clones/friendica
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Merge pull request #9365 from annando/better-plaintext 

Improve plaintext generation for language detection
This commit is contained in:
Hypolite Petovan 2020-10-05 23:11:42 -04:00 committed by GitHub
parent 24cbbd72a5 20652870b6
commit 9954bf8adf
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 22 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
src/Content/Text/BBCode.php
View File

@ -1220,6 +1220,17 @@ class BBCode
return $return; return $return;
} }
public static function removeLinks(string $bbcode)
{
$bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
$bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
$bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
$bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
$bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
return $bbcode;
}
/** /**
* Converts a BBCode message to HTML message * Converts a BBCode message to HTML message
* *

12
src/Model/Item.php
View File

@ -2482,7 +2482,17 @@ class Item
return ''; 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(); $ld = new Language();
$languages = $ld->detect($naked_body)->limit(0, 3)->close(); $languages = $ld->detect($naked_body)->limit(0, 3)->close();