http://www.perlmonks.org?node_id=1025616


in reply to Curious find while comparing grep, map, and smart match...

        redo if ( @gArray && grep { /$rInt/ } @gArray );
...
        redo if ( @mArray && map { /$rInt/ } @mArray );
...
        redo if ( @sArray && $rInt ~~ @sArray );

[download]

The three are not equivalent.    grep and map use regular expressions and have code blocks while the other does not.    Try it like this: 

        redo if @gArray && grep /$rInt/, @gArray;
...
        redo if @mArray && map /$rInt/, @mArray;
...
        redo if @sArray && /$rInt/ ~~ @sArray;

[download]

Replies are listed 'Best First'.
Re^2: Curious find while comparing grep, map, and smart match...
by dbuckhal (Chaplain) on Mar 27, 2013 at 03:17 UTC

    Wow, that, and using word boundaries from tye's hint, made a significant difference!

    before:

    Benchmark: timing 2500 iterations of grepGen, mapGen, smartGen...
   grepGen: 23 wallclock secs (22.51 usr +  0.00 sys = 22.51 CPU) @ 11
+1.06/s (n=2500)
    mapGen: 11 wallclock secs (11.42 usr +  0.00 sys = 11.42 CPU) @ 21
+8.93/s (n=2500)
  smartGen:  1 wallclock secs ( 1.05 usr +  0.00 sys =  1.05 CPU) @ 23
+92.34/s (n=2500)

    [download]

    after:

    Benchmark: timing 2500 iterations of grepGen, mapGen, smartGen...
   grepGen: 14 wallclock secs (13.70 usr +  0.00 sys = 13.70 CPU) @ 18
+2.52/s (n=2500)
    mapGen: 13 wallclock secs (13.57 usr +  0.00 sys = 13.57 CPU) @ 18
+4.20/s (n=2500)
  smartGen:  2 wallclock secs ( 1.19 usr +  0.00 sys =  1.19 CPU) @ 21
+07.93/s (n=2500)

    [download]

    That really narrowed the gap between grep and map.

[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom