stripping characters from a string

[Anonymous Monk]

hi, i just started learning perl, so sorry if this question may sound very newbish.

I have a string that i want to replace spaces with an under score and remove any '.

i know i need some sort of regex but i really don't understand how to start.

$string = "Good O'l Time";

will look like this with the regex:

Good_Ol_Time

I would appreciate any help or referrence :)

thanks!

[acid06]

You might try reading the perlre and perlretut manpages.
You can follow the links above for an online reading or you can try it on your own shell/command prompt using:


perldoc perlre
perldoc perlretut

[Samy_rio]

Try this,

use strict;
use warnings;

my $string = "Good O'l Time"; 
$string =~ y/ '/\_/d;
print $string;

[ayrnieu]

I'd prefer this, actually:

for ($string) { tr/ '/_/d; print }

[Samy_rio]

ayrnieu, OP's need

Good_Ol_Time

but your output shows

Good_O l_Time

As per perlfunc - Perl builtin functions tutorial.

y///
The transliteration operator. Same as tr///.

Updated: ayrnieu updated his answer without proper information.

[lamp]

Hi,
Please try the following code.


$string = "Good O'l Time"; 
$string =~ s/\ /_/g;
$string =~ s/\'//g;
print $string;


Thanks,
--lamp

[leocharre]

if you're trying to normalize a string so you can make a decent filename out of it- which i get the feeling here (slap me if i'm wrong..)

my $fus=qq|This Is some Funkah Lookin' & Zmellin' "straeyang"|;

$fus=~s/\W/_/g;
# now $fus equals This_Is_some_Funkah_Lookin____Zmellin_ __straeyang_

#great. super. let's clean that up
$fus=~s/_+/_/g;

# now $fup equals This_Is_some_Funkah_Lookin_Zmellin_straeyang_

#let's get rid of any _ in ends
$fus=~s/^_+|_+$//g;

print $fus; # prints This_Is_some_Funkah_Lookin_Zmellin_straeyang

#maybe im tight and i don't want any uppercase
$fus = lc($fus);

#now $fus = this_is_some_funkah_lookin_zmellin_straeyang
#anyway.. you get the idea

Maybe you want to get rid of funny chars? like non word chars? then use \W, that means *any* non word characters, as opposed to (\w which are 0-9, a-z, A-Z, _ anyway..)

[radiantmatrix]

Good post. Here's what I would do if all I wanted was your last result:

# note the lower-casing inline
($fus = lc $fus) =~ s/\W+/_/g; #all spans of 1 or more non-words becom
+e 1 '_'
$fus =~ s/^_|_$//g;            #at most one on each end, because of th
+e above

But, for the poster's question, removing all single-quotes and migrating spaces to _...

for ($string) {
   s/\s+/_/g;    #convert 1+ spaces to a single '_'
   s/'|^_|_$//g; #trim lead/trail _ and remove single quotes
}

Which is going a bit further than the direct solution of

$string =~ tr/\s'/_/d;