mirror of
https://git.sekbaer.de/Friendica/friendica.git
synced 2025-06-08 00:04:27 +02:00
Create HTTPClientFactory and introduce ImageTest
This commit is contained in:
Philipp
parent
73e8db24f9
commit
52c7948526
7 changed files with 278 additions and 134 deletions
421
src/Network/HTTPClient.php
Normal file
421
src/Network/HTTPClient.php
Normal file
|
|
@ -0,0 +1,421 @@
|
|||
<?php
|
||||
/**
|
||||
* @copyright Copyright (C) 2010-2021, the Friendica project
|
||||
*
|
||||
* @license GNU APGL version 3 or any later version
|
||||
*
|
||||
* This program is free software: you can redistribute it and/or modify
|
||||
* it under the terms of the GNU Affero General Public License as
|
||||
* published by the Free Software Foundation, either version 3 of the
|
||||
* License, or (at your option) any later version.
|
||||
*
|
||||
* This program is distributed in the hope that it will be useful,
|
||||
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
||||
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
||||
* GNU Affero General Public License for more details.
|
||||
*
|
||||
* You should have received a copy of the GNU Affero General Public License
|
||||
* along with this program. If not, see <https://www.gnu.org/licenses/>.
|
||||
*
|
||||
*/
|
||||
|
||||
namespace Friendica\Network;
|
||||
|
||||
use DOMDocument;
|
||||
use DomXPath;
|
||||
use Friendica\Core\Config\IConfig;
|
||||
use Friendica\Core\System;
|
||||
use Friendica\Util\Network;
|
||||
use Friendica\Util\Profiler;
|
||||
use GuzzleHttp\Client;
|
||||
use GuzzleHttp\Cookie\FileCookieJar;
|
||||
use GuzzleHttp\Exception\RequestException;
|
||||
use GuzzleHttp\Exception\TransferException;
|
||||
use GuzzleHttp\RequestOptions;
|
||||
use Psr\Http\Message\ResponseInterface;
|
||||
use Psr\Log\LoggerInterface;
|
||||
|
||||
/**
|
||||
* Performs HTTP requests to a given URL
|
||||
*/
|
||||
class HTTPClient implements IHTTPClient
|
||||
{
|
||||
/** @var LoggerInterface */
|
||||
private $logger;
|
||||
/** @var Profiler */
|
||||
private $profiler;
|
||||
/** @var IConfig */
|
||||
private $config;
|
||||
/** @var string */
|
||||
private $userAgent;
|
||||
/** @var Client */
|
||||
private $client;
|
||||
|
||||
public function __construct(LoggerInterface $logger, Profiler $profiler, IConfig $config, string $userAgent, Client $client)
|
||||
{
|
||||
$this->logger = $logger;
|
||||
$this->profiler = $profiler;
|
||||
$this->config = $config;
|
||||
$this->userAgent = $userAgent;
|
||||
$this->client = $client;
|
||||
}
|
||||
|
||||
protected function request(string $method, string $url, array $opts = [])
|
||||
{
|
||||
$this->profiler->startRecording('network');
|
||||
|
||||
if (Network::isLocalLink($url)) {
|
||||
$this->logger->info('Local link', ['url' => $url, 'callstack' => System::callstack(20)]);
|
||||
}
|
||||
|
||||
if (strlen($url) > 1000) {
|
||||
$this->logger->debug('URL is longer than 1000 characters.', ['url' => $url, 'callstack' => System::callstack(20)]);
|
||||
$this->profiler->stopRecording();
|
||||
return CurlResult::createErrorCurl(substr($url, 0, 200));
|
||||
}
|
||||
|
||||
$parts2 = [];
|
||||
$parts = parse_url($url);
|
||||
$path_parts = explode('/', $parts['path'] ?? '');
|
||||
foreach ($path_parts as $part) {
|
||||
if (strlen($part) <> mb_strlen($part)) {
|
||||
$parts2[] = rawurlencode($part);
|
||||
} else {
|
||||
$parts2[] = $part;
|
||||
}
|
||||
}
|
||||
$parts['path'] = implode('/', $parts2);
|
||||
$url = Network::unparseURL($parts);
|
||||
|
||||
if (Network::isUrlBlocked($url)) {
|
||||
$this->logger->info('Domain is blocked.', ['url' => $url]);
|
||||
$this->profiler->stopRecording();
|
||||
return CurlResult::createErrorCurl($url);
|
||||
}
|
||||
|
||||
$conf = [];
|
||||
|
||||
if (!empty($opts['cookiejar'])) {
|
||||
$jar = new FileCookieJar($opts['cookiejar']);
|
||||
$conf[RequestOptions::COOKIES] = $jar;
|
||||
}
|
||||
|
||||
if (!empty($opts['accept_content'])) {
|
||||
array_push($curlOptions[CURLOPT_HTTPHEADER], 'Accept: ' . $opts['accept_content']);
|
||||
}
|
||||
|
||||
if (!empty($opts['header'])) {
|
||||
$curlOptions[CURLOPT_HTTPHEADER] = array_merge($opts['header'], $curlOptions[CURLOPT_HTTPHEADER]);
|
||||
}
|
||||
|
||||
$curlOptions[CURLOPT_USERAGENT] = $this->userAgent;
|
||||
|
||||
if (!empty($opts['headers'])) {
|
||||
$this->logger->notice('Wrong option \'headers\' used.');
|
||||
$curlOptions[CURLOPT_HTTPHEADER] = array_merge($opts['headers'], $curlOptions[CURLOPT_HTTPHEADER]);
|
||||
}
|
||||
|
||||
if (!empty($opts['timeout'])) {
|
||||
$curlOptions[CURLOPT_TIMEOUT] = $opts['timeout'];
|
||||
}
|
||||
|
||||
$onHeaders = function (ResponseInterface $response) use ($opts) {
|
||||
if (!empty($opts['content_length']) &&
|
||||
$response->getHeaderLine('Content-Length') > $opts['content_length']) {
|
||||
throw new TransferException('The file is too big!');
|
||||
}
|
||||
};
|
||||
|
||||
try {
|
||||
$response = $this->client->$method($url, [
|
||||
'on_headers' => $onHeaders,
|
||||
'curl' => $curlOptions,
|
||||
]);
|
||||
return new GuzzleResponse($response, $url);
|
||||
} catch (TransferException $exception) {
|
||||
if ($exception instanceof RequestException &&
|
||||
$exception->hasResponse()) {
|
||||
return new GuzzleResponse($exception->getResponse(), $url, $exception->getCode(), '');
|
||||
} else {
|
||||
return new CurlResult($url, '', ['http_code' => $exception->getCode()], $exception->getCode(), '');
|
||||
}
|
||||
} finally {
|
||||
$this->profiler->stopRecording();
|
||||
}
|
||||
}
|
||||
|
||||
/** {@inheritDoc}
|
||||
*
|
||||
* @throws HTTPException\InternalServerErrorException
|
||||
*/
|
||||
public function head(string $url, array $opts = [])
|
||||
{
|
||||
return $this->request('head', $url, $opts);
|
||||
}
|
||||
|
||||
/**
|
||||
* {@inheritDoc}
|
||||
*/
|
||||
public function get(string $url, array $opts = [])
|
||||
{
|
||||
return $this->request('get', $url, $opts);
|
||||
}
|
||||
|
||||
/**
|
||||
* {@inheritDoc}
|
||||
*
|
||||
* @param int $redirects The recursion counter for internal use - default 0
|
||||
*
|
||||
* @throws \Friendica\Network\HTTPException\InternalServerErrorException
|
||||
*/
|
||||
public function post(string $url, $params, array $headers = [], int $timeout = 0, &$redirects = 0)
|
||||
{
|
||||
$this->profiler->startRecording('network');
|
||||
|
||||
if (Network::isLocalLink($url)) {
|
||||
$this->logger->info('Local link', ['url' => $url, 'callstack' => System::callstack(20)]);
|
||||
}
|
||||
|
||||
if (Network::isUrlBlocked($url)) {
|
||||
$this->logger->info('Domain is blocked.' . ['url' => $url]);
|
||||
$this->profiler->stopRecording();
|
||||
return CurlResult::createErrorCurl($url);
|
||||
}
|
||||
|
||||
$ch = curl_init($url);
|
||||
|
||||
if (($redirects > 8) || (!$ch)) {
|
||||
$this->profiler->stopRecording();
|
||||
return CurlResult::createErrorCurl($url);
|
||||
}
|
||||
|
||||
$this->logger->debug('Post_url: start.', ['url' => $url]);
|
||||
|
||||
curl_setopt($ch, CURLOPT_HEADER, true);
|
||||
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
|
||||
curl_setopt($ch, CURLOPT_POST, 1);
|
||||
curl_setopt($ch, CURLOPT_POSTFIELDS, $params);
|
||||
curl_setopt($ch, CURLOPT_USERAGENT, $this->userAgent);
|
||||
|
||||
if ($this->config->get('system', 'ipv4_resolve', false)) {
|
||||
curl_setopt($ch, CURLOPT_IPRESOLVE, CURL_IPRESOLVE_V4);
|
||||
}
|
||||
|
||||
@curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
|
||||
|
||||
if (intval($timeout)) {
|
||||
curl_setopt($ch, CURLOPT_TIMEOUT, $timeout);
|
||||
} else {
|
||||
$curl_time = $this->config->get('system', 'curl_timeout', 60);
|
||||
curl_setopt($ch, CURLOPT_TIMEOUT, intval($curl_time));
|
||||
}
|
||||
|
||||
if (!empty($headers)) {
|
||||
curl_setopt($ch, CURLOPT_HTTPHEADER, $headers);
|
||||
}
|
||||
|
||||
$check_cert = $this->config->get('system', 'verifyssl');
|
||||
curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, (($check_cert) ? true : false));
|
||||
|
||||
if ($check_cert) {
|
||||
@curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, 2);
|
||||
}
|
||||
|
||||
$proxy = $this->config->get('system', 'proxy');
|
||||
|
||||
if (!empty($proxy)) {
|
||||
curl_setopt($ch, CURLOPT_HTTPPROXYTUNNEL, 1);
|
||||
curl_setopt($ch, CURLOPT_PROXY, $proxy);
|
||||
$proxyuser = $this->config->get('system', 'proxyuser');
|
||||
if (!empty($proxyuser)) {
|
||||
curl_setopt($ch, CURLOPT_PROXYUSERPWD, $proxyuser);
|
||||
}
|
||||
}
|
||||
|
||||
// don't let curl abort the entire application
|
||||
// if it throws any errors.
|
||||
|
||||
$s = @curl_exec($ch);
|
||||
|
||||
$curl_info = curl_getinfo($ch);
|
||||
|
||||
$curlResponse = new CurlResult($url, $s, $curl_info, curl_errno($ch), curl_error($ch));
|
||||
|
||||
if (!Network::isRedirectBlocked($url) && $curlResponse->isRedirectUrl()) {
|
||||
$redirects++;
|
||||
$this->logger->info('Post redirect.', ['url' => $url, 'to' => $curlResponse->getRedirectUrl()]);
|
||||
curl_close($ch);
|
||||
$this->profiler->stopRecording();
|
||||
return $this->post($curlResponse->getRedirectUrl(), $params, $headers, $redirects, $timeout);
|
||||
}
|
||||
|
||||
curl_close($ch);
|
||||
|
||||
$this->profiler->stopRecording();
|
||||
|
||||
// Very old versions of Lighttpd don't like the "Expect" header, so we remove it when needed
|
||||
if ($curlResponse->getReturnCode() == 417) {
|
||||
$redirects++;
|
||||
|
||||
if (empty($headers)) {
|
||||
$headers = ['Expect:'];
|
||||
} else {
|
||||
if (!in_array('Expect:', $headers)) {
|
||||
array_push($headers, 'Expect:');
|
||||
}
|
||||
}
|
||||
$this->logger->info('Server responds with 417, applying workaround', ['url' => $url]);
|
||||
return $this->post($url, $params, $headers, $redirects, $timeout);
|
||||
}
|
||||
|
||||
$this->logger->debug('Post_url: End.', ['url' => $url]);
|
||||
|
||||
return $curlResponse;
|
||||
}
|
||||
|
||||
/**
|
||||
* {@inheritDoc}
|
||||
*/
|
||||
public function finalUrl(string $url, int $depth = 1, bool $fetchbody = false)
|
||||
{
|
||||
if (Network::isLocalLink($url)) {
|
||||
$this->logger->info('Local link', ['url' => $url, 'callstack' => System::callstack(20)]);
|
||||
}
|
||||
|
||||
if (Network::isUrlBlocked($url)) {
|
||||
$this->logger->info('Domain is blocked.', ['url' => $url]);
|
||||
return $url;
|
||||
}
|
||||
|
||||
if (Network::isRedirectBlocked($url)) {
|
||||
$this->logger->info('Domain should not be redirected.', ['url' => $url]);
|
||||
return $url;
|
||||
}
|
||||
|
||||
$url = Network::stripTrackingQueryParams($url);
|
||||
|
||||
if ($depth > 10) {
|
||||
return $url;
|
||||
}
|
||||
|
||||
$url = trim($url, "'");
|
||||
|
||||
$this->profiler->startRecording('network');
|
||||
|
||||
$ch = curl_init();
|
||||
curl_setopt($ch, CURLOPT_URL, $url);
|
||||
curl_setopt($ch, CURLOPT_HEADER, 1);
|
||||
curl_setopt($ch, CURLOPT_NOBODY, 1);
|
||||
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
|
||||
curl_setopt($ch, CURLOPT_TIMEOUT, 10);
|
||||
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
|
||||
curl_setopt($ch, CURLOPT_USERAGENT, $this->userAgent);
|
||||
|
||||
curl_exec($ch);
|
||||
$curl_info = @curl_getinfo($ch);
|
||||
$http_code = $curl_info['http_code'];
|
||||
curl_close($ch);
|
||||
|
||||
$this->profiler->stopRecording();
|
||||
|
||||
if ($http_code == 0) {
|
||||
return $url;
|
||||
}
|
||||
|
||||
if (in_array($http_code, ['301', '302'])) {
|
||||
if (!empty($curl_info['redirect_url'])) {
|
||||
return $this->finalUrl($curl_info['redirect_url'], ++$depth, $fetchbody);
|
||||
} elseif (!empty($curl_info['location'])) {
|
||||
return $this->finalUrl($curl_info['location'], ++$depth, $fetchbody);
|
||||
}
|
||||
}
|
||||
|
||||
// Check for redirects in the meta elements of the body if there are no redirects in the header.
|
||||
if (!$fetchbody) {
|
||||
return $this->finalUrl($url, ++$depth, true);
|
||||
}
|
||||
|
||||
// if the file is too large then exit
|
||||
if ($curl_info["download_content_length"] > 1000000) {
|
||||
return $url;
|
||||
}
|
||||
|
||||
// if it isn't a HTML file then exit
|
||||
if (!empty($curl_info["content_type"]) && !strstr(strtolower($curl_info["content_type"]), "html")) {
|
||||
return $url;
|
||||
}
|
||||
|
||||
$this->profiler->startRecording('network');
|
||||
|
||||
$ch = curl_init();
|
||||
curl_setopt($ch, CURLOPT_URL, $url);
|
||||
curl_setopt($ch, CURLOPT_HEADER, 0);
|
||||
curl_setopt($ch, CURLOPT_NOBODY, 0);
|
||||
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
|
||||
curl_setopt($ch, CURLOPT_TIMEOUT, 10);
|
||||
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
|
||||
curl_setopt($ch, CURLOPT_USERAGENT, $this->userAgent);
|
||||
|
||||
$body = curl_exec($ch);
|
||||
curl_close($ch);
|
||||
|
||||
$this->profiler->stopRecording();
|
||||
|
||||
if (trim($body) == "") {
|
||||
return $url;
|
||||
}
|
||||
|
||||
// Check for redirect in meta elements
|
||||
$doc = new DOMDocument();
|
||||
@$doc->loadHTML($body);
|
||||
|
||||
$xpath = new DomXPath($doc);
|
||||
|
||||
$list = $xpath->query("//meta[@content]");
|
||||
foreach ($list as $node) {
|
||||
$attr = [];
|
||||
if ($node->attributes->length) {
|
||||
foreach ($node->attributes as $attribute) {
|
||||
$attr[$attribute->name] = $attribute->value;
|
||||
}
|
||||
}
|
||||
|
||||
if (@$attr["http-equiv"] == 'refresh') {
|
||||
$path = $attr["content"];
|
||||
$pathinfo = explode(";", $path);
|
||||
foreach ($pathinfo as $value) {
|
||||
if (substr(strtolower($value), 0, 4) == "url=") {
|
||||
return $this->finalUrl(substr($value, 4), ++$depth);
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
return $url;
|
||||
}
|
||||
|
||||
/**
|
||||
* {@inheritDoc}
|
||||
*/
|
||||
public function fetch(string $url, int $timeout = 0, string $accept_content = '', string $cookiejar = '')
|
||||
{
|
||||
$ret = $this->fetchFull($url, $timeout, $accept_content, $cookiejar);
|
||||
|
||||
return $ret->getBody();
|
||||
}
|
||||
|
||||
/**
|
||||
* {@inheritDoc}
|
||||
*/
|
||||
public function fetchFull(string $url, int $timeout = 0, string $accept_content = '', string $cookiejar = '')
|
||||
{
|
||||
return $this->get(
|
||||
$url,
|
||||
[
|
||||
'timeout' => $timeout,
|
||||
'accept_content' => $accept_content,
|
||||
'cookiejar' => $cookiejar
|
||||
]
|
||||
);
|
||||
}
|
||||
}
|
||||
Loading…
Add table
Add a link
Reference in a new issue