maintenance/benchmarks/bench_utf8_title_check.php

   1 <?php
   2 /**
   3  * Benchmark for using a regexp vs. mb_check_encoding to check for UTF-8 encoding.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup Benchmark
  22  */
  23
  24 require_once __DIR__ . '/Benchmarker.php';
  25
  26 /**
  27  * This little benchmark executes the regexp formerly used in Language->checkTitleEncoding()
  28  * and compares its execution time against that of mb_check_encoding.
  29  *
  30  * @ingroup Benchmark
  31  */
  32 class BenchUtf8TitleCheck extends Benchmarker {
  33         private $data;
  34
  35         public function __construct() {
  36                 parent::__construct();
  37
  38                 // @codingStandardsIgnoreStart Ignore long line warnings.
  39                 $this->data = [
  40                         "",
  41                         "United States of America", // 7bit ASCII
  42                         "S%C3%A9rie%20t%C3%A9l%C3%A9vis%C3%A9e",
  43                         "Acteur%7CAlbert%20Robbins%7CAnglais%7CAnn%20Donahue%7CAnthony%20E.%20Zuiker%7CCarol%20Mendelsohn",
  44                         // This comes from bug 36839
  45                         "Acteur%7CAlbert%20Robbins%7CAnglais%7CAnn%20Donahue%7CAnthony%20E.%20Zuiker%7CCarol%20Mendelsohn%7C"
  46                         . "Catherine%20Willows%7CDavid%20Hodges%7CDavid%20Phillips%7CGil%20Grissom%7CGreg%20Sanders%7CHodges%7C"
  47                         . "Internet%20Movie%20Database%7CJim%20Brass%7CLady%20Heather%7C"
  48                         . "Les%20Experts%20(s%C3%A9rie%20t%C3%A9l%C3%A9vis%C3%A9e)%7CLes%20Experts%20:%20Manhattan%7C"
  49                         . "Les%20Experts%20:%20Miami%7CListe%20des%20personnages%20des%20Experts%7C"
  50                         . "Liste%20des%20%C3%A9pisodes%20des%20Experts%7CMod%C3%A8le%20discussion:Palette%20Les%20Experts%7C"
  51                         . "Nick%20Stokes%7CPersonnage%20de%20fiction%7CPersonnage%20fictif%7CPersonnage%20de%20fiction%7C"
  52                         . "Personnages%20r%C3%A9currents%20dans%20Les%20Experts%7CRaymond%20Langston%7CRiley%20Adams%7C"
  53                         . "Saison%201%20des%20Experts%7CSaison%2010%20des%20Experts%7CSaison%2011%20des%20Experts%7C"
  54                         . "Saison%2012%20des%20Experts%7CSaison%202%20des%20Experts%7CSaison%203%20des%20Experts%7C"
  55                         . "Saison%204%20des%20Experts%7CSaison%205%20des%20Experts%7CSaison%206%20des%20Experts%7C"
  56                         . "Saison%207%20des%20Experts%7CSaison%208%20des%20Experts%7CSaison%209%20des%20Experts%7C"
  57                         . "Sara%20Sidle%7CSofia%20Curtis%7CS%C3%A9rie%20t%C3%A9l%C3%A9vis%C3%A9e%7CWallace%20Langham%7C"
  58                         . "Warrick%20Brown%7CWendy%20Simms%7C%C3%89tats-Unis"
  59                 ];
  60                 // @codingStandardsIgnoreEnd
  61
  62                 $this->addDescription( "Benchmark for using a regexp vs. mb_check_encoding " .
  63                         "to check for UTF-8 encoding." );
  64         }
  65
  66         public function execute() {
  67                 $benchmarks = [];
  68                 foreach ( $this->data as $val ) {
  69                         $benchmarks[] = [
  70                                 'function' => [ $this, 'use_regexp' ],
  71                                 'args' => [ rawurldecode( $val ) ]
  72                         ];
  73                         $benchmarks[] = [
  74                                 'function' => [ $this, 'use_regexp_non_capturing' ],
  75                                 'args' => [ rawurldecode( $val ) ]
  76                         ];
  77                         $benchmarks[] = [
  78                                 'function' => [ $this, 'use_regexp_once_only' ],
  79                                 'args' => [ rawurldecode( $val ) ]
  80                         ];
  81                         $benchmarks[] = [
  82                                 'function' => [ $this, 'use_mb_check_encoding' ],
  83                                 'args' => [ rawurldecode( $val ) ]
  84                         ];
  85                 }
  86                 $this->bench( $benchmarks );
  87                 print $this->getFormattedResults();
  88         }
  89
  90         private $isutf8;
  91
  92         function use_regexp( $s ) {
  93                 $this->isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
  94                         '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
  95         }
  96
  97         function use_regexp_non_capturing( $s ) {
  98                 // Same as above with a non-capturing subgroup.
  99                 $this->isutf8 = preg_match( '/^(?:[\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
 100                         '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
 101         }
 102
 103         function use_regexp_once_only( $s ) {
 104                 // Same as above with a once-only subgroup.
 105                 $this->isutf8 = preg_match( '/^(?>[\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
 106                         '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
 107         }
 108
 109         function use_mb_check_encoding( $s ) {
 110                 $this->isutf8 = mb_check_encoding( $s, 'UTF-8' );
 111         }
 112 }
 113
 114 $maintClass = 'BenchUtf8TitleCheck';
 115 require_once RUN_MAINTENANCE_IF_MAIN;