Regex Question : Substitution

[azatoth]

Hi Guys -

I have looked in Supersearch, but can't find anything close to my query...although, i couldn't really think of any good search strings either :P

#!/usr/bin/perl -w

use strict;

# variables

my $eng;


# main 

print "String : \n";
($eng = <STDIN>);

($eng =~ s/u/a/i) if ($eng =~ /^.*u.*/g);

print $eng;

The code above does the following. If I enter
"Jump up, Helmut!",
the result is
"Jamp up, Helmut!".
What I want though, is
"Jamp ap, Helmat!".
I.E Substitute every 'u' for an 'a', instead of just the first one.

Any ideas?

Thanks

Azatoth a.k.a Captain Whiplash

Get YOUR PerlMonks Stagename here!

[ChOas]

Hi!

use a 'g' in you substitution, like you do in your search...
And if your searching every sting with a regex to see if you
want to replace it, better just take the chance of replacing it
Anyways... it won't hurt if there are no 'u's there...

$eng =~ s/u/a/i => $eng =~ s/u/a/gio

GreetZ!,

ChOas

print "profeth still\n" if /bird|devil/;

[mirod]

Just replace

$eng =~ s/u/a/i) if ($eng =~ /^.*u.*/g);

by

$eng =~ s/u/a/ig;

There is no need to test whether the substitution will work, just do it.

I think you wanted to write initially was more:

$eng =~ s/u/a/i while ($eng =~ /^.*u.*/g);

but even if this works it is really inneficient. To start with the /^.*u.*/ is strictly equivalent to /u/ except that it is slower (you should read Death to Dot Star! for more on the evils of .*)

[BooK]

Or even shorter, use the transliteration operator:

$eng =~tr/u/a/;

Documentation to be found at: tr and perlop

[$code or die]

For the question he'll need to use $eng =~tr/uU/aA/;

$code or die
Using perl at
The Spiders Web

[japhy]

No, he only wants tr/uU/a/ (which is the same as tr/uU/aa/.

japhy -- Perl and Regex Hacker

[petral]

Er, I think he wants t/Uu/Aa/

p

oopsdate: Never mind. (I gotta keep this set at a larger font.)

[Beatnik]

The funny part is, you're using g in m// while you only need to check once, and you're not using g in s/// when you want all occurances swapped.

Greetz
Beatnik
... Quidquid perl dictum sit, altum viditur.

[dsb]

Try:

$eng =~ s/u/a/gi;

Using the 'g' and 'i' modifiers to: 1. Substitute every instance of 'u' 2. Ignore case - kel -