pcgamer: meta tag change (#4670)
Some checks are pending
Build Image on Commit and Release / bake (push) Waiting to run
Lint / phpcs (7.4) (push) Waiting to run
Lint / phpcompatibility (7.4) (push) Waiting to run
Lint / executable_php_files_check (push) Waiting to run
Tests / phpunit8 (8.0) (push) Waiting to run
Tests / phpunit8 (8.1) (push) Waiting to run
Tests / phpunit8 (8.2) (push) Waiting to run
Tests / phpunit8 (8.3) (push) Waiting to run
Tests / phpunit8 (8.4) (push) Waiting to run
Tests / phpunit8 (7.4) (push) Waiting to run

* pcgamer: the parsely tags are gone, use different tags

* apply phpcs.xml rules
This commit is contained in:
Matt DeMoss 2025-08-14 13:06:39 -04:00 committed by GitHub
parent 1b584b4551
commit ee4f85cc94
No known key found for this signature in database
GPG key ID: B5690EEEBB952194

View file

@ -25,22 +25,40 @@ class PcGamerBridge extends BridgeAbstract
$articleHtml = getSimpleHTMLDOMCached($item['uri']); $articleHtml = getSimpleHTMLDOMCached($item['uri']);
// Relying on meta tags ought to be more reliable. // Relying on meta tags ought to be more reliable.
$item['title'] = $articleHtml->find('meta[name=parsely-title]', 0)->content; $item['title'] = $articleHtml->find('meta[property=og:title]', 0)->content;
$item['content'] = html_entity_decode($articleHtml->find('meta[name=description]', 0)->content); $item['content'] = html_entity_decode($articleHtml->find('meta[name=description]', 0)->content);
$item['author'] = $articleHtml->find('meta[name=parsely-author]', 0)->content;
$imageUrl = $articleHtml->find('meta[name=parsely-image-url]', 0); // TODO: parsely-author is no longer available, but it is in the application/ld+json
$item['author'] = $articleHtml->find('a[rel=author]', 0)->innertext;
$imageUrl = $articleHtml->find('meta[property=og:image]', 0);
if ($imageUrl) { if ($imageUrl) {
$item['enclosures'][] = $imageUrl->content; $item['enclosures'][] = $imageUrl->content;
} }
/* I don't know why every article has two extra tags, but because /*
one matches another common tag, "guide," it needs to be removed. */ Tags in mrf:tags are semicolon-delimited and each begins with a label and a ':'
$item['categories'] = array_diff( Example:
explode(',', $articleHtml->find('meta[name=parsely-tags]', 0)->content), "region:US;articleType:News;channel:Gaming software;"
['van_buying_guide_progressive', 'serversidehawk'] Find the tag, replace ; with \n, remove the label prefixes, then explode by newline.
*/
$item['categories'] = array_unique(
explode(
PHP_EOL,
preg_replace(
'/^[^:]+:/m',
'',
preg_replace(
'/;/',
PHP_EOL,
$articleHtml->find('meta[property=mrf:tags]', 0)->content
)
)
)
); );
$item['timestamp'] = strtotime($articleHtml->find('meta[name=pub_date]', 0)->content); $item['timestamp'] = strtotime($articleHtml->find('meta[name=pub_date]', 0)->content);
$this->items[] = $item; $this->items[] = $item;
} }
} }