No. Guzzle can use any HTTP handler to send requests. This means that Guzzlecan be used with cURL, PHP's stream wrapper, sockets, and non-blocking librarieslikeReact. You just need to configure an HTTP handlerto use a different method of sending requests.
Note
Guzzle has historically only utilized cURL to send HTTP requests. cURL isan amazing HTTP client (arguably the best), and Guzzle will continue to useit by default when it is available. It is rare, but some developers don'thave cURL installed on their systems or run into version specific issues.By allowing swappable HTTP handlers, Guzzle is now much more customizableand able to adapt to fit the needs of more developers.
Yes. You can use therequestAsync,sendAsync,getAsync,headAsync,putAsync,postAsync,deleteAsync, andpatchAsyncmethods of a client to send an asynchronous request. The client will return aGuzzleHttp\Promise\PromiseInterface object. You can chainthenfunctions off of the promise.
$promise=$client->requestAsync('GET','http://httpbin.org/get');$promise->then(function($response){echo'Got a response! '.$response->getStatusCode();});
You can force an asynchronous response to complete using thewait() methodof the returned promise.
$promise=$client->requestAsync('GET','http://httpbin.org/get');$response=$promise->wait();
cURL offers a huge number ofcustomizable options.While Guzzle normalizes many of these options across different handlers, thereare times when you need to set custom cURL options. This can be accomplishedby passing an associative array of cURL settings in thecurl key of arequest.
For example, let's say you need to customize the outgoing network interfaceused with a client.
$client->request('GET','/',['curl'=>[CURLOPT_INTERFACE=>'xxx.xxx.xxx.xxx']]);
If you use asynchronous requests with cURL multi handler and want to tweak it,additional options can be specified as an associative array in theoptions key of theCurlMultiHandler constructor.
useGuzzleHttp\Client;useGuzzleHttp\HandlerStack;useGuzzleHttp\Handler\CurlMultiHandler;$client=newClient(['handler'=>HandlerStack::create(newCurlMultiHandler(['options'=>[CURLMOPT_MAX_TOTAL_CONNECTIONS=>50,CURLMOPT_MAX_HOST_CONNECTIONS=>5,]]))]);
You can pass customstream context optionsusing thestream_context key of the request option. Thestream_contextarray is an associative array where each key is a PHP transport, and each valueis an associative array of transport options.
For example, let's say you need to customize the outgoing network interfaceused with a client and allow self-signed certificates.
$client->request('GET','/',['stream'=>true,'stream_context'=>['ssl'=>['allow_self_signed'=>true],'socket'=>['bindto'=>'xxx.xxx.xxx.xxx']]]);
You need to specify the path on disk to the CA bundle used by Guzzle forverifying the peer certificate. Seeverify.
Maximum function nesting level of '100' reached, aborting
You could run into this error if you have the XDebug extension installed andyou execute a lot of requests in callbacks. This error message comesspecifically from the XDebug extension. PHP itself does not have a functionnesting limit. Change this setting in your php.ini to increase the limit:
xdebug.max_nesting_level=1000
This can occur for a number of reasons, but if you are sending PUT, POST, orPATCH requests with anExpect:100-Continue header, a server that does notsupport this header will return a 417 response. You can work around this bysetting theexpect request option tofalse:
$client=newGuzzleHttp\Client();// Disable the expect header on a single request$response=$client->request('PUT','/',['expect'=>false]);// Disable the expect header on all client requests$client=newGuzzleHttp\Client(['expect'=>false]);
You can enable tracking of redirected URIs and status codes via thetrack_redirects option. Each redirected URI and status code will be stored in theX-Guzzle-Redirect-History and theX-Guzzle-Redirect-Status-Historyheader respectively.
The initial request's URI and the final status code will be excluded from the results.With this in mind you should be able to easily track a request's full redirect path.
For example, let's say you need to track redirects and provide both resultstogether in a single report:
// First you configure Guzzle with redirect tracking and make a request$client=newClient([RequestOptions::ALLOW_REDIRECTS=>['max'=>10,// allow at most 10 redirects.'strict'=>true,// use "strict" RFC compliant redirects.'referer'=>true,// add a Referer header'track_redirects'=>true,],]);$initialRequest='/redirect/3';// Store the request URI for later use$response=$client->request('GET',$initialRequest);// Make your request// Retrieve both Redirect History headers$redirectUriHistory=$response->getHeader('X-Guzzle-Redirect-History')[0];// retrieve Redirect URI history$redirectCodeHistory=$response->getHeader('X-Guzzle-Redirect-Status-History')[0];// retrieve Redirect HTTP Status history// Add the initial URI requested to the (beginning of) URI historyarray_unshift($redirectUriHistory,$initialRequest);// Add the final HTTP status code to the end of HTTP response historyarray_push($redirectCodeHistory,$response->getStatusCode());// (Optional) Combine the items of each array into a single result set$fullRedirectReport=[];foreach($redirectUriHistoryas$key=>$value){$fullRedirectReport[$key]=['location'=>$value,'code'=>$redirectCodeHistory[$key]];}echojson_encode($fullRedirectReport);