From a658eee7fc65b19f6bb937131cdcdd5bb4e941b2 Mon Sep 17 00:00:00 2001
From: Brian Wolff <bawolff@users.mediawiki.org>
Date: Sun, 11 Sep 2011 01:13:08 +0000
Subject: [PATCH] (bug 30722) Add an identity collation that sorts things based
 on what the unicode code point is (aka pre-1.17 behaviour).

I'm tagging this 1.18 because the original bug was for iswiktionary wanting it, so it'd be nice to get it in 1.18.
---
 RELEASE-NOTES-1.18           |  1 +
 includes/AutoLoader.php      |  1 +
 includes/Collation.php       | 26 ++++++++++++++++++++++++++
 includes/DefaultSettings.php |  4 +++-
 4 files changed, 31 insertions(+), 1 deletion(-)

diff --git a/RELEASE-NOTES-1.18 b/RELEASE-NOTES-1.18
index 19daaa1c12..144da2eb4f 100644
--- a/RELEASE-NOTES-1.18
+++ b/RELEASE-NOTES-1.18
@@ -209,6 +209,7 @@ production.
 * Introduced $wgVaryOnXFPForAPI which will cause the API to send
   Vary: X-Forwarded-Proto headers.
 * New maintenance script to refresh image metadata (maintenance/refreshImageMetadata.php)
+* (bug 30722) Add a new collation that sorts categories in ascii sort order.
 
 === Bug fixes in 1.18 ===
 * mw.util.getScript has been implemented (like wfScript in GlobalFunctions.php)
diff --git a/includes/AutoLoader.php b/includes/AutoLoader.php
index d407ca68fa..065a487207 100644
--- a/includes/AutoLoader.php
+++ b/includes/AutoLoader.php
@@ -118,6 +118,7 @@ $wgAutoloadLocalClasses = array(
 	'HttpRequest' => 'includes/HttpFunctions.old.php',
 	'IContextSource' => 'includes/RequestContext.php',
 	'IcuCollation' => 'includes/Collation.php',
+	'IdentityCollation' => 'includes/Collation.php',
 	'ImageGallery' => 'includes/ImageGallery.php',
 	'ImageHistoryList' => 'includes/ImagePage.php',
 	'ImageHistoryPseudoPager' => 'includes/ImagePage.php',
diff --git a/includes/Collation.php b/includes/Collation.php
index a6f1cfe47a..0c510b7830 100644
--- a/includes/Collation.php
+++ b/includes/Collation.php
@@ -23,6 +23,8 @@ abstract class Collation {
 		switch( $collationName ) {
 			case 'uppercase':
 				return new UppercaseCollation;
+			case 'identity':
+				return new IdentityCollation;
 			case 'uca-default':
 				return new IcuCollation( 'root' );
 			default:
@@ -99,6 +101,30 @@ class UppercaseCollation extends Collation {
 	}
 }
 
+/**
+ * Collation class that's essentially a no-op.
+ *
+ * Does sorting based on binary value of the string.
+ * Like how things were pre 1.17.
+ */
+class IdentityCollation extends Collation {
+
+	function getSortKey( $string ) {
+		return $string;
+	}
+
+	function getFirstLetter( $string ) {
+		global $wgContLang;
+		// Copied from UppercaseCollation.
+		// I'm kind of unclear on when this could happen...
+		if ( $string[0] == "\0" ) {
+			$string = substr( $string, 1 );
+		}
+		return $wgContLang->firstChar( $string );
+	}
+}
+
+
 class IcuCollation extends Collation {
 	var $primaryCollator, $mainCollator, $locale;
 	var $firstLetterData;
diff --git a/includes/DefaultSettings.php b/includes/DefaultSettings.php
index b7d9e0012e..a6c1d03ae6 100644
--- a/includes/DefaultSettings.php
+++ b/includes/DefaultSettings.php
@@ -4879,6 +4879,8 @@ $wgCategoryPagingLimit = 200;
  *
  *   - uppercase: Converts the category name to upper case, and sorts by that.
  *
+ *   - identity: Does no conversion. Sorts by binary value of the string.
+ *
  *   - uca-default: Provides access to the Unicode Collation Algorithm with
  *     the default element table. This is a compromise collation which sorts
  *     all languages in a mediocre way. However, it is better than "uppercase".
@@ -4892,7 +4894,7 @@ $wgCategoryPagingLimit = 200;
  * the sort keys in the database.
  *
  * Extensions can define there own collations by subclassing Collation
- * and using the class name as the value of this variable.
+ * and using the Collation::factory hook.
  */
 $wgCategoryCollation = 'uppercase';
 
-- 
2.20.1