Fix a normalize whitespace error in the DomCrawler

[fritzmg]

Currently the following error will occur when running the Crawler:

"exception" => Exception {
    #message: ""Symfony\Component\DomCrawler\Crawler::text()" will normalize whitespaces by default in Symfony 5.0, set the second "$normalizeWhitespace" argument se to retrieve the non-normalized version of the text."
    #code: 16384
    #file: "vendor\terminal42\escargot\src\Subscriber\RobotsSubscriber.php"
    #line: 244
    trace: {
        vendor\terminal42\escargot\src\Subscriber\RobotsSubscriber.php:244 { …}
        Terminal42\Escargot\Subscriber\RobotsSubscriber->Terminal42\Escargot\Subscriber\{closure}() {}
        vendor\symfony\dom-crawler\Crawler.php:622 { …}
        vendor\contao\contao\core-bundle\src\Search\Document.php:113 { …}
        vendor\symfony\dom-crawler\Crawler.php:355 { …}
        vendor\contao\contao\core-bundle\src\Search\Document.php:120 { …}
        vendor\contao\contao\core-bundle\src\Search\Indexer\DefaultIndexer.php:146 { …}
        vendor\contao\contao\core-bundle\src\Search\Indexer\DefaultIndexer.php:76 { …}
        vendor\contao\contao\core-bundle\src\Search\Indexer\DelegatingIndexer.php:34 { …}
        vendor\contao\contao\core-bundle\src\Crawl\Escargot\Subscriber\SearchIndexSubscriber.php:165 { …}
        vendor\terminal42\escargot\src\Escargot.php:515 { …}
        vendor\terminal42\escargot\src\Escargot.php:457 { …}
        vendor\terminal42\escargot\src\Escargot.php:330 { …}
        vendor\contao\contao\core-bundle\src\Command\CrawlCommand.php:135 { …}
        vendor\symfony\console\Command\Command.php:255 { …}
        vendor\symfony\console\Application.php:1027 { …}
        vendor\symfony\framework-bundle\Console\Application.php:97 { …}
        vendor\symfony\console\Application.php:273 { …}
        vendor\symfony\framework-bundle\Console\Application.php:83 { …}
        vendor\symfony\console\Application.php:149 { …}
        vendor\bin\contao-console:21 { …}
        vendor\bin\contao-console:21 { …}
    }
}

This PR fixes this by specifically setting the second parameter of Symfony\Component\DomCrawler\Crawler::text to true (which was previously the default).

[Toflar]

We should have noticed this with tests. Apparently that was not covered, can you add a test case for that?

[fritzmg]

Hm, the tests would already cover this - but no error occurs there. Presumably because trigger_error is already silenced in Symfony. However, I suppose that may be the custom error handler in Escargot makes the error then actually visible. May be Escargot should restore the error handler after this block?

[fritzmg]

That would explain why the error only now surfaces. The error handler in Escargot was only added 10 days ago (in terminal42/escargot@ef7d131#diff-e66b75aabb721cde7baf72cca5b86fcc33eb03e24ed5b5d550d0a0b2c476972c).

@Toflar do you want me to create an issue there?

[leofeyer]

I don't understand. If true was the default and will still be the default in the future, why do we have to set it explicitly?

[fritzmg]

It is not technically necessary with the release of terminal42/escargot 1.0.5. However, not setting the parameters will still result in

@trigger_error(sprintf('"%s()" will normalize whitespaces by default in Symfony 5.0, set the second "$normalizeWhitespace" argument to false to retrieve the non-normalized version of the text.', __METHOD__), \E_USER_DEPRECATED);

being triggered under certain circumstances.

[leofeyer]

Then we should ignore the deprecation in these places as we do not want the non-normalized version of the text.

[ausi]

If we do not want the non-normalized version we have to use ->text(null, true) in Symfony 4.

[leofeyer]

Thank you @fritzmg.