Re: [PATCH] Use get_key_arg to parse keyword arguments in json_parse_object_type

[Paul Eggert]

John Wiegley wrote:
> I approve of the patch, Vibhav. I'm suprised we don't have a SAFE_CALLOC
> macro, though, to avoid having to use memset immediately after.

I guess Vibhav's future patches will make this issue moot, as the change doesn't 
make sense by itself; it merely substitutes a less-efficient implementation for 
a more-efficient one. Perhaps Vibhav's future changes will pass the argument 
without memsetting it, for example.

It would be helpful to know what the future changes are, as it's hard to tell 
whether the proposed patch is a good idea without seeing it in the context of 
the future patches.