yogsototh/scratch
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
scratch/output/Scratch/fr/blog/2010-02-23-When-regexp-is-not-the-best-solution/index.html
Yann Esposito (Yogsototh) d11d3cba86 New blog entry mail from command line
2010-08-31 15:06:43 +02:00

270 lines
No EOL
16 KiB
HTML
Raw Blame History

<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="fr" xml:lang="fr">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
<meta name="keywords" content="programmation, regexp, expressions régulières, extension, fichier">
<link rel="shortcut icon" type="image/x-icon" href="/Scratch/img/favicon.ico" />
<link rel="stylesheet" type="text/css" href="/Scratch/assets/css/main.css" />
<link rel="stylesheet" type="text/css" href="/Scratch/css/twilight.css" />
<link rel="stylesheet" type="text/css" href="/Scratch/css/idc.css" />
<link rel="alternate" type="application/rss+xml" title="RSS" href="http://feeds.feedburner.com/yannespositocomfr"/>
<link rel="alternate" lang="fr" xml:lang="fr" title="Quand se passer des expressions régulières ?" type="text/html" hreflang="fr" href="/Scratch/fr/blog/2010-02-23-When-regexp-is-not-the-best-solution/" />
<link rel="alternate" lang="en" xml:lang="en" title="When regexp is not the best solution" type="text/html" hreflang="en" href="/Scratch/en/blog/2010-02-23-When-regexp-is-not-the-best-solution/" />
<script type="text/javascript" src="/Scratch/js/jquery-1.3.1.min.js"></script>
<script type="text/javascript" src="/Scratch/js/jquery.cookie.js"></script>
<script type="text/javascript" src="/Scratch/js/index.js"></script>
<title>Quand se passer des expressions régulières ?</title>
</head>
<body lang="fr">
<script type="text/javascript">// <![CDATA[
document.write('<div id="blackpage"><img src="/Scratch/img/loading.gif" alt="Chargement en cours..."/></div>');
// ]]>
</script>
<div id="content">
<div id="choix">
<div class="return"><a href="#entete">&darr; Menu &darr;</a></div>
<div id="choixlang">
<a href="/Scratch/en/blog/2010-02-23-When-regexp-is-not-the-best-solution/" onclick="setLanguage('en')">Switch to English</a>
</div>
</div>
<div id="titre">
<h1>
Quand se passer des expressions régulières ?
</h1>
</div>
<div class="flush"></div>
<div class="flush"></div>
<div id="afterheader">
<div class="corps">
<p>Les expressions régulières sont très utiles. Cependant, elles ne sont pas toujours la meilleure manière d&rsquo;aborder certain problème autour des chaines de caractères.
Et surtout quand les transformations que vous voulez accomplir sont simples.</p>
<p>Je voulais savoir comment récupérer le plus vite possible l&rsquo;extension d&rsquo;un nom de fichier. Il y a trois manière naturelle d&rsquo;accomplir celà&nbsp;:</p>
<div><pre class="twilight">
<span class="Comment"><span class="Comment">#</span> regexp</span>
str.<span class="Entity">match</span>(<span class="StringRegexp"><span class="StringRegexp">/</span></span><span class="StringRegexp"><span class="StringRegexp"><span class="StringRegexp">[</span>^.<span class="StringRegexp">]</span></span>*$</span><span class="StringRegexp"><span class="StringRegexp">/</span></span>);
ext<span class="Keyword">=</span><span class="Variable"><span class="Variable">$</span>&amp;</span>
<span class="Comment"><span class="Comment">#</span> split</span>
ext<span class="Keyword">=</span>str.<span class="Entity">split</span>(<span class="String"><span class="String">'</span>.<span class="String">'</span></span>)[<span class="Keyword">-</span><span class="Constant">1</span>]
<span class="Comment"><span class="Comment">#</span> File module</span>
ext<span class="Keyword">=</span><span class="Support">File</span>.<span class="Entity">extname</span>(str)
</pre></div>
<p>A première vue, je pensais que l&rsquo;expression régulière serait plus rapide que le <code>split</code> parce qu&rsquo;il pouvait y avoir plusieurs de <code>.</code> dans un nom de fichier. Mais la majorité du temps il n&rsquo;y a qu&rsquo;un seul point par nom de fichier. C&rsquo;est pourquoi j&rsquo;ai réalisé que le <code>split</code> serait plus rapide. Mais pas le plus rapide possible. Il y a une fonction qui est dédiée à faire ce travail dans un module standard de ruby&nbsp;; le module <code>File</code>.</p>
<p>Voici le code pour faire un benchmark&nbsp;:</p>
<div><div class="code"><div class="file"><a href="/Scratch/fr/blog/2010-02-23-When-regexp-is-not-the-best-solution/code/regex_benchmark_ext.rb"> &#x27A5; regex_benchmark_ext.rb </a></div><div class="withfile">
<pre class="twilight">
<span class="Comment"><span class="Comment">#</span>!/usr/bin/env ruby</span>
<span class="Keyword">require</span> <span class="String"><span class="String">'</span>benchmark<span class="String">'</span></span>
n<span class="Keyword">=</span><span class="Constant">80000</span>
tab<span class="Keyword">=</span>[ <span class="String"><span class="String">'</span>/accounts/user.json<span class="String">'</span></span>,
<span class="String"><span class="String">'</span>/accounts/user.xml<span class="String">'</span></span>,
<span class="String"><span class="String">'</span>/user/titi/blog/toto.json<span class="String">'</span></span>,
<span class="String"><span class="String">'</span>/user/titi/blog/toto.xml<span class="String">'</span></span> ]
puts <span class="String"><span class="String">&quot;</span>Get extname<span class="String">&quot;</span></span>
<span class="Support">Benchmark</span>.<span class="Entity">bm</span> <span class="Keyword">do </span>|<span class="Variable">x</span>|
x.<span class="Entity">report</span>(<span class="String"><span class="String">&quot;</span>regexp:<span class="String">&quot;</span></span>) { n.<span class="Entity">times</span> <span class="Keyword">do </span>
str<span class="Keyword">=</span>tab[<span class="Entity">rand</span>(<span class="Constant">4</span>)];
str.<span class="Entity">match</span>(<span class="StringRegexp"><span class="StringRegexp">/</span></span><span class="StringRegexp"><span class="StringRegexp"><span class="StringRegexp">[</span>^.<span class="StringRegexp">]</span></span>*$</span><span class="StringRegexp"><span class="StringRegexp">/</span></span>);
ext<span class="Keyword">=</span><span class="Variable"><span class="Variable">$</span>&amp;</span>;
<span class="Keyword">end</span> }
x.<span class="Entity">report</span>(<span class="String"><span class="String">&quot;</span> split:<span class="String">&quot;</span></span>) { n.<span class="Entity">times</span> <span class="Keyword">do </span>
str<span class="Keyword">=</span>tab[<span class="Entity">rand</span>(<span class="Constant">4</span>)];
ext<span class="Keyword">=</span>str.<span class="Entity">split</span>(<span class="String"><span class="String">'</span>.<span class="String">'</span></span>)[<span class="Keyword">-</span><span class="Constant">1</span>]&nbsp;;
<span class="Keyword">end</span> }
x.<span class="Entity">report</span>(<span class="String"><span class="String">&quot;</span> File:<span class="String">&quot;</span></span>) { n.<span class="Entity">times</span> <span class="Keyword">do </span>
str<span class="Keyword">=</span>tab[<span class="Entity">rand</span>(<span class="Constant">4</span>)];
ext<span class="Keyword">=</span><span class="Support">File</span>.<span class="Entity">extname</span>(str);
<span class="Keyword">end</span> }
<span class="Keyword">end</span>
</pre>
</div></div></div>
<p>Et voici les résultats&nbsp;:</p>
<pre class="twilight">
Get extname
user system total real
regexp: 2.550000 0.020000 2.570000 ( 2.693407)
split: 1.080000 0.050000 1.130000 ( 1.190408)
File: 0.640000 0.030000 0.670000 ( 0.717748)
</pre>
<p>En conclusion, les fonction dédiées sont meilleures que votre façon de faire (la plupart du temps).</p>
<h2 id="chemin-complet-dun-fichier-sans-lextension">Chemin complet d&rsquo;un fichier sans l&rsquo;extension</h2>
<div><div class="code"><div class="file"><a href="/Scratch/fr/blog/2010-02-23-When-regexp-is-not-the-best-solution/code/regex_benchmark_strip.rb"> &#x27A5; regex_benchmark_strip.rb </a></div><div class="withfile">
<pre class="twilight">
<span class="Comment"><span class="Comment">#</span>!/usr/bin/env ruby</span>
<span class="Keyword">require</span> <span class="String"><span class="String">'</span>benchmark<span class="String">'</span></span>
n<span class="Keyword">=</span><span class="Constant">80000</span>
tab<span class="Keyword">=</span>[ <span class="String"><span class="String">'</span>/accounts/user.json<span class="String">'</span></span>,
<span class="String"><span class="String">'</span>/accounts/user.xml<span class="String">'</span></span>,
<span class="String"><span class="String">'</span>/user/titi/blog/toto.json<span class="String">'</span></span>,
<span class="String"><span class="String">'</span>/user/titi/blog/toto.xml<span class="String">'</span></span> ]
puts <span class="String"><span class="String">&quot;</span>remove extension<span class="String">&quot;</span></span>
<span class="Support">Benchmark</span>.<span class="Entity">bm</span> <span class="Keyword">do </span>|<span class="Variable">x</span>|
x.<span class="Entity">report</span>(<span class="String"><span class="String">&quot;</span> File:<span class="String">&quot;</span></span>) { n.<span class="Entity">times</span> <span class="Keyword">do </span>
str<span class="Keyword">=</span>tab[<span class="Entity">rand</span>(<span class="Constant">4</span>)];
path<span class="Keyword">=</span><span class="Support">File</span>.<span class="Entity">expand_path</span>(str,<span class="Support">File</span>.<span class="Entity">basename</span>(str,<span class="Support">File</span>.<span class="Entity">extname</span>(str)));
<span class="Keyword">end</span> }
x.<span class="Entity">report</span>(<span class="String"><span class="String">&quot;</span>chomp:<span class="String">&quot;</span></span>) { n.<span class="Entity">times</span> <span class="Keyword">do </span>
str<span class="Keyword">=</span>tab[<span class="Entity">rand</span>(<span class="Constant">4</span>)];
ext<span class="Keyword">=</span><span class="Support">File</span>.<span class="Entity">extname</span>(str);
path<span class="Keyword">=</span>str.<span class="Entity">chomp</span>(ext);
<span class="Keyword">end</span> }
<span class="Keyword">end</span>
</pre>
</div></div></div>
<p>et voici les résultats&nbsp;:</p>
<pre class="twilight">
remove extension
user system total real
File: 0.970000 0.060000 1.030000 ( 1.081398)
chomp: 0.820000 0.040000 0.860000 ( 0.947432)
</pre>
<p>En conclusion du ce second benchmark. Un fonction simple est meilleure que trois fonctions dédiées. Pas de surprise, mais c&rsquo;est toujours bien de savoir.</p>
</div>
<div id="choixrss">
<a id="rss" href="http://feeds.feedburner.com/yannespositocomfr">
s'abonner
</a>
</div>
<script type="text/javascript">
$(document).ready(function(){
$('#comment').hide();
$('#clickcomment').click(showComments);
});
function showComments() {
$('#comment').show();
$('#clickcomment').fadeOut();
}
document.write('<div id="clickcomment">Commentaires</div>');
</script>
<div class="flush"></div>
<div class="corps" id="comment">
<h2 class="first">commentaires</h2>
<noscript>
</noscript>
<script type="text/javascript">
var idcomments_acct = 'a307f0044511ff1b5cfca573fc0a52e7';
var idcomments_post_id = '/Scratch/fr/blog/2010-02-23-When-regexp-is-not-the-best-solution/';
var idcomments_post_url = 'http://yannesposito.com/Scratch/fr/blog/2010-02-23-When-regexp-is-not-the-best-solution/';
</script>
<span id="IDCommentsPostTitle" style="display:none"></span>
<script type='text/javascript' src='/Scratch/js/genericCommentWrapperV2.js'></script>
</div>
<div id="entete" class="corps_spaced">
<div id="liens">
<ul><li><a href="/Scratch/fr/">Acceuil</a></li>
<li><a href="/Scratch/fr/blog/">Blog</a></li>
<li><a href="/Scratch/fr/about/">À propos</a></li>
<li><a href="/Scratch/fr/contact/">Contact</a></li></ul>
</div>
<div class="flush"></div>
<hr/>
<div id="next_before_articles">
<div id="previous_articles">
articles précédents
<div class="previous_article">
<a href="/Scratch/fr/blog/2010-02-18-split-a-file-by-keyword/">&larr; découper un fichier par mots clés</a>
</div>
<div class="previous_article">
<a href="/Scratch/fr/blog/2010-02-16-All-but-something-regexp--2-/">&larr; Tout sauf quelquechose en expression régulière.</a>
</div>
<div class="previous_article">
<a href="/Scratch/fr/blog/2010-02-15-All-but-something-regexp/">&larr; Expression régulière pour tout sauf quelquechose</a>
</div>
</div>
<div id="next_articles">
articles suivants
<div class="next_article">
<a href="/Scratch/fr/blog/2010-03-22-Git-Tips/">Astuces Git&rarr; </a>
</div>
<div class="next_article">
<a href="/Scratch/fr/blog/2010-03-23-Encapsulate-git/">Encapsuler git&rarr; </a>
</div>
<div class="next_article">
<a href="/Scratch/fr/blog/2010-05-17-at-least-this-blog-revive/">Je reviens à la vie !&rarr; </a>
</div>
</div>
<div class="flush"></div>
</div>
</div>
<div id="bottom">
<div>
<a rel="license" href="http://creativecommons.org/licenses/by-sa/3.0/deed.fr">Droits de reproduction ©, Yann Esposito</a>
</div>
<div id="lastmod">
Écrit le : 23/02/2010
</div>
<div>
Site entièrement réalisé avec
<a href="http://www.vim.org">Vim</a>
et
<a href="http://nanoc.stoneship.org">nanoc</a>
</div>
<div>
<a href="/Scratch/fr/validation/">Validation</a>
<a href="http://validator.w3.org/check?uri=referer"> [xhtml] </a>
.
<a href="http://jigsaw.w3.org/css-validator/check/referer?profile=css3"> [css] </a>
.
<a href="http://validator.w3.org/feed/check.cgi?url=http%3A//yannesposito.com/Scratch/fr/blog/feed/feed.xml">[rss]</a>
</div>
</div>
<div class="clear"></div>
</div>
</body>
</html>
Reference in a new issue View git blame Copy permalink