Re: do separate hashes share common keys?

in reply to do separate hashes share common keys?

In terms of documentation, the perldelta for 5.20.0 notes that "for example hash keys were already copy-on-write" when it describes the COW implementation.

https://metacpan.org/pod/distribution/perl/pod/perl5200delta.pod#Internal-Changes

I'm not sure when it was introduced, but jcb's estimate in 11128352 looks reasonable to me.

(Edit: look for the ninth bullet point under the link to find the quoted text).

Comment on Re: do separate hashes share common keys?

Replies are listed 'Best First'.
Re^2: do separate hashes share common keys? by LanX (Saint) on Feb 14, 2021 at 13:05 UTC
thanks, but I'm confused about the "for example hash keys were already copy-on-write" comment. By principle hash-keys are (and must be) immutable. Immutable data can be shared w/o allocating new memory, while mutable objects need to be copied to separate space. COW is a performance trick to "lazily" delay that copy to the time when the mutation (the write) happens. So there must be a misunderstanding, since hash-keys are always read-only, there is never a write. Cheers Rolf _{(addicted to the Perl Programming Language :) Wikisyntax for the Monastery}	[reply]

Replies are listed 'Best First'.

Re^2: do separate hashes share common keys?
by LanX (Saint) on Feb 14, 2021 at 13:05 UTC

"for example hash keys were already copy-on-write"

By principle hash-keys are (and must be) immutable.

Immutable data can be shared w/o allocating new memory, while mutable objects need to be copied to separate space.

COW is a performance trick to "lazily" delay that copy to the time when the mutation (the write) happens.

So there must be a misunderstanding, since hash-keys are always read-only, there is never a write.

Cheers Rolf
_{(addicted to the Perl Programming Language :)

Wikisyntax for the Monastery}

[reply]

In Section Seekers of Perl Wisdom