friendica
/
friendica
mirror of https://github.com/friendica/friendica
6
0
Fork 1
Code Issues Packages Projects Releases 1 Wiki Activity
friendica/src/Network/HTTPRequest.php

524 lines
14 KiB
PHP
Raw Blame History

<?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\App;
use Friendica\Core\Config\IConfig;
use Friendica\Core\System;
use Friendica\Util\Network;
use Friendica\Util\Profiler;
use GuzzleHttp\Client;
use GuzzleHttp\Exception\RequestException;
use GuzzleHttp\Exception\TransferException;
use Psr\Http\Message\RequestInterface;
use Psr\Http\Message\ResponseInterface;
use Psr\Http\Message\UriInterface;
use Psr\Log\LoggerInterface;
/**
* Performs HTTP requests to a given URL
*/
class HTTPRequest implements IHTTPRequest
{
/** @var LoggerInterface */
private $logger;
/** @var Profiler */
private $profiler;
/** @var IConfig */
private $config;
/** @var string */
private $baseUrl;
public function __construct(LoggerInterface $logger, Profiler $profiler, IConfig $config, App\BaseURL $baseUrl)
{
$this->logger = $logger;
$this->profiler = $profiler;
$this->config = $config;
$this->baseUrl = $baseUrl->get();
}
/** {@inheritDoc}
*
* @throws HTTPException\InternalServerErrorException
*/
public function head(string $url, array $opts = [])
{
$opts['nobody'] = true;
return $this->get($url, $opts);
}
/**
* {@inheritDoc}
*/
public function get(string $url, bool $binary = false, 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);
}
$curlOptions = [];
$curlOptions[CURLOPT_HEADER] = true;
if (!empty($opts['cookiejar'])) {
$curlOptions[CURLOPT_COOKIEJAR] = $opts["cookiejar"];
$curlOptions[CURLOPT_COOKIEFILE] = $opts["cookiejar"];
}
// These settings aren't needed. We're following the location already.
// $curlOptions[CURLOPT_FOLLOWLOCATION] =true;
// $curlOptions[CURLOPT_MAXREDIRS] = 5;
if (!empty($opts['accept_content'])) {
if (empty($curlOptions[CURLOPT_HTTPHEADER])) {
$curlOptions[CURLOPT_HTTPHEADER] = [];
}
array_push($curlOptions[CURLOPT_HTTPHEADER], 'Accept: ' . $opts['accept_content']);
}
if (!empty($opts['header'])) {
if (empty($curlOptions[CURLOPT_HTTPHEADER])) {
$curlOptions[CURLOPT_HTTPHEADER] = [];
}
$curlOptions[CURLOPT_HTTPHEADER] = array_merge($opts['header'], $curlOptions[CURLOPT_HTTPHEADER]);
}
$curlOptions[CURLOPT_RETURNTRANSFER] = true;
$curlOptions[CURLOPT_USERAGENT] = $this->getUserAgent();
$range = intval($this->config->get('system', 'curl_range_bytes', 0));
if ($range > 0) {
$curlOptions[CURLOPT_RANGE] = '0-' . $range;
}
// Without this setting it seems as if some webservers send compressed content
// This seems to confuse curl so that it shows this uncompressed.
/// @todo We could possibly set this value to "gzip" or something similar
$curlOptions[CURLOPT_ENCODING] = '';
if (!empty($opts['headers'])) {
$this->logger->notice('Wrong option \'headers\' used.');
if (empty($curlOptions[CURLOPT_HTTPHEADER])) {
$curlOptions[CURLOPT_HTTPHEADER] = [];
}
$curlOptions[CURLOPT_HTTPHEADER] = array_merge($opts['headers'], $curlOptions[CURLOPT_HTTPHEADER]);
}
if (!empty($opts['nobody'])) {
$curlOptions[CURLOPT_NOBODY] = $opts['nobody'];
}
$curlOptions[CURLOPT_CONNECTTIMEOUT] = 10;
if (!empty($opts['timeout'])) {
$curlOptions[CURLOPT_TIMEOUT] = $opts['timeout'];
} else {
$curl_time = $this->config->get('system', 'curl_timeout', 60);
$curlOptions[CURLOPT_TIMEOUT] = intval($curl_time);
}
// by default we will allow self-signed certs
// but you can override this
$check_cert = $this->config->get('system', 'verifyssl');
$curlOptions[CURLOPT_SSL_VERIFYPEER] = ($check_cert) ? true : false;
if ($check_cert) {
$curlOptions[CURLOPT_SSL_VERIFYHOST] = 2;
}
$proxy = $this->config->get('system', 'proxy');
if (!empty($proxy)) {
$curlOptions[CURLOPT_HTTPPROXYTUNNEL] = 1;
$curlOptions[CURLOPT_PROXY] = $proxy;
$proxyuser = $this->config->get('system', 'proxyuser');
if (!empty($proxyuser)) {
$curlOptions[CURLOPT_PROXYUSERPWD] = $proxyuser;
}
}
if ($this->config->get('system', 'ipv4_resolve', false)) {
$curlOptions[CURLOPT_IPRESOLVE] = CURL_IPRESOLVE_V4;
}
if ($binary) {
$curlOptions[CURLOPT_BINARYTRANSFER] = 1;
}
$logger = $this->logger;
$onRedirect = function(
RequestInterface $request,
ResponseInterface $response,
UriInterface $uri
) use ($logger) {
$logger->notice('Curl redirect.', ['url' => $request->getUri(), 'to' => $uri]);
};
$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!');
}
};
$client = new Client([
'allow_redirect' => [
'max' => 8,
'on_redirect' => $onRedirect,
'track_redirect' => true,
'strict' => true,
'referer' => true,
],
'on_headers' => $onHeaders,
'sink' => tempnam(get_temppath(), 'guzzle'),
'curl' => $curlOptions
]);
try {
$response = $client->get($url);
return new GuzzleResponse($response, $url);
} catch (TransferException $exception) {
if ($exception instanceof RequestException &&
$exception->hasResponse()) {
return new GuzzleResponse($exception->getResponse(), $url, $exception->getCode(), $exception->getMessage());
} else {
return new CurlResult($url, '', ['http_code' => $exception->getCode()], $exception->getCode(), $exception->getMessage());
}
} finally {
$this->profiler->stopRecording();
}
}
/**
* {@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, $this->getUserAgent());
if ($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, $this->getUserAgent());
curl_exec($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->getUserAgent());
$body = curl_exec($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}
*
* @param int $redirects The recursion counter for internal use - default 0
*
* @throws \Friendica\Network\HTTPException\InternalServerErrorException
*/
public function fetch(string $url, int $timeout = 0, string $accept_content = '', string $cookiejar = '', &$redirects = 0)
{
$ret = $this->fetchFull($url, $timeout, $accept_content, $cookiejar, $redirects);
return $ret->getBody();
}
/**
* {@inheritDoc}
*
* @param int $redirects The recursion counter for internal use - default 0
*
* @throws \Friendica\Network\HTTPException\InternalServerErrorException
*/
public function fetchFull(string $url, int $timeout = 0, string $accept_content = '', string $cookiejar = '', &$redirects = 0)
{
return $this->get(
$url,
[
'timeout' => $timeout,
'accept_content' => $accept_content,
'cookiejar' => $cookiejar
]
);
}
/**
* {@inheritDoc}
*/
public function getUserAgent()
{
return
FRIENDICA_PLATFORM . " '" .
FRIENDICA_CODENAME . "' " .
FRIENDICA_VERSION . '-' .
DB_UPDATE_VERSION . '; ' .
$this->baseUrl;
}
}
View Git Blame Copy Permalink