Dan Sugalski <[EMAIL PROTECTED]> wrote:
> foreach loopvar (1..10000) {
> total[loopvar] = somedata[loopvar] + otherdata[loopvar];
> }
> If either (or both) of somedata or otherdata isn't an array of PMCs,
> there's going to be an awful lot of temporary PMC creation, which is a
> waste.
You would need one temporary PMC per operand, which can be assigned a
value and it can be created outside of the loop. To achieve the value
semantics that above addition has, you have to create new destination
PMCs anyway (presuming that C<total> holds PMCs).
> ... We're trying to encourage the use of more compact aggregate
> representations with Parrot, as currently Perl/Python/Ruby all have really
> portly aggregates and part of the design goals is to allow for compact
> versions in places we can be compact.
I don't see the point of "compact" with ops-files 60 times of the
current size (with all ki?c? permutations) and huge classes files
implementing all the keyed variants.
When only add_p_k_p_k_p_k is implemented, we have additional checks for
NULL keys, additional calls to key_integer() to extract the array
index, and additonal code to set the value of the key(s).
Please read (again) my proposal in:
Subject: Re: keyed vtables
Date: Mon, 18 Aug 2003 09:36:40 +0200
> Dan
leo
