Re^2: Setting accessor with Object::Tiny

sub _set { my $self = shift; my ($key, $value); while (($key, $value) = splice @_, 0, 2) { $self->{$key} = $value +} }
[download]

For this, why not just:

c:\@Work\Perl\monks>perl -wMstrict -MData::Dump -le
"sub _set {
   my $self = shift;
   %$self = (%$self, @_);
   }
 ;;
 my $hr = bless { qw(y 25  z 26) };
 ;;
 $hr->_set(qw(a 1  b 2  c 3));
 dd $hr;
"
bless({ a => 1, b => 2, c => 3, "y" => 25, z => 26 }, "main")
[download]

Give a man a fish: <%-{-{-{-<

Comment on Re^2: Setting accessor with Object::Tiny Select or Download Code

Replies are listed 'Best First'.
Re^3: Setting accessor with Object::Tiny by jcb (Parson) on Aug 12, 2019 at 05:10 UTC
Because I had not thought of that? :-) It figures that there would be a more compact way to do that. Although if I had objects being mutated in an inner loop, I would want to compare both ways, in case changing a few keys is faster than setting the whole hash.	[reply]
Re^4: Setting accessor with Object::Tiny (updated) by AnomalousMonk (Archbishop) on Aug 12, 2019 at 05:31 UTC
`while (($key, $value) = splice @_, 0, 2) { $self->{$key} = $value }` Although I notice that the `while`-loop version allows assignment of an unpaired key/value (i.e., if the shift-ed `@_` array has an odd number of elements, the last element/key of the array has no value and is assigned as undef) without a warning. The list assignment version must be changed to something like `%$self = (%$self, @_, @_ & 1 ? undef : ());` to avoid a warning (if this is desireable), and this, I must admit, is a bit more messy. Update: But I guess a simple `no warnings 'misc';` `%$self = (%$self, @_);` would suppress that warning just as well. Give a man a fish: `<%-{-{-{-<`	[reply] [d/l] [select]
Re^5: Setting accessor with Object::Tiny (updated) by jcb (Parson) on Aug 12, 2019 at 06:23 UTC
In my code `_set` methods are always called with `key => value` pairs. Silently assigning `undef` is actually a bug, but (I said that code had not been tested) could be fixed by adding `die "unbalanced assignment group: ".join(', ', @_) if @_ & 1;` which is also a bit more messy. I personally believe that `no warnings` (along with `no strict`) should be syntax-highlighted in bright bold primary red and generally avoided. There are a few times I have had to write `no strict`, usually when doing some types of meta-programming, but it is very rare in my code. I am not sure if I have ever written `no warnings`.	[reply] [d/l] [select]