Friendica/friendica-github
1
0
Fork 0
mirror of https://github.com/friendica/friendica synced 2025-04-30 19:44:24 +02:00
Code Issues Projects Releases Packages Wiki Activity

Improve plaintext generation for language detection

Browse source
This commit is contained in:
Michael 2020-10-05 12:50:18 +00:00
parent cc42c0ba27
commit 16224a7001
2 changed files with 24 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

12
src/Model/Item.php
View file

@ -2476,7 +2476,17 @@ class Item
return '';
}
$naked_body = BBCode::toPlaintext($item['body'], false);
// Convert attachments to links
$naked_body = BBCode::removeAttachment($item['body']);
// Remove links and pictures
$naked_body = BBCode::removeLinks($naked_body);
// Convert the title and the body to plain text
$naked_body = trim($item['title'] . "\n" . BBCode::toPlaintext($naked_body));
// Remove possibly remaining links
$naked_body = preg_replace(Strings::autoLinkRegEx(), '', $naked_body);
$ld = new Language();
$languages = $ld->detect($naked_body)->limit(0, 3)->close();