On Tue, 25 Sep 2001, Philippe M . Chiasson wrote:
> Now, this patch should be very close to what it will end up being.
looks great!
> Worth noting now, I generalized the tie'ing mechanism and ended up with:
>
> SV *mpxs_typemap_output_tied()
> void *mpxs_typemap_input_tied()
very nice.
and use of SvCUR as iterator state is very clever :)
but i think FIRSTKEY is wrong, currently:
return ((apr_table_entry_t *)t->a.elts)[SvCUR(SvRV(tsv))++].key;
should be:
SvCUR(SvRV(tsv)) = 1; /* nextkey after we return firstkey (index 0) below */
return ((apr_table_entry_t *)t->a.elts)[0].key;
that's need to support:
for (1,2) {
while (my($key,$val) = each %$table) {
...
i'm in the process of applying/testing your patch, but source_scan is
indeed mucho brokeno. trying to sort that out.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
