change [A-Z] to U, [a-z] to L with one regex?

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

Hello dear Monks :) Say, I have following code:

my $r = "SomEThingWithMIXEDRegiSter";
$r =~ s/[A-Z]/U/g;
$r =~ s/[a-z]/L/g;
print $r;
[download]

Is there a way to make changes made by second and third strings with one regex? I mean - change all uppercases to "U" AND all lowercases to "L" at once? Thanks

Comment on change [A-Z] to U, [a-z] to L with one regex? Download Code

Replies are listed 'Best First'.
Re: change [A-Z] to U, [a-z] to L with one regex? by punch_card_don (Curate) on Mar 19, 2013 at 12:10 UTC
It may be useful to stop and ask why one would want to do that. Is there some compelling practical reason? Sometimes we just get caught up in the mindset that it's more elegant to condense as much as possible, but it's not always. The original code `$r =~ s/[A-Z]/U/g; $r =~ s/[a-z]/L/g;` [download] is so crystal clear and simple. It's probably pretty darn efficient too. Why garble it up just to save a line? Time flies like an arrow. Fruit flies like a banana.	[reply] [d/l]
Re: change [A-Z] to U, [a-z] to L with one regex? by BrowserUk (Patriarch) on Mar 19, 2013 at 11:02 UTC
`$r =~ s[([a-zA-Z])][$1 lt '[' ? 'U' : 'L']ge;;` [download] But this will be vastly more efficient: `$r =~ tr[A-Z][U]; $r =~ tr[a-z][L];` [download] With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday' Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error. "Science is about questioning the status quo. Questioning authority". In the absence of evidence, opinion is indistinguishable from prejudice.	[reply] [d/l] [select]
Re: change [A-Z] to U, [a-z] to L with one regex? by hdb (Monsignor) on Mar 19, 2013 at 11:00 UTC
tr is the one that does it: `my $r = "SomEThingWithMIXEDRegiSter"; #$r =~ tr/[A-Z][a-z]/UUUUUUUUUUUUUUUUUUUUUUUUUULLLLLLLLLLLLLLLLLLLLLLL +LLL/; eval( '$r =~ tr/[A-Z][a-z]/'.('U' x 26).('L' x 26).'/' ); print $r;` [download] The string in eval expands to the commented line above it. For every complex problem there is an answer that is clear, simple, and wrong. H. L. Mencken	[reply] [d/l]
Re^2: change [A-Z] to U, [a-z] to L with one regex? by jwkrahn (Abbot) on Mar 19, 2013 at 21:22 UTC
`#$r =~ tr/[A-Z][a-z]/UUUUUUUUUUUUUUUUUUUUUUUUUULLLLLLLLLLLLLLLLLLLLLLL +LLL/; eval( '$r =~ tr/[A-Z][a-z]/'.('U' x 26).('L' x 26).'/' );` [download] `[A-Z]` is 28 characters, not 26, so your expression is wrong. This will work: `$r =~ tr/A-Za-z/UUUUUUUUUUUUUUUUUUUUUUUUUUL/;` [download]	[reply] [d/l] [select]
Re^3: change [A-Z] to U, [a-z] to L with one regex? by BrowserUk (Patriarch) on Mar 19, 2013 at 23:47 UTC
A-Z is 28 characters, not 26, Since when? (Did unicrap fuck that up n'all?) With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday' Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error. "Science is about questioning the status quo. Questioning authority". In the absence of evidence, opinion is indistinguishable from prejudice.	[reply]
Re^4: change [A-Z] to U, [a-z] to L with one regex? by jwkrahn (Abbot) on Mar 20, 2013 at 02:15 UTC
Re^3: change [A-Z] to U, [a-z] to L with one regex? by hdb (Monsignor) on Mar 20, 2013 at 08:27 UTC
Thanks for your comment. I did not realize that in tr you would not need [] and that they would be treated as characters in their own right. And indeed in my script a "Z" would be translated into "L". W/o the brackets it would work. I should read the manuals more often/carefully. For every complex problem there is an answer that is clear, simple, and wrong. H. L. Mencken	[reply]
Re: change [A-Z] to U, [a-z] to L with one regex? by daxim (Curate) on Mar 19, 2013 at 11:07 UTC
There are more than just upper-case and lower-case letter characters. Decide what you want to do with those you haven't considered yet, and put them into the map. `use Unicode::UCD qw(charinfo); my %category_map = ( Lu => 'U', # letter, upper Ll => 'L', # letter, lower ); my $r = 'SomEThingWithMIXEDRegiSter'; $r =~ s { (.) # capture the character } { $category_map{ charinfo(ord $1)->{category} } }egmsx; # ULLUULLLLULLLUUUUUULLLULLL` [download]	[reply] [d/l]
Re^2: change [A-Z] to U, [a-z] to L with one regex? by ikegami (Patriarch) on Mar 20, 2013 at 07:00 UTC
Simpler solution not limited to non-ASCII chars: `s/(\p{Lu})\|\p{Ll}/ defined($1) ? 'U' : 'L' /eg;` [download]	[reply] [d/l]
Re^2: change [A-Z] to U, [a-z] to L with one regex? by LanX (Saint) on Mar 20, 2013 at 01:20 UTC
what about punctuation and numbers? =) my 2¢ `DB<141> $r = 'SomEThingWithMIXEDRegiSter;.?=356'; => "SomEThingWithMIXEDRegiSter;.?=356" DB<142> $r =~ s/(.)/ $1 eq "\u$1" ? ($1 eq "\l$1" ? $1 : "U" ) : ($1 + eq "\l$1" ? "L" : "X" ) /ge => 33 DB<143> $r => "ULLUULLLLULLLUUUUUULLLULLL;.?=356"` [download] Cheers Rolf ( addicted to the Perl Programming Language)	[reply] [d/l]
Re^3: change [A-Z] to U, [a-z] to L with one regex? by ikegami (Patriarch) on Mar 20, 2013 at 06:58 UTC
I'm pretty sure `$1 eq "\u$1"` can fail for some characters it shouldn't fail for.	[reply] [d/l]
Re: change [A-Z] to U, [a-z] to L with one regex? by rjt (Curate) on Mar 19, 2013 at 23:05 UTC
You can always use `$_`: `$_ = "SomEThingWithMIXEDRegiSter\n"; y/A-Z/U/, y/a-z/L/; print;` [download] Yes, this is technically still two expressions, but as others have stated, it'll be efficient, and to me, anyway, the comma operator has good semantic meaning, here, even though a semicolon would do just as well. Update: Corrected bracket typo in y///. Thanks jwkrahn.	[reply] [d/l] [select]
Re^2: change [A-Z] to U, [a-z] to L with one regex? by jwkrahn (Abbot) on Mar 20, 2013 at 05:32 UTC
`y/[A-Z]/U/, y/[a-z]/L/;` [download] Doctrin didn't ask to change the characters `[` or `]` to `U` or `L` `y/A-Z/U/, y/a-z/L/;` [download]	[reply] [d/l] [select]

Back to Seekers of Perl Wisdom