CURL问题(多个)

时间:2009-09-07 01:39:28

标签: php multithreading curl libcurl curl-multi

我遇到了curl_multi_*的问题,我想创建一个接收的类/函数,可以说1000个URL,并且一次处理所有这些URL 5,所以当URL完成下载时,它将会将现有的插槽分配给尚未处理的新URL。

我见过curl_multi的some implementations,但是没有一个允许我做我想做的事情,我相信解决方案在于curl_multi_select的使用,但是文档不太清楚,用户注意事项没有多大帮助。

有谁能请我提供一些如何实现这样一个功能的例子?

1 个答案:

答案 0 :(得分:7)

这是一种方法。此脚本将一次获取任意数量的URL,并在每个URL完成时添加一个新的URL(因此它总是获取 $ maxConcurrent 页面)。

$sites = array('http://example.com', 'http://google.com', 'http://stackoverflow.com');
$concurrent = 2;   // Any number.

$mc = new MultiCurl($sites, $concurrent);
$mc->process();

echo '</pre>';

class MultiCurl
{
    private $allToDo;
    private $multiHandle;
    private $maxConcurrent = 2;
    private $currentIndex  = 0;
    private $info          = array();
    private $options       = array(CURLOPT_RETURNTRANSFER => true,
                                   CURLOPT_FOLLOWLOCATION => true,
                                   CURLOPT_MAXREDIRS      => 3,
                                   CURLOPT_TIMEOUT        => 3);

    public function __construct($todo, $concurrent)
    {
        $this->allToDo = $todo;
        $this->maxConcurrent = $concurrent;
        $this->multiHandle = curl_multi_init();
    }

    public function process()
    {
        $running = 0;
        do {
            $this->_addHandles(min(array($this->maxConcurrent - $running, $this->_moreToDo())));
            while ($exec = curl_multi_exec($this->multiHandle, $running) === -1) {
            }
            curl_multi_select($this->multiHandle);
            while ($multiInfo = curl_multi_info_read($this->multiHandle, $msgs)) {
                $this->_showData($multiInfo);
                curl_multi_remove_handle($this->multiHandle, $multiInfo['handle']);
                curl_close($multiInfo['handle']);
            }
        } while ($running || $this->_moreTodo());
        return $this;
    }    

    private function _addHandles($num)
    {
        while ($num-- > 0) {
            $handle = curl_init($this->allToDo[$this->currentIndex]);
            curl_setopt_array($handle, $this->options);
            curl_multi_add_handle($this->multiHandle, $handle);
            $this->info[$handle]['url'] = $this->allToDo[$this->currentIndex];
            $this->currentIndex++;
        }
    }        

    private function _moreToDo()
    {
        return count($this->allToDo) - $this->currentIndex;
    }

    private function _showData($multiInfo)
    {
        $this->info[$multiInfo['handle']]['multi'] = $multiInfo;
        $this->info[$multiInfo['handle']]['curl']  = curl_getinfo($multiInfo['handle']);
        //print_r($this->info[$multiInfo['handle']]);
        $content = curl_multi_getcontent($multiInfo['handle']);
        echo $this->info[$multiInfo['handle']]['url'] . ' - ' . strlen($content) . ' bytes<br />';
        //echo htmlspecialchars($content);
    }
}
相关问题