On Thu, 2013-07-18 at 21:34 -0400, Tom Lane wrote:
> Yeah, this is randomly different from everywhere else in PG. The more
> usual convention if you want typedefs for both the struct and the
> pointer type is that the pointer type is FooBar and the struct type is
> FooBarData.
I think that is mor
On 07/18/2013 09:20 PM, Peter Eisentraut wrote:
The new jsonfuncs.c has some confusing typedef scheme. For example, it
has a bunch of definitions like this:
typedef struct getState
{
...
} getState, *GetState;
So GetState is a pointer to getState. I have never seen that kind of
conventi
Peter Eisentraut writes:
> The new jsonfuncs.c has some confusing typedef scheme. For example, it
> has a bunch of definitions like this:
> typedef struct getState
> {
> ...
> } getState, *GetState;
> So GetState is a pointer to getState. I have never seen that kind of
> convention before.
The new jsonfuncs.c has some confusing typedef scheme. For example, it
has a bunch of definitions like this:
typedef struct getState
{
...
} getState, *GetState;
So GetState is a pointer to getState. I have never seen that kind of
convention before.
This then leads to code like
GetState
