Front page | perl.perl5.porters |
Postings from November 2021
On Wed, Nov 10, 2021 at 3:46 PM Ovid via perl5-porters <
perl5-porters@perl.org> wrote:
> > On Wednesday, 10 November 2021, 20:41:57 CET, Oodler 577 via
> perl5-porters <perl5-porters@perl.org> wrote:
>
>
>
> > 'state' can be easily replicated with lexical
> > closures:
> >
> > {
> > my $foo;
> > sub bar {
> > ++$foo;
> > #...do stuff
> > return $foo;
> > }
> > }
> >
> > What's the form for 'permanent'?
>
> Ironically, depending on the surrounding code, the example you give would
> be for "permanent" variables, not "state" variables. What we currently use
> as "state" variables get reset in a dynamic scope, as my original code
> shows. Your example seems to match the "state" documentation, but doesn't
> match the state behavioe.
>
> "state" could be replaced with lexical closures when done correctly. When
> done correctly, "permanent" could, also. "When done correctly" is the
> problem.
>
> The docs for "state" *seem* to imply the semantics I would like to see for
> "permanent", but the behavior is different. Static lexical scope and
> dynamic lexical scope are different beasts.
>
> There are a couple of times I've been bitten by thinking "state" would
> imply the semantics of what I describe as "permanent". This is usually when
> I am building code dynamically and I hate building scaffolding to remember
> that.
>
> So ...
>
>
> - "my" reinitializes every lexical scope
> - "state" reinitializes every dynamic lexical scope
> - "permanent" never reinitializes (which is what the docs for "state"
> imply to me)
>
>
> I suspect (but can't prove) that "permanent" probably matches what most
> developers think of as "state", and those developers who can appreciate
> "dynamic lexical scope" would hopefully understand the difference.
>
> #!/usr/bin/env perl
>
> sub foo { # static lexical scope
> ...
> return sub { # dynamic lexical scope
> ...
> }
> }
>
Thanks for explaining, this is new terminology for me - unrelated to the
concept of dynamic scope used in the documentation (runtime scoping).
Perhaps an addition to the documentation could be made, but this seems like
expected behavior to me.
-Dan
Thread Previous
|
Thread Next