Changeset 31349 for trunk/src/wp-includes/wp-db.php

Timestamp:

02/06/2015 04:50:19 AM (10 years ago)

Author:

pento

Message:

WPDB: If a site is using the utf8 charset, and their version of MySQL supports utf8mb4, auto-upgrade them to utf8mb4.

This patch also resizes some indexes, to allow for the 767 byte index size limit in standard MySQL installs.

See #21212

File:

• trunk/src/wp-includes/wp-db.php (modified) (8 diffs)

trunk/src/wp-includes/wp-db.php

@@ -625 +625 @@
         }
 
-        $this->init_charset();
-
         $this->dbuser = $dbuser;
         $this->dbpassword = $dbpassword;
@@ -718 +716 @@
         if ( function_exists('is_multisite') && is_multisite() ) {
             $this->charset = 'utf8';
-            if ( defined( 'DB_COLLATE' ) && DB_COLLATE )
+            if ( defined( 'DB_COLLATE' ) && DB_COLLATE ) {
                 $this->collate = DB_COLLATE;
-            else
+            } else {
                 $this->collate = 'utf8_general_ci';
+            }
         } elseif ( defined( 'DB_COLLATE' ) ) {
             $this->collate = DB_COLLATE;
         }
 
-        if ( defined( 'DB_CHARSET' ) )
+        if ( defined( 'DB_CHARSET' ) ) {
             $this->charset = DB_CHARSET;
+        }
+
+        if ( ( $this->use_mysqli && ! ( $this->dbh instanceof mysqli ) )
+          || ( empty( $this->dbh ) || ! ( $this->dbh instanceof mysqli ) ) ) {
+            return;
+        }
+
+        if ( 'utf8' === $this->charset && $this->has_cap( 'utf8mb4' ) ) {
+            $this->charset = 'utf8mb4';
+        }
+
+        if ( 'utf8mb4' === $this->charset && ( ! $this->collate || stripos( $this->collate, 'utf8_' ) === 0 ) ) {
+            $this->collate = 'utf8mb4_unicode_ci';
+        }
     }
 
@@ -1477 +1490 @@
             return false;
         } elseif ( $this->dbh ) {
+            if ( ! $this->has_connected ) {
+                $this->init_charset();
+            }
+
             $this->has_connected = true;
+
             $this->set_charset( $this->dbh );
+
             $this->ready = true;
             $this->set_sql_mode();
@@ -2250 +2269 @@
      *
      * @since 4.2.0
-     * @access protected
+     * @access public
      *
      * @param string $table  Table name.
@@ -2257 +2276 @@
      *               character set. {@see WP_Error} object if there was an error.
      */
-    protected function get_col_charset( $table, $column ) {
+    public function get_col_charset( $table, $column ) {
         $tablekey = strtolower( $table );
         $columnkey = strtolower( $column );
@@ -2357 +2376 @@
             'ujis'    => 'EUC-JP',
             'utf32'   => 'UTF-32',
-            'utf8mb4' => 'UTF-8',
         );
 
@@ -2392 +2410 @@
             }
 
-            // utf8(mb3) can be handled by regex, which is a bunch faster than a DB lookup.
-            if ( 'utf8' === $charset || 'utf8mb3' === $charset ) {
+            // utf8 can be handled by regex, which is a bunch faster than a DB lookup.
+            if ( 'utf8' === $charset || 'utf8mb3' === $charset || 'utf8mb4' === $charset ) {
                 $regex = '/
                     (
@@ -2401 +2419 @@
                         |   [\xE1-\xEC][\x80-\xBF]{2}
                         |   \xED[\x80-\x9F][\x80-\xBF]
-                        |   [\xEE-\xEF][\x80-\xBF]{2}
-                        ){1,50}                          # ...one or more times
+                        |   [\xEE-\xEF][\x80-\xBF]{2}';
+
+                if ( 'utf8mb4' === $charset) {
+                    $regex .= '
+                        |    \xF0[\x90-\xBF][\x80-\xBF]{2} # four-byte sequences   11110xxx 10xxxxxx * 3
+                        |    [\xF1-\xF3][\x80-\xBF]{3}
+                        |    \xF4[\x80-\x8F][\x80-\xBF]{2}
+                    ';
+                }
+
+                $regex .= '){1,50}                          # ...one or more times
                     )
                     | .                                  # anything else