<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=ISO-8859-1">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<font face="Arial">Hello,<br>
<br>
I would like to slice the following string with regexp:<br>
<br>
<font face="Courier New, Courier, monospace">texte = "voir
<a class="moz-txt-link-freetext" href="http://help.scilab.org/docs/5.4.1/fr_FR/regexp.html">http://help.scilab.org/docs/5.4.1/fr_FR/regexp.html</a>";<br>
[iD, iF, matchs, captures] =
regexp(texte,"|[^\s]+\:\/\/([^\/]+\/)+([^\/]*$)|i"); captures<br>
<br>
</font></font>returning<br>
<br>
<font face="Courier New, Courier, monospace"><big><small>!fr_FR/
regexp.html !
</small></big></font><br>
<big><br>
</big>while i would expect <br>
<br>
<font face="Courier New, Courier, monospace">help.scilab.org/ docs/
5.4.1/ fr_FR/ regexp.html<br>
</font><br>
since the capturing parentheses are iterated <i>at leas</i>t once
by the ending + (as {1,} would do) in<br>
<font face="Arial"><font face="Courier New, Courier, monospace">([^\/]+\/)+<br>
<br>
</font></font>This slicing it likely possible with strtok() or
strindex(), but i do not understand why with regexp<br>
the iteration is not performed. Nor with <br>
<font face="Arial"><font face="Courier New, Courier, monospace">([^\/]+?\/)+</font></font><br>
<br>
Any hint would be appreciated<br>
Regards<br>
Samuel<br>
<br>
<font face="Arial"><font face="Courier New, Courier, monospace"></font></font>
</body>
</html>