On 12 Jun 2001 18:45:23 +0200, Jos Boumans wrote:
> Please, if you try and flame posts, get your facts straight.
>
> 1st: - is a range operator, and hence needs not be escaped when it's not
> indicating a range, ie, at the beginning or end of a []
> so this regex is not 'wrong'. feel free to try it.
I am thourghly ashamed of myself for not checking before posting. I am
sorry.
> 2nd: the regex is purposely written verbose, seeing this is a newbie list and we
> want to teach them, not scare them away by hard to follow code
Since I started reading _Learning Perl_ a month and half ago, just
finished reading up on some of the stranger stuff in _Perl Programming_
(2nd edition, If only I had know 3rd was coming out!), and have barely
begun Conway's _Object Oriented Perl_ I consider myself to be a newbie.
I want to be scared by new and strange things. It encourages me to hit
the docs and figure out what is going on. That said I don't think it is
wrong to give simple answers to simple questions; I just believe that
the simple answers (the ones thay can use immediately) should be
followed up by scarier, tighter answers. This method solves the first
problem (How do I do this?) and gives them (hopefully, although I fudged
it this time by fscking up the - bit) more to think about for the
future.
> on that note: ^\w might confuse, since, when used inside a range ( [] ),
> means something totally different then outside it... hence, for comprehension i
> teach \W instead
> (for those wondering: outside a range, ^ means 'match beginning of
> string', inside a range it means 'not')
>
> so the short version would be:
> if (/[\W.-/]){ print "illegal string" }
Correct me if I am wrong, but I don't think ^\w is the same as \W. It
is my understanding (and I checked the docs on this one <grin />) that
/[\W.-]/ expands to /[[^a-zA-Z0-9_].-]/ which, when translated to
English, means match any non-word character, period, or underbar. As
opposed to /[^\w.-]/ which means, again in English, match any character
that is not a word character, period, or underbar.
<code>
#!/usr/bin/perl -w
use strict;
$_ = "this_should.pass";
print "testing ($_) with ", '/[\W.-]/ = ';
if (/[\W.-]/) { #fixed typo (]/ instead of /])
print "illegal string\n";
} else {
print "Good!\n";
}
print "testing ($_) with ", '/[^\w.-]/ = ';
if (/[^\w.-]/) {
print "you cheated!\n";
} else {
print "Good!\n";
}
</code>
<output>
testing (this_should.pass) with /[\W.-]/ = illegal string
testing (this_should.pass) with /[^\w.-]/ = Good!
</output>
This output would seem to say that /[\W.-]/ is not equal to /[^\w.-]/.
>
> the verbose version would be an explicit check of every character, as described
> below
>
> hope this clears things up,
>
> Jos Boumans
>
> PS html tags end </tag>
Not for single tags (ie <br> and <hr>) in the latest version. Since XML
is now HTML's base (replacing SGML) <br> is illegal. You must say
<br></br> or <br />.
>
<snip />
--
Today is Pungenday, the 17th day of Confusion in the YOLD 3167
Or not.
