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 used in Language->checkTitleEncoding()
  28  * and compares its execution time against that of mb_check_encoding, if available.
  29  *
  30  * @ingroup Benchmark
  31  */
  32 class BenchUtf8TitleCheck extends Benchmarker {
  33         private $canRun;
  34
  35         private $data;
  36
  37         public function __construct() {
  38                 parent::__construct();
  39
  40                 // @codingStandardsIgnoreStart Ignore long line warnings.
  41                 $this->data = array(
  42                         "",
  43                         "United States of America", // 7bit ASCII
  44                         "S%C3%A9rie%20t%C3%A9l%C3%A9vis%C3%A9e",
  45                         "Acteur%7CAlbert%20Robbins%7CAnglais%7CAnn%20Donahue%7CAnthony%20E.%20Zuiker%7CCarol%20Mendelsohn",
  46                         // This comes from bug 36839
  47                         "Acteur%7CAlbert%20Robbins%7CAnglais%7CAnn%20Donahue%7CAnthony%20E.%20Zuiker%7CCarol%20Mendelsohn%7C"
  48                         . "Catherine%20Willows%7CDavid%20Hodges%7CDavid%20Phillips%7CGil%20Grissom%7CGreg%20Sanders%7CHodges%7C"
  49                         . "Internet%20Movie%20Database%7CJim%20Brass%7CLady%20Heather%7C"
  50                         . "Les%20Experts%20(s%C3%A9rie%20t%C3%A9l%C3%A9vis%C3%A9e)%7CLes%20Experts%20:%20Manhattan%7C"
  51                         . "Les%20Experts%20:%20Miami%7CListe%20des%20personnages%20des%20Experts%7C"
  52                         . "Liste%20des%20%C3%A9pisodes%20des%20Experts%7CMod%C3%A8le%20discussion:Palette%20Les%20Experts%7C"
  53                         . "Nick%20Stokes%7CPersonnage%20de%20fiction%7CPersonnage%20fictif%7CPersonnage%20de%20fiction%7C"
  54                         . "Personnages%20r%C3%A9currents%20dans%20Les%20Experts%7CRaymond%20Langston%7CRiley%20Adams%7C"
  55                         . "Saison%201%20des%20Experts%7CSaison%2010%20des%20Experts%7CSaison%2011%20des%20Experts%7C"
  56                         . "Saison%2012%20des%20Experts%7CSaison%202%20des%20Experts%7CSaison%203%20des%20Experts%7C"
  57                         . "Saison%204%20des%20Experts%7CSaison%205%20des%20Experts%7CSaison%206%20des%20Experts%7C"
  58                         . "Saison%207%20des%20Experts%7CSaison%208%20des%20Experts%7CSaison%209%20des%20Experts%7C"
  59                         . "Sara%20Sidle%7CSofia%20Curtis%7CS%C3%A9rie%20t%C3%A9l%C3%A9vis%C3%A9e%7CWallace%20Langham%7C"
  60                         . "Warrick%20Brown%7CWendy%20Simms%7C%C3%89tats-Unis"
  61                 );
  62                 // @codingStandardsIgnoreEnd
  63
  64                 $this->canRun = function_exists( 'mb_check_encoding' );
  65
  66                 if ( $this->canRun ) {
  67                         $this->mDescription = "Benchmark for using a regexp vs. mb_check_encoding " .
  68                                 "to check for UTF-8 encoding.";
  69                         mb_internal_encoding( 'UTF-8' );
  70                 } else {
  71                         $this->mDescription = "CANNOT RUN benchmark using mb_check_encoding: function not available.";
  72                 }
  73         }
  74
  75         public function execute() {
  76                 if ( !$this->canRun ) {
  77                         return;
  78                 }
  79                 $benchmarks = array();
  80                 foreach ( $this->data as $val ) {
  81                         $benchmarks[] = array(
  82                                 'function' => array( $this, 'use_regexp' ),
  83                                 'args' => array( rawurldecode( $val ) )
  84                         );
  85                         $benchmarks[] = array(
  86                                 'function' => array( $this, 'use_regexp_non_capturing' ),
  87                                 'args' => array( rawurldecode( $val ) )
  88                         );
  89                         $benchmarks[] = array(
  90                                 'function' => array( $this, 'use_regexp_once_only' ),
  91                                 'args' => array( rawurldecode( $val ) )
  92                         );
  93                         $benchmarks[] = array(
  94                                 'function' => array( $this, 'use_mb_check_encoding' ),
  95                                 'args' => array( rawurldecode( $val ) )
  96                         );
  97                 }
  98                 $this->bench( $benchmarks );
  99                 print $this->getFormattedResults();
 100         }
 101
 102         private $isutf8;
 103
 104         function use_regexp( $s ) {
 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_regexp_non_capturing( $s ) {
 110                 // Same as above with a non-capturing subgroup.
 111                 $this->isutf8 = preg_match( '/^(?:[\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
 112                         '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
 113         }
 114
 115         function use_regexp_once_only( $s ) {
 116                 // Same as above with a once-only subgroup.
 117                 $this->isutf8 = preg_match( '/^(?>[\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
 118                         '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
 119         }
 120
 121         function use_mb_check_encoding( $s ) {
 122                 $this->isutf8 = mb_check_encoding( $s, 'UTF-8' );
 123         }
 124 }
 125
 126 $maintClass = 'BenchUtf8TitleCheck';
 127 require_once RUN_MAINTENANCE_IF_MAIN;