From: Aaron Schulz <aschulz@wikimedia.org>
Date: Tue, 23 Apr 2019 23:59:35 +0000 (-0700)
Subject: Reinstate small category refresh logic in LinksDeletionUpdate
X-Git-Tag: 1.34.0-rc.0~1881^2
X-Git-Url: https://git.heureux-cyclage.org/?a=commitdiff_plain;h=d2331a68230d9e1793422f16fcd029f27f55e2cb;p=lhc%2Fweb%2Fwiklou.git

Reinstate small category refresh logic in LinksDeletionUpdate

Add new Category::refreshCountsIfSmall() method that will do a non-locking
SELECT with LIMIT before deciding whether to do a full locking SELECT and
refresh. Call this from LinksDeletionUpdate.

Bug: T18036
Change-Id: I9de8311565988453b8e29a7f3d95d758182fcec1
---

diff --git a/includes/Category.php b/includes/Category.php
index 77f1212fcf..34ac0e1936 100644
--- a/includes/Category.php
+++ b/includes/Category.php
@@ -41,6 +41,8 @@ class Category {
 	const LOAD_ONLY = 0;
 	const LAZY_INIT_ROW = 1;
 
+	const ROW_COUNT_SMALL = 100;
+
 	private function __construct() {
 	}
 
@@ -431,28 +433,57 @@ class Category {
 	 * @since 1.32
 	 */
 	public function refreshCountsIfEmpty() {
+		return $this->refreshCountsIfSmall( 0 );
+	}
+
+	/**
+	 * Call refreshCounts() if there are few entries in the categorylinks table
+	 *
+	 * Due to lock errors or other failures, the precomputed counts can get out of sync,
+	 * making it hard to know when to delete the category row without checking the
+	 * categorylinks table.
+	 *
+	 * This method will do a non-locking select first to reduce contention.
+	 *
+	 * @param int $maxSize Only refresh if there are this or less many backlinks
+	 * @return bool Whether links were refreshed
+	 * @since 1.34
+	 */
+	public function refreshCountsIfSmall( $maxSize = self::ROW_COUNT_SMALL ) {
 		$dbw = wfGetDB( DB_MASTER );
+		$dbw->startAtomic( __METHOD__ );
 
-		$hasLink = $dbw->selectField(
+		$typeOccurances = $dbw->selectFieldValues(
 			'categorylinks',
-			'1',
+			'cl_type',
 			[ 'cl_to' => $this->getName() ],
-			__METHOD__
+			__METHOD__,
+			[ 'LIMIT' => $maxSize + 1 ]
 		);
-		if ( !$hasLink ) {
-			$this->refreshCounts(); // delete any category table entry
 
-			return true;
+		if ( !$typeOccurances ) {
+			$doRefresh = true; // delete any category table entry
+		} elseif ( count( $typeOccurances ) <= $maxSize ) {
+			$countByType = array_count_values( $typeOccurances );
+			$doRefresh = !$dbw->selectField(
+				'category',
+				'1',
+				[
+					'cat_title' => $this->getName(),
+					'cat_pages' => $countByType['page'] ?? 0,
+					'cat_subcats' => $countByType['subcat'] ?? 0,
+					'cat_files' => $countByType['file'] ?? 0
+				],
+				__METHOD__
+			);
+		} else {
+			$doRefresh = false; // category is too big
 		}
 
-		$hasBadRow = $dbw->selectField(
-			'category',
-			'1',
-			[ 'cat_title' => $this->getName(), 'cat_pages <= 0' ],
-			__METHOD__
-		);
-		if ( $hasBadRow ) {
-			$this->refreshCounts(); // clean up this row
+		$dbw->endAtomic( __METHOD__ );
+
+		if ( $doRefresh ) {
+			$this->refreshCounts(); // update the row
 
 			return true;
 		}
diff --git a/includes/deferred/LinksDeletionUpdate.php b/includes/deferred/LinksDeletionUpdate.php
index 0e24134770..8bc9dfa589 100644
--- a/includes/deferred/LinksDeletionUpdate.php
+++ b/includes/deferred/LinksDeletionUpdate.php
@@ -73,7 +73,7 @@ class LinksDeletionUpdate extends LinksUpdate implements EnqueueableDataUpdate {
 			// T166757: do the update after the main job DB commit
 			DeferredUpdates::addCallableUpdate( function () use ( $title ) {
 				$cat = Category::newFromName( $title->getDBkey() );
-				$cat->refreshCountsIfEmpty();
+				$cat->refreshCountsIfSmall();
 			} );
 		}