From 252ef37028af515326bf281bf40ad80792540a2d Mon Sep 17 00:00:00 2001
From: Andrew Garrett <werdna@users.mediawiki.org>
Date: Mon, 9 Feb 2009 23:18:37 +0000
Subject: [PATCH] Apply changes made live on Wikimedia cluster related to
 preprocessor caching to subversion. Patch worked on by Tim Starling and
 myself.

Changes from the original patch (r46936):
* Add versioning to the cache, so the cache can be purged.
* Only cache preprocessor output for input of over a certain amount (default of 1000 bytes).
---
 includes/DefaultSettings.php          |  5 +++
 includes/parser/Preprocessor_DOM.php  | 45 ++++++++++++++++++++-------
 includes/parser/Preprocessor_Hash.php | 43 +++++++++++++++++++------
 3 files changed, 73 insertions(+), 20 deletions(-)

diff --git a/includes/DefaultSettings.php b/includes/DefaultSettings.php
index 949fd819bd..8171586771 100644
--- a/includes/DefaultSettings.php
+++ b/includes/DefaultSettings.php
@@ -3708,3 +3708,8 @@ $wgEnforceHtmlIds = true;
  * false = use Go button & Advanced search link
  */
 $wgUseTwoButtonsSearchForm = true;
+
+/**
+ * Preprocessor caching threshold
+ */
+$wgPreprocessorCacheThreshold = 1000;
\ No newline at end of file
diff --git a/includes/parser/Preprocessor_DOM.php b/includes/parser/Preprocessor_DOM.php
index 66b5e1d3e3..bd33c10249 100644
--- a/includes/parser/Preprocessor_DOM.php
+++ b/includes/parser/Preprocessor_DOM.php
@@ -6,6 +6,8 @@
 class Preprocessor_DOM implements Preprocessor {
 	var $parser, $memoryLimit;
 
+	const CACHE_VERSION = 1;
+
 	function __construct( $parser ) {
 		$this->parser = $parser;
 		$mem = ini_get( 'memory_limit' );
@@ -63,19 +65,37 @@ class Preprocessor_DOM implements Preprocessor {
 	 */
 	function preprocessToObj( $text, $flags = 0 ) {
 		wfProfileIn( __METHOD__ );
+		global $wgMemc, $wgPreprocessorCacheThreshold;
 		
-		global $wgMemc;
-		$cacheKey = wfMemcKey( 'preprocess-xml', md5($text), $flags );
-		
-		if ( $xml = $wgMemc->get( $cacheKey ) ) {
-			// From the cache
-			wfDebugLog( "Preprocessor", "Loaded preprocessor XML from memcached (key $cacheKey)" );
-		} else {
-			$xml = $this->preprocessToXml( $text, $flags );
-			$wgMemc->set( $cacheKey, $xml, 86400 );
-			wfDebugLog( "Preprocessor", "Saved preprocessor XML to memcached (key $cacheKey)" );
+		$xml = false;
+		$cacheable = strlen( $text ) > $wgPreprocessorCacheThreshold;
+		if ( $cacheable ) {
+			wfProfileIn( __METHOD__.'-cacheable' );
+
+			$cacheKey = wfMemcKey( 'preprocess-xml', md5($text), $flags );
+			$cacheValue = $wgMemc->get( $cacheKey );
+			if ( $cacheValue ) {
+				$version = substr( $cacheValue, 0, 8 );
+				if ( intval( $version ) == self::CACHE_VERSION ) {
+					$xml = substr( $cacheValue, 8 );
+					// From the cache
+					wfDebugLog( "Preprocessor", "Loaded preprocessor XML from memcached (key $cacheKey)" );
+				}
+			}
+		}
+		if ( $xml === false ) {
+			if ( $cacheable ) {
+				wfProfileIn( __METHOD__.'-cache-miss' );
+				$xml = $this->preprocessToXml( $text, $flags );
+				$cacheValue = sprintf( "%08d", self::CACHE_VERSION ) . $xml;
+				$wgMemc->set( $cacheKey, $cacheValue, 86400 );
+				wfProfileOut( __METHOD__.'-cache-miss' );
+				wfDebugLog( "Preprocessor", "Saved preprocessor XML to memcached (key $cacheKey)" );
+			} else {
+				$xml = $this->preprocessToXml( $text, $flags );
+			}
+
 		}
-		
 		wfProfileIn( __METHOD__.'-loadXML' );
 		$dom = new DOMDocument;
 		wfSuppressWarnings();
@@ -91,6 +111,9 @@ class Preprocessor_DOM implements Preprocessor {
 		}
 		$obj = new PPNode_DOM( $dom->documentElement );
 		wfProfileOut( __METHOD__.'-loadXML' );
+		if ( $cacheable ) {
+			wfProfileOut( __METHOD__.'-cacheable' );
+		}
 		wfProfileOut( __METHOD__ );
 		return $obj;
 	}
diff --git a/includes/parser/Preprocessor_Hash.php b/includes/parser/Preprocessor_Hash.php
index 0bebdb7f24..f46ee40ce9 100644
--- a/includes/parser/Preprocessor_Hash.php
+++ b/includes/parser/Preprocessor_Hash.php
@@ -8,6 +8,8 @@
  */
 class Preprocessor_Hash implements Preprocessor {
 	var $parser;
+	
+	const CACHE_VERSION = 1;
 
 	function __construct( $parser ) {
 		$this->parser = $parser;
@@ -45,14 +47,30 @@ class Preprocessor_Hash implements Preprocessor {
 	 */
 	function preprocessToObj( $text, $flags = 0 ) {
 		wfProfileIn( __METHOD__ );
+	
+	
+		// Check cache.
+		global $wgMemc, $wgPreprocessorCacheThreshold;
 		
-		global $wgMemc;
-		$cacheKey = wfMemckey( 'preprocessor-hash', md5( $text ), $flags );
-		
-		if ( $obj = $wgMemc->get( $cacheKey ) ) {
-			wfDebugLog( "Preprocessor", "Got preprocessor_hash output from cache" );
-			wfProfileOut( __METHOD__ );
-			return $obj;
+		$cacheable = strlen( $text ) > $wgPreprocessorCacheThreshold;
+		if ( $cacheable ) {
+			wfProfileIn( __METHOD__.'-cacheable' );
+
+			$cacheKey = wfMemcKey( 'preprocess-hash', md5($text), $flags );
+			$cacheValue = $wgMemc->get( $cacheKey );
+			if ( $cacheValue ) {
+				$version = substr( $cacheValue, 0, 8 );
+				if ( intval( $version ) == self::CACHE_VERSION ) {
+					$hash = unserialize( substr( $cacheValue, 8 ) );
+					// From the cache
+					wfDebugLog( "Preprocessor",
+						"Loaded preprocessor hash from memcached (key $cacheKey)" );
+					wfProfileOut( __METHOD__.'-cacheable' );
+					wfProfileOut( __METHOD__ );
+					return $hash;
+				}
+			}
+			wfProfileIn( __METHOD__.'-cache-miss' );
 		}
 
 		$rules = array(
@@ -626,10 +644,17 @@ class Preprocessor_Hash implements Preprocessor {
 		$rootNode = new PPNode_Hash_Tree( 'root' );
 		$rootNode->firstChild = $stack->rootAccum->firstNode;
 		$rootNode->lastChild = $stack->rootAccum->lastNode;
-		wfProfileOut( __METHOD__ );
 		
-		$wgMemc->set( $cacheKey, $rootNode, 86400 );
+		// Cache
+		if ($cacheable) {
+			$cacheValue = sprintf( "%08d", self::CACHE_VERSION ) . serialize( $rootNode );;
+			$wgMemc->set( $cacheKey, $cacheValue, 86400 );
+			wfProfileOut( __METHOD__.'-cache-miss' );
+			wfProfileOut( __METHOD__.'-cacheable' );
+			wfDebugLog( "Preprocessor", "Saved preprocessor Hash to memcached (key $cacheKey)" );
+		}
 		
+		wfProfileOut( __METHOD__ );
 		return $rootNode;
 	}
 }
-- 
2.20.1