I'm not about to write my own version here, just offer some pointers I find
makes life easier for myself, and they may help others too if you're not
bound to "I don't want to use CPAN".
These days, I find it easier and safer to just use
> use Path::Tiny qw(path);
>
> my $infile = path('numbers.txt')->openr;
> my $outfile = path('sumfile.txt')->openw;
Those will open files and do the right things with error handling by
default.
If you decide you want a unfiltered IO stream ( ie: not subject to the CRLF
filter on win32 ), its easy to simply switch to
openr_raw and openw_raw, and it will do the right thing.
Likewise, if you decide you want utf8, just openr_utf8 and openw_utf8 (
there's a lot of complex messy bugs in utf8 io that this avoids )
Similarly, instead of:
> chomp( my @lines = <$infile> );
You can simply do:
> my (@lines) = path('numbers.txt')->lines({ chomp => 1 });
( And again, you can just bolt on the _raw and _utf8 suffixes and get the
right result )
If you want to avoid handling close yourself and you have a mass of data
you just want to write and have errors handled correctly:
> path('sumfile.txt')->spew(map { "$_\n"} @lines );
And again, you can just bolt on _utf8 and _raw suffixes and get the right
thing.
Spew also adds extra protection by writing to a temporary file , and then
moving it over the original one, so that if spew for some reason fails half
way through, the original file should not be affected.
--
Kent
*KENTNL* - https://metacpan.org/author/KENTNL
