>> the problem with all of these is that it's breaking away from the 1:1
>> mapping of the original apr_brigade_(p)flatten APIs.  if we're going
>> to do
>> that to make them more perlish, I'd much rather see
>>
>>   my $data = $bb->flatten(30);  # calls apr_brigade_flatten
>>   my $data = $bb->flatten($p);  # calls apr_brigade_pflatten
>>
>> and throw away the length component, since it's very rarely important
>> anyway, has the possibility of being wrong (at least that's my reading of
>> the API), and can be computed (more accurately) if required.
> 
> 
> Sounds good to me. It makes the API so much simpler to use.
> 
> I just thought that since it's similar to read() it could behave like one.
> 
> I find it strange that you need the pool to allocate the unknown len,
> but you don't need one to allocate a certain amount of bytes. To make
> the API consistent I'd suggest it to be:
> 
>   my $data = $bb->flatten($p, [$len]);
> 
> i.e. always pass $p, and use apr_pcalloc to allocate $len bytes if that
> optional argument is passed.

done.  take a look when you get the chance just to make sure I did the
allocation properly.

--Geoff


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to