爬虫开发

app/Console/Commands/CrawlerHealthCheckCommand.php

@@ -0,0 +1,41 @@
+<?php
+
+declare(strict_types=1);
+
+namespace App\Console\Commands;
+
+use Illuminate\Console\Command;
+use Illuminate\Support\Facades\Http;
+
+class CrawlerHealthCheckCommand extends Command
+{
+    protected $signature = 'crawler:health-check';
+
+    protected $description = '采集器依赖与配置自检';
+
+    public function handle(): int
+    {
+        $checks = [
+            ['item' => 'Queue Connection', 'status' => (string) config('queue.default'), 'detail' => '当前队列连接'],
+            ['item' => 'Browserless Endpoint', 'status' => (string) (config('crawler.browserless_endpoint') ?: 'not-configured'), 'detail' => 'JS渲染服务'],
+            ['item' => 'AI Endpoint', 'status' => (string) (config('crawler.openai_compatible_endpoint') ?: 'not-configured'), 'detail' => 'AI兜底抽取'],
+            ['item' => 'Alert Email', 'status' => (string) (config('crawler.default_alert_email') ?: 'not-configured'), 'detail' => '默认告警邮箱'],
+        ];
+
+        $browserlessEndpoint = (string) config('crawler.browserless_endpoint', '');
+
+        if ($browserlessEndpoint !== '') {
+            try {
+                $response = Http::timeout(5)->get($browserlessEndpoint);
+                $checks[] = ['item' => 'Browserless Reachable', 'status' => $response->status() < 500 ? 'ok' : 'degraded', 'detail' => 'HTTP '.$response->status()];
+            } catch (\Throwable $exception) {
+                $checks[] = ['item' => 'Browserless Reachable', 'status' => 'failed', 'detail' => $exception->getMessage()];
+            }
+        }
+
+        $this->table(['Item', 'Status', 'Detail'], $checks);
+
+        return self::SUCCESS;
+    }
+}
+