        // 2. 提取標題
        $title = $this->extractTitle($html);
        if (!$title) return $this->log("標題提取失敗");

        // 3. 提取內容
        $content = $this->extractContent($html);
        if (!$content) return $this->log("內容提取失敗 (檢查 Marker)");

        // 4. 清洗
        $cleanContent = $this->cleanHtml($content);

        // 5. 儲存
        $filename = $this->sanitizeFilename($title) . '.html';
        $saved = $this->saveFile($filename, $cleanContent);
        
        if ($saved) {
            $this->log("✅ 采集成功: {$title} -> {$filename}");
        } else {
            $this->log("❌ 儲存失敗");
        }
    }

    private function fetch($url) {
        $ch = curl_init($url);
        curl_setopt_array($ch, [
            CURLOPT_RETURNTRANSFER => 1,
            CURLOPT_CONNECTTIMEOUT => 10,
            CURLOPT_TIMEOUT => $this->timeout,
            CURLOPT_USERAGENT => $this->userAgent,
            CURLOPT_SSL_VERIFYPEER => false,
            CURLOPT_FOLLOWLOCATION => true,
            CURLOPT_ENCODING => 'gzip,deflate',
        ]);
        $html = curl_exec($ch);
        $errno = curl_errno($ch);
        curl_close($ch);
        if ($errno) {
            $this->log("cURL Error [$errno]: " . curl_strerror($errno));
            return false;
        }
        // 編碼統一
        if (mb_detect_encoding($html, 'GBK,UTF-8,BIG5') !== 'UTF-8') {
            $html = mb_convert_encoding($html, 'UTF-8', mb_detect_encoding($html, 'GBK,UTF-8,BIG5'));
        }
        return $html;
    }

    private function extractTitle($html) {
        if (preg_match('/]*>(.*?)<\/title>/is', $html, $m)) {
            return trim(strip_tags($m[1]));
        }
        // 備用：h1
        if (preg_match('/]*>(.*?)<\/h1>/is', $html, $m)) {
            return trim(strip_tags($m[1]));
        }
        return false;
    }

    private function extractContent($html) {
        // 策略：找到 startMarker 後的內容，直到 endMarker
        $startPos = strpos($html, $this->startMarker);
        if ($startPos === false) return false;
        $startPos += strlen($this->startMarker);
        
        $endPos = strpos($html, $this->endMarker, $startPos);
        if ($endPos === false) {
            // 無結束標記，取剩餘或固定長度
            return substr($html, $startPos, 50000);
        }
        return substr($html, $startPos, $endPos - $startPos);
    }

    private function cleanHtml($html) {
        // 1. 移除 script, style, iframe, noscript 等
        $html = preg_replace('/<(script|style|iframe|noscript)[^>]*>.*?<\/\1>/is', '', $html);
        // 2. 移除所有標籤，保留文字 (或保留 p, br, img, a 等)
        // 此處示範「保留基本排版標籤」
        $allowed = '

';
        $html = strip_tags($html, $allowed);
        // 3. 解碼實體
        $html = html_entity_decode($html, ENT_QUOTES | ENT_HTML5, 'UTF-8');
        // 4. 修正圖片連結為絕對路徑 (簡易)
        $html = preg_replace_callback('/]+src=["\']([^"\']+)["\']/i', function($m) {
            $src = $m[1];
            if (strpos($src, 'http') !== 0) {
                $src = $this->resolveUrl($src);
            }
            return str_replace($m[1], $src, $m[0]);
        }, $html);
        // 5. 壓縮空白
        $html = preg_replace('/\s+/', ' ', $html);
        return trim($html);
    }

    private function resolveUrl($relative) {
        $base = parse_url($this->targetUrl);
        $baseUrl = $base['scheme'] . '://' . $base['host'] . (isset($base['port']) ? ':' . $base['port'] : '');
        $path = dirname($base['path']);
        if ($relative[0] === '/') return $baseUrl . $relative;
        return $baseUrl . $path . '/' . $relative;
    }

    private function sanitizeFilename($title) {
        $filename = preg_replace('/[\\\\\/:*?"<>|]/', '_', $title); // 非法字元
        $filename = mb_substr($filename, 0, 100, 'UTF-8'); // 限制長度
        return $filename ?: 'article_' . time();
    }

    private function saveFile($filename, $content) {
        if (!is_dir($this->outputDir)) {
            mkdir($this->outputDir, 0755, true);
        }
        $path = $this->outputDir . $filename;
        // 避免覆蓋
        $counter = 1;
        $origPath = $path;
        while (file_exists($path)) {
            $info = pathinfo($origPath);
            $path = $info['dirname'] . '/' . $info['filename'] . "_{$counter}." . $info['extension'];
            $counter++;
        }
        return file_put_contents($path, $content, LOCK_EX) !== false;
    }

    private function log($msg) {
        echo "[" . date('H:i:s') . "] {$msg}\n";
    }
}

// ===== 執行 =====
// 可從 CLI 執行: php scraper.php
// 或 Web 觸發 (需注意超時)
$scraper = new SimpleArticleScraper();
// $scraper->targetUrl = $_GET['url'] ?? $scraper->targetUrl; // Web 模式需驗證 URL 安全性
$scraper->run();
?>

flowchart TD
    A[開始] --> B{抓取 HTML}
    B -- 失敗 --> C[記錄錯誤/重試] --> Z[結束]
    B -- 成功 --> D[提取 Title]
    D --> E[定位內容區塊
startMarker ~ endMarker]
    E -- 失敗 --> C
    E -- 成功 --> F[HTML 清洗
去標籤/實體/絕對化圖片]
    F --> G[生成安全檔名]
    G --> H[寫入檔案]
    H --> I[成功/失敗日誌]
    I --> Z