wikimedia/mediawiki-extensions-Linter
1
0
Fork 0
mirror of https://gerrit.wikimedia.org/r/mediawiki/extensions/Linter synced 2024-11-23 23:44:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

Use estimateRowCount() instead of actually counting everything

Browse source 
On large wikis with lots of lint errors, counting the entire table can
be problematic from a performance perspective, sometimes taking minutes.

Instead, use Database::estimateRowCount(), which uses EXPLAIN SELECT
COUNT(*) to get an approximate value for the number of rows. We do make
sure that if the category actually has no rows, that it will return 0.

This should be considered a temporary solution, and we should look into
doing something like the SiteStats incremental updates table in the long
run.

Bug: T184280
Change-Id: I2d4dcc615477fd60e41dfed4a3d1a3ad52a9f4af
This commit is contained in:
Kunal Mehta 2018-02-01 12:59:08 -08:00
parent 7ba5f25eeb
commit db5e5e9003
1 changed files with 55 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

58
includes/Database.php
View file

@ -230,14 +230,53 @@ class Database {
* @return int[]
*/
public function getTotalsForPage() {
return $this->getTotals( [ 'linter_page' => $this->pageId ] );
return $this->getTotalsAccurate( [ 'linter_page' => $this->pageId ] );
}
/**
* @param array $conds Query conditions
* Get an estimate of how many rows are there for the
* specified category with EXPLAIN SELECT COUNT(*).
* If the category actually has no rows, then 0 will
* be returned.
*
* @param int $catId
*
* @return int
*/
private function getTotalsEstimate( $catId ) {
$dbr = wfGetDB( DB_REPLICA );
// First see if there are no rows, since the distinction
// between 0 and 1 is important. And estimateRowCount seems
// to never return 0.
$rows = $dbr->selectRowCount(
'linter',
'*',
[ 'linter_cat' => $catId ],
__METHOD__,
[ 'LIMIT' => 1 ]
);
if ( $rows === 0 ) {
return 0;
}
// Now we can just estimate
return $dbr->estimateRowCount(
'linter',
'*',
[ 'linter_cat' => $catId ],
__METHOD__
);
}
/**
* This uses COUNT(*), which is accurate, but can be significantly
* slower depending upon how many rows are in the database.
*
* @param array $conds
*
* @return int[]
*/
public function getTotals( $conds = [] ) {
private function getTotalsAccurate( $conds = [] ) {
$rows = wfGetDB( DB_REPLICA )->select(
'linter',
[ 'linter_cat', 'COUNT(*) AS count' ],
@ -260,4 +299,17 @@ class Database {
return $ret;
}
/**
* @return int[]
*/
public function getTotals() {
$ret = [];
foreach ( $this->categoryManager->getVisibleCategories() as $cat ) {
$id = $this->categoryManager->getCategoryId( $cat );
$ret[$cat] = $this->getTotalsEstimate( $id );
}
return $ret;
}
}