OI::M::Finally({IO::All})
Get it?
(Meant to post this to perlmonks, but the site's submission form is missing the [Add] button right now).
Anyway, as a long time user of File::Slurp who utilizes both idioms below often:
# slurp into a scalar
$content = slurp("file.txt");
# slurp into an array
@lines = slurp("file.txt");
I'm willing to switch to IO::All's:
# slurp into a scalar
io("file.txt") > $content;But there's no array counterpart:
# slurp into an array
io("file.txt") > @lines;I've hopelessly tried several variants:
io("file.txt") > \@lines;
io("file.txt") >> @lines;
io("file.txt") >> \@lines;
@lines = @{ io("file.txt") };
@lines = <io("file.txt")>;
None of them works. Do I really have to settle with:
@lines = io("file.txt")->slurp;But where's the coolness of using IO::All in this case?
I'm loving the other features of IO::All though, like append, tie, copy, path, dir, etc. Yes, it's totally mad.
A programmer (mostly Perl 5 nowadays).
My CPAN ID: SHARYANTO. I'm sedusedan on perlmonks. My twitter is stevenharyanto (but I don't tweet much). Follow me on github: sharyanto.
Hmm, this should work
@lines = @{ io("file.txt") };
Since in IO::All::File, @{} is overloaded to deref it as Tie::File
Tested with IO::All 0.46
It doesn't look like operator overloading will yield array output, but instead of slurp, all will be a bit nicer. :)
Apropos:
Letting subroutines participate in pipelines, perhaps by overloading the '|' operator, would be really useful. I briefly played around with something like this several years ago, but didn't end up with anything I liked.
Its not really the same thing, but you might want to look at Zoidberg (why not Zoidberg?).
Good to see that Zoidberg's alive again. My zsh muscle memory was a bit too strong the last time I tried it, but it was a really nice merger of Perl and Unix the shell.
That said, I was thinking of something that worked within Perl's syntax, e.g.
io("file.txt") | sub { ... } | io("| frob > file2.txt");
But "|"'s precedence is a pain, there's the question of line-at-a-time vs. whole-output, etc.
This seems to suggest that it's theoretically doable:
perl -e'use overload ">" => sub { use Data::Dumper; print Dumper \@_ }; my @a; my $o = bless []; $o > \@a'The overload sub clearly gets passed an arrayref; the arrayref doesn't get numified first or anything annoying like that.
Someone just needs to make a few adjustments to the following IO::All methods to detect when the scalar is an arrayref and DWIM.
It was never very broken. There was a test failure that got introduced by some esoteric change between 5.8 and 5.10. I fixed that, one other rare bug and simplified the build process. Jaap gave me co-maint and there you go!
Hi gugod, you're right it does work, but: 1) it turns on chomping by default; 2) it needs rw access. This fails if you're not root, for example: @lines = @{ io("/etc/passwd") };