Legoktm has uploaded a new change for review.

  https://gerrit.wikimedia.org/r/177021

Change subject: Add removeInvalidEmails.php maintenance script
......................................................................

Add removeInvalidEmails.php maintenance script

The script scans the user table and removes emails
that are technically invalid which don't pass
Sanitizer::validateEmail(), and wouldn't be allowed
today. Confirmed emails are skipped entirely since
they had to be valid at some point.

Bug: T76512
Change-Id: I3cc6396ff6d8b738846b7716b4b0cddc9bf9e1a4
---
A maintenance/removeInvalidEmails.php
1 file changed, 56 insertions(+), 0 deletions(-)


  git pull ssh://gerrit.wikimedia.org:29418/mediawiki/core 
refs/changes/21/177021/1

diff --git a/maintenance/removeInvalidEmails.php 
b/maintenance/removeInvalidEmails.php
new file mode 100644
index 0000000..e838e1f
--- /dev/null
+++ b/maintenance/removeInvalidEmails.php
@@ -0,0 +1,56 @@
+<?php
+
+require_once __DIR__ . '/Maintenance.php';
+
+/**
+ * A script to remove emails that are invalid from
+ * the user_email column of the user table. Emails
+ * are validated before users can add them, but
+ * this was not always the case so older users may
+ * have invalid ones.
+ */
+class RemoveInvalidEmails extends Maintenance {
+       public function __construct() {
+               parent::__construct();
+               $this->setBatchSize( 500 );
+       }
+       public function execute() {
+               $dbr = $this->getDB( DB_SLAVE );
+               $dbw = $this->getDB( DB_MASTER );
+               do {
+                       $rows = $dbr->select(
+                               'user',
+                               array( 'user_id', 'user_email' ),
+                               array( 'user_email != ""', 
'user_email_authenticated IS NULL' ),
+                               __METHOD__,
+                               array( 'LIMIT' => $this->mBatchSize )
+                       );
+                       $count = $rows->numRows();
+                       $badIds = array();
+                       foreach ( $rows as $row ) {
+                               if ( !Sanitizer::validateEmail( 
$row->user_email ) ) {
+                                       $badIds[] = $row->user_id;
+                               }
+                       }
+
+                       if ( $badIds ) {
+                               $badCount = count( $badIds );
+                               $this->output( "Removing $badCount emails from 
the database.\n" );
+                               $dbw->update(
+                                       'user',
+                                       array( 'user_email' => '' ),
+                                       array( 'user_id' => $badIds ),
+                                       __METHOD__
+                               );
+                               foreach ( $badIds as $badId ) {
+                                       User::newFromId( $badId 
)->invalidateCache();
+                               }
+                               wfWaitForSlaves();
+                       }
+               } while ( $count !== 0 );
+               $this->output( "Done.\n" );
+       }
+}
+
+$maintClass = 'RemoveInvalidEmails';
+require_once RUN_MAINTENANCE_IF_MAIN;
\ No newline at end of file

-- 
To view, visit https://gerrit.wikimedia.org/r/177021
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings

Gerrit-MessageType: newchange
Gerrit-Change-Id: I3cc6396ff6d8b738846b7716b4b0cddc9bf9e1a4
Gerrit-PatchSet: 1
Gerrit-Project: mediawiki/core
Gerrit-Branch: master
Gerrit-Owner: Legoktm <[email protected]>

_______________________________________________
MediaWiki-commits mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits

Reply via email to