nrdvana

#! /usr/bin/perl
use v5.10;
use strict;
use warnings;
use Pod::Usage;
use Getopt::Long;
use POSIX 'dup2';

our $VERSION= 0.001;

nrdvana

diff --git a/toke.c b/toke.c
index e6ff0c4f74..3c32e70dfd 100644
--- a/toke.c
+++ b/toke.c
@@ -9174,6 +9174,13 @@ yyl_try(pTHX_ char *s)
                 return tok;
             goto retry_bufptr;
         }
+        if (UTF && s + 2 < PL_bufend && *s == '\xE2' && s[1] == '\x88' && s[2] == '\x9E') {
+            pl_yylval.opval = newSVOP(OP_CONST, 0, newSVnv(NV_INF));

nrdvana

#! /usr/bin/env perl
use strict;
use warnings;
use Benchmark qw( :hireswallclock cmpthese );

{
        package OP_Point;
        use v5.26;
        use Object::Pad;
        class OP_Point {

nrdvana

package Mock::RelationalData;
use Moo 2;

=head1 SYNOPSIS

  my $reldata= Mock::RelationalData->new;
  
  # Define custom mock data generators
  $reldata->add_generator("words" => ...);
  

nrdvana

My Nitpicks with the Design of Moo

Verbosity

The attribute declaraions can get a little bit verbose. Moo(se) isn't bad, but it seems like we ought to be able to do a little better...

has foo => ( is => 'rw', isa => Type, lazy_build => 1, trigger => 1 );