Path: utzoo!utgpu!news-server.csri.toronto.edu!torsqnt!hybrid!scifi!bywater!uunet!convex!news
From: tchrist@convex.COM (Tom Christiansen)
Newsgroups: comp.lang.perl
Subject: Re: Can this be done more slickly
Message-ID: <1991Mar16.033051.2482@convex.com>
Date: 16 Mar 91 03:30:51 GMT
References: <VICTOR.91Mar15171726@irt.watson.ibm.com>
Sender: news@convex.com (news access account)
Reply-To: tchrist@convex.COM (Tom Christiansen)
Distribution: comp
Organization: CONVEX Software Development, Richardson, TX
Lines: 43
Nntp-Posting-Host: pixel.convex.com

From the keyboard of victor@ibm.com:
:I wrote the function given below to transform a string with -'s in it
:so that all alphabetics were lower case, except at the beginning of
:the string or immediately after a dash, in which case it would be
:upper case.  The follow works correctly, but I was wondering (in the
:JAPH spirit), if there is a more clever (or concise) way of doing it.

:sub normalize {
:	local(@pieces) = split(/-/,$_[0]);
:	local($i);
:	foreach $i (0..$#pieces) {
:	    $pieces[$i] =~ tr/A-Z/a-z/;
:	    substr($pieces[$i],0,1) =~ tr/a-z/A-Z/;
:	}
:	join('-',@pieces);
:    }

Before 4.0, you have to use this:

    sub normalize { 
       local($_) = @_;
       tr/A-Z/a-z/;
       substr($_,$[,1) =~ tr/a-z/A-Z/;
       s/(-[a-z])/$x = tr#a-z#A-Z#, $x/ge;
       $_;
    }


But at 4.0beta or above, this works, and runs in just 40% the time of 
my previous example:

    sub normalize { 
       local($_) = @_;
       tr/A-Z/a-z/;
       substr($_,$[,1) =~ tr/a-z/A-Z/;
       s/-([a-z])/-\u$1/g;
       $_;
    }

The first two tr's could have been done with \L and \u escapes, but 
the tr's run a good faster that they do in those cases.

--tom