Author: mikel_maron

Description:
If Categories are listed in an @optional section (typically the last section), they won't be assigned to the article.

This patch strips the categories and applies to the new page, before applying layout directives.

@@ -252,11 +252,15 @@

	// fetch the layout object
	$layout_title = Title::newFromURL ( "Layout:" . $layout_slug );
	$layout_article = new Article ( $layout_title );

+ $layout_cats = '';
+

	/* if the layout article exists, pre-fill the textarea with its
	 * wiki text. if it doesn't exist, do nothing (no error) */
	if ( ( $layout_text = $layout_article->fetchContent() ) !== false ) {

+ $layout_text = UW_Layouts_stripCats($layout_text, $layout_cats);
+ $text .= $layout_cats;
+

		/* break the layout text into sections by splitting
		 * at header level =one= or ==two==, and iterate */
		$nodes = preg_split ( "/^(==?[^=].*)$/mi", $layout_text, - 1, PREG_SPLIT_DELIM_CAPTURE );

@@ -282,3 +286,39 @@

	return true;

}
+
+
+function UW_Layouts_stripCats($texttostrip,&$catsintext){
+ global $wgContLang, $wgOut;
+
+ # Get localised namespace string:
+ $m_catString = strtolower( $wgContLang->getNsText( NS_CATEGORY ) );
+ # The regular expression to find the category links:
+ $m_pattern = "\[\[({$m_catString}|category):(.*?)\]\]";
+
+ $m_replace = "$2";
+ # The container to store the processed text:
+ $m_cleanText = '';
+
+ # Check linewise for category links:
+ foreach( explode( "\n", $texttostrip ) as $m_textLine ) {
+ # Filter line through pattern and store the result:
+ $m_cleanText .= trim( preg_replace( "/{$m_pattern}/i", "", $m_textLine ) ) . "\n";
+
+ # Check if we have found a category, else proceed with next line:
+ if( preg_match_all( "/{$m_pattern}/i", $m_textLine,$catsintext2,PREG_SET_ORDER) ){
+ foreach( $catsintext2 as $local_cat => $m_prefix ) {
+ //Set first letter to upper case to match MediaWiki standard
+ $strFirstLetter = substr($m_prefix[2], 0,1);
+ strtoupper($strFirstLetter);
+ $newString = strtoupper($strFirstLetter) . substr($m_prefix[2], 1);
+ $catsintext .= "[[" . $m_catString . ":" . $newString . "]]\n";
+ }
+ # Get the category link from the original text and store it in our list:
+ preg_replace( "/.*{$m_pattern}/i", $m_replace, $m_textLine,-1,$intNumber );
+ }
+
+ }
+
+ return $m_cleanText;
+}

Version: unspecified
Severity: minor