Knah-Tsaeb/MyShaarli
Knah-Tsaeb/MyShaarli
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

Merge pull request #1665 from ArthurHoaro/fix/metadata-regexes-2

Browse source 
Fix metadata extract regex (2)
This commit is contained in:
ArthurHoaro 2020-12-29 11:43:39 +01:00 committed by GitHub
commit 8ed5fbef8f
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
2 changed files with 14 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
application/bookmark/LinkUtils.php
View file

@ -68,11 +68,13 @@ function html_extract_tag($tag, $html)
$properties = implode('|', $propertiesKey); $properties = implode('|', $propertiesKey);
// We need a OR here to accept either 'property=og:noquote' or 'property="og:unrelated og:my-tag"' // We need a OR here to accept either 'property=og:noquote' or 'property="og:unrelated og:my-tag"'
$orCondition = '["\']?(?:og:)?' . $tag . '["\']?|["\'][^\'"]*?(?:og:)?' . $tag . '[^\'"]*?[\'"]'; $orCondition = '["\']?(?:og:)?' . $tag . '["\']?|["\'][^\'"]*?(?:og:)?' . $tag . '[^\'"]*?[\'"]';
// Support quotes in double quoted content, and the other way around
$content = 'content=(["\'])((?:(?!\1).)*)\1';
// Try to retrieve OpenGraph tag. // Try to retrieve OpenGraph tag.
$ogRegex = '#<meta[^>]+(?:' . $properties . ')=(?:' . $orCondition . ')[^>]*content=(["\'])([^\1]*?)\1.*?>#'; $ogRegex = '#<meta[^>]+(?:' . $properties . ')=(?:' . $orCondition . ')[^>]*' . $content . '.*?>#';
// If the attributes are not in the order property => content (e.g. Github) // If the attributes are not in the order property => content (e.g. Github)
// New regex to keep this readable... more or less. // New regex to keep this readable... more or less.
$ogRegexReverse = '#<meta[^>]+content=(["\'])([^\1]*?)\1[^>]+(?:' . $properties . ')=(?:' . $orCondition . ').*?>#'; $ogRegexReverse = '#<meta[^>]+' . $content . '[^>]+(?:' . $properties . ')=(?:' . $orCondition . ').*?>#';
if ( if (
preg_match($ogRegex, $html, $matches) > 0 preg_match($ogRegex, $html, $matches) > 0

10
tests/bookmark/LinkUtilsTest.php
View file

@ -245,6 +245,16 @@ public function testHtmlExtractNonExistentOgTag()
$this->assertFalse(html_extract_tag('description', $html)); $this->assertFalse(html_extract_tag('description', $html));
} }
public function testHtmlExtractDescriptionFromGoogleRealCase(): void
{
$html = 'id="gsr"><meta content="Fêtes de fin d\'année" property="twitter:title"><meta '.
'content="Bonnes fêtes de fin d\'année ! #GoogleDoodle" property="twitter:description">'.
'<meta content="Bonnes fêtes de fin d\'année ! #GoogleDoodle" property="og:description">'.
'<meta content="summary_large_image" property="twitter:card"><meta co'
;
$this->assertSame('Bonnes fêtes de fin d\'année ! #GoogleDoodle', html_extract_tag('description', $html));
}
/** /**
* Test the header callback with valid value * Test the header callback with valid value
*/ */