From 1603ed2f22411d98fb8f6162af9a28eb5340c87f Mon Sep 17 00:00:00 2001
From: Tim Starling <tstarling@users.mediawiki.org>
Date: Wed, 20 Jan 2010 01:50:16 +0000
Subject: [PATCH] Fixes for r60599: * Split $wgFixArchaicUnicode into two
 separate variables, one for Malayalam and one for Arabic * Clarified
 documentation and switched them both on by default * Removed accidentally
 added variable LanguageAr::$normalizeArray

---
 RELEASE-NOTES                    |  5 +++--
 includes/DefaultSettings.php     | 28 ++++++++++++++++++----------
 languages/classes/LanguageAr.php |  6 ++----
 languages/classes/LanguageMl.php |  4 ++--
 4 files changed, 25 insertions(+), 18 deletions(-)

diff --git a/RELEASE-NOTES b/RELEASE-NOTES
index 3efa86ff8c..a6d71e8675 100644
--- a/RELEASE-NOTES
+++ b/RELEASE-NOTES
@@ -294,8 +294,9 @@ it from source control: http://www.mediawiki.org/wiki/Download_from_SVN
 * (bug 19791) Add URL of file source as comment to thumbs (for ImageMagick)
 * (bug 21946) Sorted wikitables do not properly handle minus signs
 * (bug 18885) Red links for media files do not support shared repositories
-* Added $wgFixArchaicUnicode, which, if enabled, converts some deprecated
-  Unicode sequences in Arabic and Malayalam text to their Unicode 5.1
+* Added $wgFixArabicUnicode, to convert deprecated presentation forms in 
+  Arabic text to their modern equivalents, and $wgFixMalayalamUnicode, to 
+  convert ZWJ-based chillu sequences in Malayalam text to their Unicode 5.1
   equivalents.
 * (bug 22051) Returing false in SpecialContributionsBeforeMainOutput hook now
   stops normal output
diff --git a/includes/DefaultSettings.php b/includes/DefaultSettings.php
index fd38d3ce87..5f9e8982dc 100644
--- a/includes/DefaultSettings.php
+++ b/includes/DefaultSettings.php
@@ -870,17 +870,25 @@ $wgOutputEncoding = 'UTF-8';
 $wgEditEncoding   = '';
 
 /**
- * Set this to true to clean up archaic Unicode sequences in Arabic and
- * Malayalam text. Currently only works if $wgLanguageCode is set to Arabic
- * or Malayalam.
+ * Set this to true to replace Arabic presentation forms with their standard 
+ * forms in the U+0600-U+06FF block. This only works if $wgLanguageCode is
+ * set to "ar".
  *
- * Enabling this is generally a good idea for new wikis, since it fixes a few
- * technical problems to do with editing these languages. However, if it's
- * enabled on an existing wiki, pages which contain the problematic characters
- * in their page titles may become inaccessible. Running maintenance/cleanupTitles.php
- * after enabling it may fix this.
+ * Note that pages with titles containing presentation forms will become 
+ * inaccessible, run maintenance/cleanupTitles.php to fix this.
  */
-$wgFixArchaicUnicode = false;
+$wgFixArabicUnicode = true;
+
+/**
+ * Set this to true to replace ZWJ-based chillu sequences in Malayalam text
+ * with their Unicode 5.1 equivalents. This only works if $wgLanguageCode is 
+ * set to "ml". Note that some clients (even new clients as of 2010) do not 
+ * support these characters. 
+ *
+ * If you enable this on an existing wiki, run maintenance/cleanupTitles.php to
+ * fix any ZWJ sequences in existing page titles.
+ */
+$wgFixMalayalamUnicode = true;
 
 /**
  * Locale for LC_CTYPE, to work around http://bugs.php.net/bug.php?id=45132
@@ -4318,4 +4326,4 @@ $wgOldChangeTagsIndex = false;
  * Extensions can add mwEmbed modules via adding paths to their loader.js to
  * $wgExtensionJavascriptLoader[] = path/to/loader.js
  */
-$wgExtensionJavascriptLoader = array();
\ No newline at end of file
+$wgExtensionJavascriptLoader = array();
diff --git a/languages/classes/LanguageAr.php b/languages/classes/LanguageAr.php
index 2ccc08968d..70a53f8675 100644
--- a/languages/classes/LanguageAr.php
+++ b/languages/classes/LanguageAr.php
@@ -6,8 +6,6 @@
  * @author Niklas Laxström
  */
 class LanguageAr extends Language {
-	var $normalizeArray;
-
 	function convertPlural( $count, $forms ) {
 		if ( !count($forms) ) { return ''; }
 		$forms = $this->preConvertPlural( $forms, 6 );
@@ -36,9 +34,9 @@ class LanguageAr extends Language {
 	 * performance impact of enabling it for all languages.
 	 */
 	function normalize( $s ) {
-		global $wgFixArchaicUnicode;
+		global $wgFixArabicUnicode;
 		$s = parent::normalize( $s );
-		if ( $wgFixArchaicUnicode ) {
+		if ( $wgFixArabicUnicode ) {
 			$s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
 		}
 		return $s;
diff --git a/languages/classes/LanguageMl.php b/languages/classes/LanguageMl.php
index 316ea81a51..98b5fc961c 100644
--- a/languages/classes/LanguageMl.php
+++ b/languages/classes/LanguageMl.php
@@ -12,9 +12,9 @@ class LanguageMl extends Language {
 	 * performance impact of enabling it for all languages.
 	 */
 	function normalize( $s ) {
-		global $wgFixArchaicUnicode;
+		global $wgFixMalayalamUnicode;
 		$s = parent::normalize( $s );
-		if ( $wgFixArchaicUnicode ) {
+		if ( $wgFixMalayalamUnicode ) {
 			$s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
 		}
 		return $s;
-- 
2.20.1