17290669502

Committed 28 Aug 2025 08:44AM UTC coverage: 74.685%. First build

Build # 17290669502

Build Type

Pull #1162

github

Committed by

web-flow

Commit Message

Merge b636485a6 into cdedf1aeb

Pull Request Pull Request #1162: [FEATURE] Extract error information in SubProcess crawl strategy (TYPO3v11)

Run Details

0 of 17 new or added lines in 1 file covered. (0.0%)

1953 of 2615 relevant lines covered (74.68%)

3.72 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

16.44

/Classes/CrawlStrategy/SubProcessExecutionStrategy.php

<?php

declare(strict_types=1);

namespace AOE\Crawler\CrawlStrategy;

/*
 * (c) 2020 AOE GmbH <dev@aoe.com>
 *
 * This file is part of the TYPO3 Crawler Extension.
 *
 * It is free software; you can redistribute it and/or modify it under
 * the terms of the GNU General Public License, either version 2
 * of the License, or any later version.
 *
 * For the full copyright and license information, please read the
 * LICENSE.txt file that was distributed with this source code.
 *
 * The TYPO3 project - inspiring people to share!
 */

use AOE\Crawler\Configuration\ExtensionConfigurationProvider;
use AOE\Crawler\Utility\PhpBinaryUtility;
use Psr\Http\Message\UriInterface;
use Psr\Log\LoggerAwareInterface;
use Psr\Log\LoggerAwareTrait;
use TYPO3\CMS\Core\Core\Environment;
use TYPO3\CMS\Core\Utility\CommandUtility;
use TYPO3\CMS\Core\Utility\ExtensionManagementUtility;
use TYPO3\CMS\Core\Utility\GeneralUtility;

/**
 * Executes another process via shell_exec() to include cli/bootstrap.php which in turn
 * includes the index.php for frontend.
 */
class SubProcessExecutionStrategy implements LoggerAwareInterface, CrawlStrategy
{
    use LoggerAwareTrait;

    /**
     * @var array
     */
    protected $extensionSettings;

    public function __construct(?ExtensionConfigurationProvider $configurationProvider = null)
    {
        $configurationProvider = $configurationProvider ?? GeneralUtility::makeInstance(ExtensionConfigurationProvider::class);
        $settings = $configurationProvider->getExtensionConfiguration();
        $this->extensionSettings = is_array($settings) ? $settings : [];
    }

    /**
     * Fetches a URL by calling a shell script.
     *
     * @return array|bool|mixed
     */
    public function fetchUrlContents(UriInterface $url, string $crawlerId)
    {
        $url = (string) $url;
        $parsedUrl = parse_url($url);

        if ($parsedUrl === false) {
            $this->logger->debug(
                sprintf('Could not parse_url() for string "%s"', $url),
                ['crawlerId' => $crawlerId]
            );
            return false;
        }

        if (! isset($parsedUrl['scheme']) || ! in_array($parsedUrl['scheme'], ['', 'http', 'https'], true)) {
            $this->logger->debug(
                sprintf('Scheme does not match for url "%s"', $url),
                ['crawlerId' => $crawlerId]
            );
            return false;
        }

        if (! is_array($parsedUrl)) {
            return [];
        }

        $requestHeaders = $this->buildRequestHeaders($parsedUrl, $crawlerId);

        $commandParts = [
            ExtensionManagementUtility::extPath('crawler') . 'cli/bootstrap.php',
            $this->getFrontendBasePath(),
            $url,
            base64_encode(serialize($requestHeaders)),
        ];
        $commandParts = CommandUtility::escapeShellArguments($commandParts);
        $cmd = escapeshellcmd(PhpBinaryUtility::getPhpBinary());
        $cmd .= ' ' . implode(' ', $commandParts);

        $startTime = microtime(true);
        $content = $this->executeShellCommand($cmd);
        $this->logger->info($url . ' ' . (microtime(true) - $startTime));

        if ($content === null) {
            return false;
        }
        if (str_contains($content, 'typo3-error-page')) {
            preg_match('#class="typo3-error-page-statuscode">(.+?)</#s', $content, $matchStatus);
            preg_match('#class="typo3-error-page-title">(.+?)</#s', $content, $matchTitle);
            preg_match('#class="typo3-error-page-message">(.+?)</#s', $content, $matchMessage);
            $message = trim($matchStatus[1] ?? '')
                . ' ' . trim($matchTitle[1] ?? '')
                . ' - ' . trim($matchMessage[1] ?? '');
            $this->logger->debug(
                sprintf('Error while opening "%s" - %s', $url, $message),
                [
                    'crawlerId' => $crawlerId,
                ]
            );
            return [
                'errorlog' => [$message],
                'content' => $content,
            ];
        }

        return ['content' => $content];
    }

    private function buildRequestHeaders(array $url, string $crawlerId): array
    {
        $reqHeaders = [];
        $reqHeaders[] = 'GET ' . $url['path'] . (isset($url['query']) ? '?' . $url['query'] : '') . ' HTTP/1.0';
        $reqHeaders[] = 'Host: ' . $url['host'];
        $reqHeaders[] = 'Connection: close';
        if (isset($url['user'], $url['pass']) && $url['user'] !== '' && $url['pass'] !== '') {
            $reqHeaders[] = 'Authorization: Basic ' . base64_encode($url['user'] . ':' . $url['pass']);
        }
        $reqHeaders[] = 'X-T3crawler: ' . $crawlerId;
        $reqHeaders[] = 'User-Agent: TYPO3 crawler';
        return $reqHeaders;
    }

    /**
     * Executes a shell command and returns the outputted result.
     *
     * @param string $command Shell command to be executed
     * @return string|null Outputted result of the command execution
     */
    private function executeShellCommand($command)
    {
        return shell_exec($command);
    }

    /**
     * Gets the base path of the website frontend.
     * (e.g. if you call http://mydomain.com/cms/index.php in
     * the browser the base path is "/cms/")
     *
     * @return string Base path of the website frontend
     */
    private function getFrontendBasePath()
    {
        $frontendBasePath = '/';

        // Get the path from the extension settings:
        if (isset($this->extensionSettings['frontendBasePath']) && $this->extensionSettings['frontendBasePath']) {
            $frontendBasePath = $this->extensionSettings['frontendBasePath'];
        // If empty, try to use config.absRefPrefix:
        } elseif (isset($GLOBALS['TSFE']->absRefPrefix) && ! empty($GLOBALS['TSFE']->absRefPrefix)) {
            $frontendBasePath = $GLOBALS['TSFE']->absRefPrefix;
        // If not in CLI mode the base path can be determined from $_SERVER environment:
        } elseif (! Environment::isCli()) {
            $frontendBasePath = GeneralUtility::getIndpEnv('TYPO3_SITE_PATH');
        }

        // Base path must be '/<pathSegements>/':
        if ($frontendBasePath !== '/') {
            $frontendBasePath = '/' . ltrim($frontendBasePath, '/');
            $frontendBasePath = rtrim($frontendBasePath, '/') . '/';
        }

        return $frontendBasePath;
    }
}

1	<?php
2
3	declare(strict_types=1);
4
5	namespace AOE\Crawler\CrawlStrategy;
6
7	/*
8	* (c) 2020 AOE GmbH <dev@aoe.com>
9	*
10	* This file is part of the TYPO3 Crawler Extension.
11	*
12	* It is free software; you can redistribute it and/or modify it under
13	* the terms of the GNU General Public License, either version 2
14	* of the License, or any later version.
15	*
16	* For the full copyright and license information, please read the
17	* LICENSE.txt file that was distributed with this source code.
18	*
19	* The TYPO3 project - inspiring people to share!
20	*/
21
22	use AOE\Crawler\Configuration\ExtensionConfigurationProvider;
23	use AOE\Crawler\Utility\PhpBinaryUtility;
24	use Psr\Http\Message\UriInterface;
25	use Psr\Log\LoggerAwareInterface;
26	use Psr\Log\LoggerAwareTrait;
27	use TYPO3\CMS\Core\Core\Environment;
28	use TYPO3\CMS\Core\Utility\CommandUtility;
29	use TYPO3\CMS\Core\Utility\ExtensionManagementUtility;
30	use TYPO3\CMS\Core\Utility\GeneralUtility;
31
32	/**
33	* Executes another process via shell_exec() to include cli/bootstrap.php which in turn
34	* includes the index.php for frontend.
35	*/
36	class SubProcessExecutionStrategy implements LoggerAwareInterface, CrawlStrategy
37	{
38	use LoggerAwareTrait;
39
40	/**
41	* @var array
42	*/
43	protected $extensionSettings;
44
45	public function __construct(?ExtensionConfigurationProvider $configurationProvider = null)
46	{
47	$configurationProvider = $configurationProvider ?? GeneralUtility::makeInstance(ExtensionConfigurationProvider::class);	3✔
48	$settings = $configurationProvider->getExtensionConfiguration();	3✔
49	$this->extensionSettings = is_array($settings) ? $settings : [];	3✔
50	}
51
52	/**
53	* Fetches a URL by calling a shell script.
54	*
55	* @return array\|bool\|mixed
56	*/
57	public function fetchUrlContents(UriInterface $url, string $crawlerId)
58	{
59	$url = (string) $url;	1✔
60	$parsedUrl = parse_url($url);	1✔
61
62	if ($parsedUrl === false) {	1✔
63	$this->logger->debug(	×
64	sprintf('Could not parse_url() for string "%s"', $url),	×
65	['crawlerId' => $crawlerId]	×
66	);	×
67	return false;	×
68	}
69
70	if (! isset($parsedUrl['scheme']) \|\| ! in_array($parsedUrl['scheme'], ['', 'http', 'https'], true)) {	1✔
71	$this->logger->debug(	1✔
72	sprintf('Scheme does not match for url "%s"', $url),	1✔
73	['crawlerId' => $crawlerId]	1✔
74	);	1✔
75	return false;	1✔
76	}
77
78	if (! is_array($parsedUrl)) {	×
79	return [];	×
80	}
81
82	$requestHeaders = $this->buildRequestHeaders($parsedUrl, $crawlerId);	×
83
84	$commandParts = [	×
85	ExtensionManagementUtility::extPath('crawler') . 'cli/bootstrap.php',	×
86	$this->getFrontendBasePath(),	×
87	$url,	×
88	base64_encode(serialize($requestHeaders)),	×
89	];	×
90	$commandParts = CommandUtility::escapeShellArguments($commandParts);	×
91	$cmd = escapeshellcmd(PhpBinaryUtility::getPhpBinary());	×
92	$cmd .= ' ' . implode(' ', $commandParts);	×
93
94	$startTime = microtime(true);	×
95	$content = $this->executeShellCommand($cmd);	×
96	$this->logger->info($url . ' ' . (microtime(true) - $startTime));	×
97
98	if ($content === null) {	×
99	return false;	×
100	}
NEW 101	if (str_contains($content, 'typo3-error-page')) {	×
NEW 102	preg_match('#class="typo3-error-page-statuscode">(.+?)</#s', $content, $matchStatus);	×
NEW 103	preg_match('#class="typo3-error-page-title">(.+?)</#s', $content, $matchTitle);	×
NEW 104	preg_match('#class="typo3-error-page-message">(.+?)</#s', $content, $matchMessage);	×
NEW 105	$message = trim($matchStatus[1] ?? '')	×
NEW 106	. ' ' . trim($matchTitle[1] ?? '')	×
NEW 107	. ' - ' . trim($matchMessage[1] ?? '');	×
NEW 108	$this->logger->debug(	×
NEW 109	sprintf('Error while opening "%s" - %s', $url, $message),	×
NEW 110	[	×
NEW 111	'crawlerId' => $crawlerId,	×
NEW 112	]	×
NEW 113	);	×
NEW 114	return [	×
NEW 115	'errorlog' => [$message],	×
NEW 116	'content' => $content,	×
NEW 117	];	×
118	}
119
120	return ['content' => $content];	×
121	}
122
123	private function buildRequestHeaders(array $url, string $crawlerId): array
124	{
125	$reqHeaders = [];	×
126	$reqHeaders[] = 'GET ' . $url['path'] . (isset($url['query']) ? '?' . $url['query'] : '') . ' HTTP/1.0';	×
127	$reqHeaders[] = 'Host: ' . $url['host'];	×
128	$reqHeaders[] = 'Connection: close';	×
129	if (isset($url['user'], $url['pass']) && $url['user'] !== '' && $url['pass'] !== '') {	×
130	$reqHeaders[] = 'Authorization: Basic ' . base64_encode($url['user'] . ':' . $url['pass']);	×
131	}
132	$reqHeaders[] = 'X-T3crawler: ' . $crawlerId;	×
133	$reqHeaders[] = 'User-Agent: TYPO3 crawler';	×
134	return $reqHeaders;	×
135	}
136
137	/**
138	* Executes a shell command and returns the outputted result.
139	*
140	* @param string $command Shell command to be executed
141	* @return string\|null Outputted result of the command execution
142	*/
143	private function executeShellCommand($command)
144	{
145	return shell_exec($command);	×
146	}
147
148	/**
149	* Gets the base path of the website frontend.
150	* (e.g. if you call http://mydomain.com/cms/index.php in
151	* the browser the base path is "/cms/")
152	*
153	* @return string Base path of the website frontend
154	*/
155	private function getFrontendBasePath()
156	{
157	$frontendBasePath = '/';	×
158
159	// Get the path from the extension settings:
160	if (isset($this->extensionSettings['frontendBasePath']) && $this->extensionSettings['frontendBasePath']) {	×
161	$frontendBasePath = $this->extensionSettings['frontendBasePath'];	×
162	// If empty, try to use config.absRefPrefix:
163	} elseif (isset($GLOBALS['TSFE']->absRefPrefix) && ! empty($GLOBALS['TSFE']->absRefPrefix)) {	×
164	$frontendBasePath = $GLOBALS['TSFE']->absRefPrefix;	×
165	// If not in CLI mode the base path can be determined from $_SERVER environment:
166	} elseif (! Environment::isCli()) {	×
167	$frontendBasePath = GeneralUtility::getIndpEnv('TYPO3_SITE_PATH');	×
168	}
169
170	// Base path must be '/<pathSegements>/':
171	if ($frontendBasePath !== '/') {	×
172	$frontendBasePath = '/' . ltrim($frontendBasePath, '/');	×
173	$frontendBasePath = rtrim($frontendBasePath, '/') . '/';	×
174	}
175
176	return $frontendBasePath;	×
177	}
178	}

tomasnorre / crawler / 17290669502

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous