From: Brian Wolff <bawolff@users.mediawiki.org>
Date: Fri, 24 Dec 2010 09:53:08 +0000 (+0000)
Subject: (Bug 26410) In an internal link, a plus sign is treated as a space if the link
X-Git-Tag: 1.31.0-rc.0~33118
X-Git-Url: http://git.cyclocoop.org/%24href?a=commitdiff_plain;h=ff5394caed9fd799aa85ca8c77031bc9852492f1;p=lhc%2Fweb%2Fwiklou.git

(Bug 26410) In an internal link, a plus sign is treated as a space if the link
has a % sign in it, but interperted as a + if no % sign is present.

This changes how the parser interperts links slightly. However:
*I can't imagine anyone is relying on this behaviour
*Things should be consistent. a + sign shouldn't magically change meaning
if there is a % sign somewhere else in the link.
*Pages are allowed to contain % signs in their title, and + signs,
you should be able to link to such pages just by typing there name without
resorting to %2B.
*If you have a page named foo%+ having [[{{PAGENAME}}]] link to a different
page seems inherently wrong.
*The previous behaviour seemed accidental.
---

diff --git a/RELEASE-NOTES b/RELEASE-NOTES
index 0304627666..b2823b04fa 100644
--- a/RELEASE-NOTES
+++ b/RELEASE-NOTES
@@ -40,6 +40,8 @@ it from source control: http://www.mediawiki.org/wiki/Download_from_SVN
 * (bug 18372) File types blacklisted by $wgFileBlacklist will no longer be shown as
   "Permitted file types" on the upload form
 * (bug 26379) importImages.php gives more descriptive error message on failure.
+* (Bug 26410) + signs are no longer treated as spaces in internal links if 
+  link has a % sign in it.
 
 === API changes in 1.18 ===
 * (bug 26339) Throw warning when truncating an overlarge API result
diff --git a/includes/Linker.php b/includes/Linker.php
index 70d2b8a4c0..2d067402a9 100644
--- a/includes/Linker.php
+++ b/includes/Linker.php
@@ -1102,7 +1102,7 @@ class Linker {
 
 		# fix up urlencoded title texts (copied from Parser::replaceInternalLinks)
 		if ( strpos( $match[1], '%' ) !== false ) {
-			$match[1] = str_replace( array( '<', '>' ), array( '&lt;', '&gt;' ), urldecode( $match[1] ) );
+			$match[1] = str_replace( array( '<', '>' ), array( '&lt;', '&gt;' ), rawurldecode( $match[1] ) );
 		}
 
 		# Handle link renaming [[foo|text]] will show link as "text"
diff --git a/includes/Title.php b/includes/Title.php
index 089d2a8c94..a5d9884f9a 100644
--- a/includes/Title.php
+++ b/includes/Title.php
@@ -411,9 +411,7 @@ class Title {
 				// and URL-decode links
 				if ( strpos( $m[1], '%' ) !== false ) {
 					// Match behavior of inline link parsing here;
-					// don't interpret + as " " most of the time!
-					// It might be safe to just use rawurldecode instead, though.
-					$m[1] = urldecode( ltrim( $m[1], ':' ) );
+					$m[1] = rawurldecode( ltrim( $m[1], ':' ) );
 				}
 				$title = Title::newFromText( $m[1] );
 				// If the title is a redirect to bad special pages or is invalid, return null
diff --git a/includes/parser/Parser.php b/includes/parser/Parser.php
index 8a72ba0dc6..95e0ff4b77 100644
--- a/includes/parser/Parser.php
+++ b/includes/parser/Parser.php
@@ -1726,14 +1726,14 @@ class Parser {
 				# fix up urlencoded title texts
 				if ( strpos( $m[1], '%' ) !== false ) {
 					# Should anchors '#' also be rejected?
-					$m[1] = str_replace( array('<', '>'), array('&lt;', '&gt;'), urldecode( $m[1] ) );
+					$m[1] = str_replace( array('<', '>'), array('&lt;', '&gt;'), rawurldecode( $m[1] ) );
 				}
 				$trail = $m[3];
 			} elseif ( preg_match( $e1_img, $line, $m ) ) { # Invalid, but might be an image with a link in its caption
 				$might_be_img = true;
 				$text = $m[2];
 				if ( strpos( $m[1], '%' ) !== false ) {
-					$m[1] = urldecode( $m[1] );
+					$m[1] = rawurldecode( $m[1] );
 				}
 				$trail = "";
 			} else { # Invalid form; output directly
@@ -4507,7 +4507,7 @@ class Parser {
 			}
 
 			if ( strpos( $matches[0], '%' ) !== false ) {
-				$matches[1] = urldecode( $matches[1] );
+				$matches[1] = rawurldecode( $matches[1] );
 			}
 			$tp = Title::newFromText( $matches[1] );
 			$nt =& $tp;