NetNews Usenet Archive 1992 #30/NN_1992_30.iso/spool/comp/lang/perl/7520

Text File | 1992-12-21 | 1.8 KB | 56 lines

Newsgroups: comp.lang.perl
Path: sparky!uunet!mcsun!sun4nl!ruuinf!henkp
From: [e-mail] (Henk Penning)
Subject: Re: In 5.0, are we gonna get...
Sender: [e-mail]
Message-ID: <[e-mail]>
Date: Fri, 18 Dec 1992 14:01:12 GMT
References: <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Organization: Utrecht University, Dept. of Computer Science
Lines: 44

In <[e-mail]> [e-mail] (Gustaf Neumann) writes:

> [ ... ] how about
>
>        $sum = eval join('+ ',1,34,$x,45,-8,0.1);
>
>ok, the string conversion could be avoided, but you can do it now...

>Gustaf Neumann          [e-mail], [e-mail]

  I would like to write:

    $max = reduce { $a >  $b ? $a : $b } @row

-- Another idea would be to have a reduce in array-context,
   for example to squeeze out adjacent equal elements:

     @uniq = reduce { ($a eq $b) ? ($a) : ($a,$b) } @row

   Here the ab-sub is evaluated in array-context. The first time
   $a is the first element of @row, later on the last element
   of the partial result.

-- Computing the difference of adjacent elements diff[i]=row[i+1]-row[i]

     @diff = reduce { ($b-$a,$b) } @row ; pop @diff ;

-- Range compression: 1,3,4,5,7,22 -> 1,3..5,7,22

   sub compress
     { if ( $a =~ /^(\d+)\.\.(\d+)$/ )
     { ( $b == $2 + 1 ) ? ("$1..$b") : ($a,$b) ; }
       else
     { ( $b == $a + 1 ) ? ("$a..$b") : ($a,$b) ; }
     }

   @compress = reduce compress @row ;

-- The examples have not been tested (;-) but the code looks nice.
   The scalar-reduce is much more important to have than the
   array-reduce, but I think there is a case for the latter too.

                ===  HenkP  ===