Re: Reading a huge input line in parts (Handles multi-digit numbers!)

[BrowserUk]

How about reading a block at a time and spliting that?:

sub genBufferedGetNum {
    my @buf = do{ local $/ = \4096;  split ' ', scalar <>; }
    my $leftover = pop @buf;
    return sub {
        unless( @buf ) {
            unless( eof  ) {
                @buf = do{ local $/ = \4096; split ' ', $leftover . <>
+ };
                $leftover = pop @buf;
            }
            else {
                die 'premature eof' if $leftover != 0;
                return $leftover; # last number
            }
        }
        return shift @buf;
    };    
}

my $getNum = genBufferedGetNum();

while( my $num = getNum->() ) {
    ## do stuff
}

[GotToBTru]

Could this not be simplified?

sub genBufferedGetNum {
    return sub {
                @buf = do{ local $/ = \10; split ' ', <> };
                return @buf;
    };
}

my $getNum = genBufferedGetNum();

while( my @part  = $getNum->() ) {
  print @part, "\n";
}

I shortened the buffer size for testing purposes

$: cat tb.dat
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0
$:

$: cat tb.dat | perl tb.pl
01234
56789
01234
56789
01234
56789
0

[BrowserUk]

Could this not be simplified?

And what happens when your buffer size splits a multi-digit number in two?

Ie. Run your code against this input:

123 456 789 1

And it produces: 123 456 78 9 1

And doesn't notice that the last number is supposed to be 0.