regexp problem using custom markup

kidd has asked for the wisdom of the Perl Monks concerning the following question:

I have this problem that has been driving me nuts, and I can't find the answer...

Lets say I have this code:

my $text = 'This test creates a [color=blue] blue [/color] word. And a
+ [color=red] red [/color] one';
[download]

As you can see its a string with special command [color=blue] wich I want to substitute for "<font color="blue">" and [/color] to "</font>"...

Now here is the code I've been trying out:

my ($color,@colors);
my $new = $text;

@colors = $new =~ m/\[color=.*\]/gi;

print @colors;
[download]

Now...according to me "@colors" should print out:

[color=blue]
[color=red]
[download]

Wich it isnt happening...could someone help me out... THANKS

Comment on regexp problem using custom markup Select or Download Code

Replies are listed 'Best First'.
Re: Regexp by dws (Chancellor) on Jul 25, 2002 at 20:40 UTC
Try `@colors = $new =~ m/(\[color=.*?\])/gi;` Adding the '?' makes the regex non-greedy. When it's greedy, it will merrily swallow everying until it reaches the last ']' in the target string.	[reply] [d/l]
Re: Regexp by thelenm (Vicar) on Jul 25, 2002 at 20:42 UTC
What you're probably seeing is the greedy behavior of dot-star. `.` will match as much as it possibly can, so in your example, it should match "`blue] blue [/color] word. And a [color=red] red [/color`", not "`blue`". You could change your regular expression to one of the following options: `# non-greedy dot-star @colors = $new =~ m/(\[color=.?\])/gi; # say what you mean: match non-] characters @colors = $new =~ m/(\[color=[^\]]*\])/gi;` [download] -- Mike `-- just,my${.02}`	[reply] [d/l] [select]
Re: Regexp by elusion (Curate) on Jul 25, 2002 at 20:42 UTC
This code should work: `@colors = $new =~ m/(\[color=.?\])/gi;` First off, `@colors` will receive values that are kept by parentheses, and you don't have any in your regex. And second, the . in your regex will make it match `"[color=blue] blue [/color] word. And a [color=red] red [/color]"` all at once because regexes are greedy -- they match as much as possible. By adding the ? behind .* it matches as little as possible. elusion : http://matt.diephouse.com	[reply] [d/l] [select]
Re: Regexp by mkmcconn (Chaplain) on Jul 25, 2002 at 20:40 UTC
You are matching, but you are not capturing. You need the parens around the match. `@colors = $new =~ m/(\[color=.\])/gi;` [download] mkmcconn update* also, read Death to Dot Star!	[reply] [d/l]
Re: Re: Regexp by Anonymous Monk on Jul 25, 2002 at 21:24 UTC
If there are no parentheses, it returns a list of all the matched strings, as if there were parentheses around the whole pattern. -- perlop / Regexp Quote-Like Operators -Anomo	[reply]
Re: Re: Re: Regexp by mkmcconn (Chaplain) on Jul 25, 2002 at 21:36 UTC
Thank you Anomo - you are right. Without the /g operator it will return '1'; but, it will capture just fine using the operator: something that frankly I have never used or noticed. Thanks again. mkmcconn	[reply]
Re: Regexp by kidd (Curate) on Jul 25, 2002 at 20:45 UTC
Thanks for your answers...I added the ? and it worked great...	[reply]
OT: Deprecated Element by BorgCopyeditor (Friar) on Jul 25, 2002 at 21:55 UTC
Not what you were asking for, but maybe informative anyway: `<font>` has been deprecated. Inline styles (or classes) via CSS are preferred, in order to better separate content data (HTML markup) from presentation data (styling). BCE --Your punctuation skills are insufficient!	[reply] [d/l]


Clear questions and runnable code get the best and fastest answer
	PerlMonks