Merge pull request #1449 from acelaya-forks/feature/title-html-entities

Feature/title html entities
This commit is contained in:
Alejandro Celaya 2022-05-22 10:12:54 +02:00 committed by GitHub
commit 2b69f5eff4
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
3 changed files with 26 additions and 3 deletions

View file

@ -4,6 +4,23 @@ All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com), and this project adheres to [Semantic Versioning](https://semver.org).
## [Unreleased]
### Added
* *Nothing*
### Changed
* *Nothing*
### Deprecated
* *Nothing*
### Removed
* *Nothing*
### Fixed
* [#1448](https://github.com/shlinkio/shlink/issues/1448) Fixed HTML entities not being properly parsed when auto-resolving page titles.
## [3.1.1] - 2022-05-09
### Added
* *Nothing*

View file

@ -13,6 +13,7 @@ use Shlinkio\Shlink\Core\Exception\InvalidUrlException;
use Shlinkio\Shlink\Core\Options\UrlShortenerOptions;
use Throwable;
use function html_entity_decode;
use function preg_match;
use function str_contains;
use function str_starts_with;
@ -71,7 +72,7 @@ class UrlValidator implements UrlValidatorInterface, RequestMethodInterface
$collectedBody .= $body->read(1024);
}
preg_match(TITLE_TAG_VALUE, $collectedBody, $matches);
return isset($matches[1]) ? trim($matches[1]) : null;
return isset($matches[1]) ? $this->normalizeTitle($matches[1]) : null;
}
/**
@ -101,4 +102,9 @@ class UrlValidator implements UrlValidatorInterface, RequestMethodInterface
return null;
}
}
private function normalizeTitle(string $title): string
{
return html_entity_decode(trim($title));
}
}

View file

@ -128,7 +128,7 @@ class UrlValidatorTest extends TestCase
$result = $this->urlValidator->validateUrlWithTitle('http://foobar.com/12345/hello?foo=bar', true);
self::assertEquals('Resolved title', $result);
self::assertEquals('Resolved "title"', $result);
$request->shouldHaveBeenCalledOnce();
}
@ -162,7 +162,7 @@ class UrlValidatorTest extends TestCase
private function respWithTitle(): Response
{
$body = $this->createStreamWithContent('<title data-foo="bar"> Resolved title</title>');
$body = $this->createStreamWithContent('<title data-foo="bar"> Resolved &quot;title&quot; </title>');
return new Response($body, 200, ['Content-Type' => 'TEXT/html; charset=utf-8']);
}