backend/app/Command/spider/BaseSpider.php

<?php

namespace App\Command\spider;

use App\Model\AppArticle;
use App\Model\AppSpiderArticle;
use Hyperf\Command\Command;
use Hyperf\Contract\StdoutLoggerInterface;
use Hyperf\Coroutine\Coroutine;
use Hyperf\Di\Annotation\Inject;
use Laminas\Stdlib\ArrayUtils;
use Swoole\Coroutine\Channel;
use Swoole\Timer;
use Symfony\Component\Console\Input\InputInterface;
use Symfony\Component\Console\Input\InputOption;
use Symfony\Component\Console\Output\OutputInterface;
use function Hyperf\Coroutine\co;

class BaseSpider extends Command
{
    /**
     * 最大协程数量
     * @var int
     */
    protected int $maxCo = 10;

    protected ?\Swoole\Coroutine\Channel $channel = null;

    /**
     * @var string
     */
    protected string $baseUrl = '';

    #[Inject]
    protected ?StdoutLoggerInterface $logger = null;

    protected array $coroutineList = [];

    protected const PLATFORM = '';

    private bool $isInit = false;

    protected int|bool $timer = false;

    protected array $commandConfigure = [];

    private function init()
    {
        // 因为最外层还有个父协程, 所以加一
        $this->channel = new Channel($this->maxCo + 1);
        $this->timer = Timer::tick(1000 * 30, function () use (&$coList) {
//            count(\Swoole\Coroutine::getElapsed());
            var_dump(count($this->coroutineList));
//            var_dump($list);
        });
        for ($i = 0; $i < $this->maxCo + 1; $i++) {
            $this->channel->push(1);
        }
    }

    public function configure()
    {
        parent::configure();
        $this->addOption('prod', '', InputOption::VALUE_NEGATABLE, '是否关闭devMode.', false);
    }

    public static function getPlatform(): string
    {
        return static::PLATFORM;
    }

    public function getBaseUrl(): string
    {
        return rtrim($this->baseUrl, '/');
    }

    protected function getArticleModel(array $condition)
    {
        return AppSpiderArticle::query()->where($condition)->first() ?: new AppSpiderArticle();
    }

    protected function request(string $url): array
    {
        $ch = curl_init();
        curl_setopt_array($ch, array(
            CURLOPT_URL => $url,
            CURLOPT_RETURNTRANSFER => true,
            CURLOPT_ENCODING => '',
            CURLOPT_MAXREDIRS => 10,
            CURLOPT_TIMEOUT => 15,
            CURLOPT_FOLLOWLOCATION => true,
            CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
            CURLOPT_CUSTOMREQUEST => 'GET',
        ));
//        curl_setopt($ch, CURLOPT_URL, $url);
//        curl_setopt($ch, CURLOPT_HEADER, false);
//        curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
//        curl_setopt($ch, CURLOPT_TIMEOUT, 10);
        $result = curl_exec($ch);
        curl_close($ch);
        $httpCode = curl_getinfo($ch,CURLINFO_HTTP_CODE);
        return [$result, $httpCode];
    }

    protected function returnPool()
    {
        return $this->channel->push(1);
    }

    protected function getPool(): bool
    {
        return $this->channel->pop();
    }

    protected function createCoroutine(\Closure $func): void
    {
        if ($this->isInit === false) {
            $this->isInit = true;
            $this->init();
        }
        $this->getPool();
        $cid = co(function () use ($func) {
            \Co\defer(function() {
                unset($this->coroutineList[Coroutine::id()]);
                $this->returnPool();
            });

            $func();
        });
        $this->coroutineList[$cid] = 1;
    }

    protected function debugPrint(array|string $message = '', $level = 0)
    {
        if ($this->getCommandConfigure('prod') === false) {
            $printTime = date('H:i:s');
            echo "[spider-debug][$printTime]" . print_r($message, true) . PHP_EOL;
        }
    }

    /**
     * 用于单元测试
     * @param string $methodName
     * @param $args
     * @return mixed
     */
    public function testMethod(string $methodName, $args = [])
    {
        return $this->{$methodName}(...$args);
    }

    public function setCommandConfigure($options): void
    {
        $this->commandConfigure = $options;
    }

    public function getCommandConfigure($key = null, $defaultValue = null)
    {
        if (!$key) {
            return $this->commandConfigure;
        }

        return $this->commandConfigure[$key] ?? $defaultValue;
    }

    public function execute(InputInterface $input, OutputInterface $output): int
    {
        $this->setCommandConfigure($input->getOptions());
        return 0;
    }
}