我正在做一个简单的应用程序，从15个不同的网址读取json数据。我有一个特殊需求，需要在服务器端执行此操作。我使用file_get_contents($url)。
由于我正在使用file_get_contents($url)，所以我编写了一个简单的脚本，如下：
$websites = array(
$url1,
$url2,
$url3,
...
$url15
);
foreach ($websites as $website) {
$data[] = file_get_contents($website);
}
证明它非常慢，因为它等待第一个请求，然后才执行下一个请求。
do {curl_multi_exec($master,$running);if($running>0){curl_multi_select($mh,1);}} while($running > 0);那么它将只使用约1%的CPU资源，而不是100%的CPU资源（一个更好的循环仍然可以构建，这将是更好的for(;;){curl_multi_exec($mh,$running);if($running<1)break;curl_multi_select($mh,1);}）。 - hanshenrik我不太喜欢现有答案的任何方法。
Timo的代码：可能会在CURLM_CALL_MULTI_PERFORM期间进行sleep/select()，这是错误的。当($still_running > 0 && $exec != CURLM_CALL_MULTI_PERFORM)时，它也可能无法进行睡眠，这可能会使代码以100％的CPU使用率（1个核心）旋转而没有任何原因。
Sudhir的代码：当$still_running > 0时不会进行睡眠，并且spam-call 异步函数curl_multi_exec()直到所有内容都被下载完成，这会导致PHP在所有内容被下载完成之前使用100％的CPU（1个CPU核心），换句话说，它在下载时未能进行睡眠。
这里提供了一种没有上述问题的方法：
$websites = array(
"http://google.com",
"http://example.org"
// $url2,
// $url3,
// ...
// $url15
);
$mh = curl_multi_init();
foreach ($websites as $website) {
$worker = curl_init($website);
curl_setopt_array($worker, [
CURLOPT_RETURNTRANSFER => 1
]);
curl_multi_add_handle($mh, $worker);
}
for (;;) {
$still_running = null;
do {
$err = curl_multi_exec($mh, $still_running);
} while ($err === CURLM_CALL_MULTI_PERFORM);
if ($err !== CURLM_OK) {
// handle curl multi error?
}
if ($still_running < 1) {
// all downloads completed
break;
}
// some haven't finished downloading, sleep until more data arrives:
curl_multi_select($mh, 1);
}
$results = [];
while (false !== ($info = curl_multi_info_read($mh))) {
if ($info["result"] !== CURLE_OK) {
// handle download error?
}
$results[curl_getinfo($info["handle"], CURLINFO_EFFECTIVE_URL)] = curl_multi_getcontent($info["handle"]);
curl_multi_remove_handle($mh, $info["handle"]);
curl_close($info["handle"]);
}
curl_multi_close($mh);
var_export($results);
$websites = array(
"http://google.com",
"http://example.org"
// $url2,
// $url3,
// ...
// $url15
);
var_dump(fetch_urls($websites,50));
function fetch_urls(array $urls, int $max_connections, int $timeout_ms = 10000, bool $return_fault_reason = true): array
{
if ($max_connections < 1) {
throw new InvalidArgumentException("max_connections MUST be >=1");
}
foreach ($urls as $key => $foo) {
if (! is_string($foo)) {
throw new \InvalidArgumentException("all urls must be strings!");
}
if (empty($foo)) {
unset($urls[$key]); // ?
}
}
unset($foo);
// DISABLED for benchmarking purposes: $urls = array_unique($urls); // remove duplicates.
$ret = array();
$mh = curl_multi_init();
$workers = array();
$work = function () use (&$ret, &$workers, &$mh, $return_fault_reason) {
// > If an added handle fails very quickly, it may never be counted as a running_handle
while (1) {
do {
$err = curl_multi_exec($mh, $still_running);
} while ($err === CURLM_CALL_MULTI_PERFORM);
if ($still_running < count($workers)) {
// some workers finished, fetch their response and close them
break;
}
$cms = curl_multi_select($mh, 1);
// var_dump('sr: ' . $still_running . " c: " . count($workers)." cms: ".$cms);
}
while (false !== ($info = curl_multi_info_read($mh))) {
// echo "NOT FALSE!";
// var_dump($info);
{
if ($info['msg'] !== CURLMSG_DONE) {
continue;
}
if ($info['result'] !== CURLE_OK) {
if ($return_fault_reason) {
$ret[$workers[(int) $info['handle']]] = print_r(array(
false,
$info['result'],
"curl_exec error " . $info['result'] . ": " . curl_strerror($info['result'])
), true);
}
} elseif (CURLE_OK !== ($err = curl_errno($info['handle']))) {
if ($return_fault_reason) {
$ret[$workers[(int) $info['handle']]] = print_r(array(
false,
$err,
"curl error " . $err . ": " . curl_strerror($err)
), true);
}
} else {
$ret[$workers[(int) $info['handle']]] = curl_multi_getcontent($info['handle']);
}
curl_multi_remove_handle($mh, $info['handle']);
assert(isset($workers[(int) $info['handle']]));
unset($workers[(int) $info['handle']]);
curl_close($info['handle']);
}
}
// echo "NO MORE INFO!";
};
foreach ($urls as $url) {
while (count($workers) >= $max_connections) {
// echo "TOO MANY WORKERS!\n";
$work();
}
$neww = curl_init($url);
if (! $neww) {
trigger_error("curl_init() failed! probably means that max_connections is too high and you ran out of system resources", E_USER_WARNING);
if ($return_fault_reason) {
$ret[$url] = array(
false,
- 1,
"curl_init() failed"
);
}
continue;
}
$workers[(int) $neww] = $url;
curl_setopt_array($neww, array(
CURLOPT_RETURNTRANSFER => 1,
CURLOPT_SSL_VERIFYHOST => 0,
CURLOPT_SSL_VERIFYPEER => 0,
CURLOPT_TIMEOUT_MS => $timeout_ms
));
curl_multi_add_handle($mh, $neww);
// curl_multi_exec($mh, $unused_here); LIKELY TO BE MUCH SLOWER IF DONE IN THIS LOOP: TOO MANY SYSCALLS
}
while (count($workers) > 0) {
// echo "WAITING FOR WORKERS TO BECOME 0!";
// var_dump(count($workers));
$work();
}
curl_multi_close($mh);
return $ret;
}
这将下载整个列表，并且不会同时下载超过50个URL（但即使使用这种方法，所有结果也会存储在RAM中，因此即使使用这种方法也可能会耗尽RAM；如果想将其存储在数据库中而不是内存中，则可以修改curl_multi_getcontent部分以将其存储在数据库中而不是存储在持久的RAM变量中）。
$return_fault_reason代表什么意思？ - Ali Niaz$return_fault_reason 是一个参数，用于告知下载失败时是应该被忽略还是需要附带错误信息；我已经更新了代码，加入了 $return_fault_reason 参数。 - hanshenrikdo {
curl_multi_exec($multiHandle, $stillRunning);
// just block the loop until there's activity on any curl_multi connection
curl_multi_select($multiHandle, 1);
} while ($stillRunning > 0);
function mtime(){
return microtime(true);
}
function ptime($prev){
$t = microtime(true) - $prev;
$t = $t * 1000;
return str_pad($t, 20, 0, STR_PAD_RIGHT);
}
// This function exists to add compatibility for CURLM_CALL_MULTI_PERFORM for old curl versions, on modern curl it will only run once and be the equivalent of calling curl_multi_exec
function curl_multi_exec_full($mh, &$still_running) {
// In theory curl_multi_exec should never return CURLM_CALL_MULTI_PERFORM (-1) because it has been deprecated
// In practice it sometimes does
// So imagine that this just runs curl_multi_exec once and returns it's value
do {
$state = curl_multi_exec($mh, $still_running);
// curl_multi_select($mh, $timeout) simply blocks for $timeout seconds while curl_multi_exec() returns CURLM_CALL_MULTI_PERFORM
// We add it to prevent CPU 100% usage in case this thing misbehaves (especially for old curl on windows)
} while ($still_running > 0 && $state === CURLM_CALL_MULTI_PERFORM && curl_multi_select($mh, 0.1));
return $state;
}
// This function replaces curl_multi_select and makes the name make more sense, since all we're doing is waiting for curl, it also forces a minimum sleep time between requests to avoid excessive CPU usage.
function curl_multi_wait($mh, $minTime = 0.001, $maxTime = 1){
$umin = $minTime*1000000;
$start_time = microtime(true);
// it sleeps until there is some activity on any of the descriptors (curl files)
// it returns the number of descriptors (curl files that can have activity)
$num_descriptors = curl_multi_select($mh, $maxTime);
// if the system returns -1, it means that the wait time is unknown, and we have to decide the minimum time to wait
// but our `$timespan` check below catches this edge case, so this `if` isn't really necessary
if($num_descriptors === -1){
usleep($umin);
}
$timespan = (microtime(true) - $start_time);
// This thing runs very fast, up to 1000 times for 2 urls, which wastes a lot of CPU
// This will reduce the runs so that each interval is separated by at least minTime
if($timespan < $umin){
usleep($umin - $timespan);
//print "sleep for ".($umin - $timeDiff).PHP_EOL;
}
}
$handles = [
[
CURLOPT_URL=>"http://example.com/",
CURLOPT_HEADER=>false,
CURLOPT_RETURNTRANSFER=>true,
CURLOPT_FOLLOWLOCATION=>false,
],
[
CURLOPT_URL=>"http://www.php.net",
CURLOPT_HEADER=>false,
CURLOPT_RETURNTRANSFER=>true,
CURLOPT_FOLLOWLOCATION=>false,
// this function is called by curl for each header received
// This complies with RFC822 and RFC2616, please do not suggest edits to make use of the mb_ string functions, it is incorrect!
// https://dev59.com/3mox5IYBdhLWcg3wekIr#41135574
CURLOPT_HEADERFUNCTION=>function($ch, $header)
{
print "header from http://www.php.net: ".$header;
//$header = explode(':', $header, 2);
//if (count($header) < 2){ // ignore invalid headers
// return $len;
//}
//$headers[strtolower(trim($header[0]))][] = trim($header[1]);
return strlen($header);
}
]
];
//create the multiple cURL handle
$mh = curl_multi_init();
$chandles = [];
foreach($handles as $opts) {
// create cURL resources
$ch = curl_init();
// set URL and other appropriate options
curl_setopt_array($ch, $opts);
// add the handle
curl_multi_add_handle($mh, $ch);
$chandles[] = $ch;
}
//execute the multi handle
$prevRunning = null;
$count = 0;
do {
$time = mtime();
// $running contains the number of currently running requests
$status = curl_multi_exec_full($mh, $running);
$count++;
print ptime($time).": curl_multi_exec status=$status running $running".PHP_EOL;
// One less is running, meaning one has finished
if($running < $prevRunning){
print ptime($time).": curl_multi_info_read".PHP_EOL;
// msg: The CURLMSG_DONE constant. Other return values are currently not available.
// result: One of the CURLE_* constants. If everything is OK, the CURLE_OK will be the result.
// handle: Resource of type curl indicates the handle which it concerns.
while ($read = curl_multi_info_read($mh, $msgs_in_queue)) {
$info = curl_getinfo($read['handle']);
if($read['result'] !== CURLE_OK){
// handle the error somehow
print "Error: ".$info['url'].PHP_EOL;
}
if($read['result'] === CURLE_OK){
/*
// This will automatically follow the redirect and still give you control over the previous page
// TODO: max redirect checks and redirect timeouts
if(isset($info['redirect_url']) && trim($info['redirect_url'])!==''){
print "running redirect: ".$info['redirect_url'].PHP_EOL;
$ch3 = curl_init();
curl_setopt($ch3, CURLOPT_URL, $info['redirect_url']);
curl_setopt($ch3, CURLOPT_HEADER, 0);
curl_setopt($ch3, CURLOPT_RETURNTRANSFER, 1);
curl_setopt($ch3, CURLOPT_FOLLOWLOCATION, 0);
curl_multi_add_handle($mh,$ch3);
}
*/
print_r($info);
$body = curl_multi_getcontent($read['handle']);
print $body;
}
}
}
// Still running? keep waiting...
if ($running > 0) {
curl_multi_wait($mh);
}
$prevRunning = $running;
} while ($running > 0 && $status == CURLM_OK);
//close the handles
foreach($chandles as $ch){
curl_multi_remove_handle($mh, $ch);
}
curl_multi_close($mh);
print $count.PHP_EOL;
CURLM_CALL_MULTI_PERFORM（因此是 CCMP）有误，如果你得到 CCMP，就不应该运行 select()，而应该再次调用 multi_exec()。更糟糕的是，从 2012 年左右开始，curl 不再返回 CCMP，所以你的 $state === CCMP 检查将始终失败，这意味着你的执行循环将在第一次迭代后总是退出。 - hanshenrikcurl_multi_wait函数内部，注意它在代码下面逐个计算进程完成情况，这意味着我们不关心curl_multi_exec_full是否在一个循环中完成或运行选择，新版curl不会这样。 - Timo Huovinencurl_multi_select 功能，它会像睡眠一样起作用。如果将其移除，我稍微有些担心会降低代码的“鲁棒性”，但也可以接受。 - Timo Huovinencurl_multi_exec_full 中保留 curl_multi_select 对 CCMP 有什么坏处吗？ - Timo Huovinen