Hi Rafi - thank you for that description. I'll have to dig into that code a bit more to get a feel for it.
The only concern I have is the implementation of PN_HANDLE. If I'm correct, you can't directly share PN_HANDLE's across compilation units due to the use of static variables. In other words, if I have PN_HANDLE(RAFI); in my rafi_private.h header, each compilation unit will define a RAFI handle with a different underlying value. Since it's not a compilation-time error, this could result in a very hard to diagnose run time bug. Perhaps a simple #defined constant would be safer? ----- Original Message ----- > From: "Rafael Schloming" <r...@alum.mit.edu> > To: users@qpid.apache.org > Sent: Monday, March 9, 2015 3:10:47 PM > Subject: Re: proton application context API deprecated in 0.9? > > On Mon, Mar 9, 2015 at 10:35 AM, Ken Giusti <kgiu...@redhat.com> wrote: > > > Hi, > > > > I just noticed that the 0.9rc1 marks the > > pn_XXX_get_context()/pn_XXX_set_context() set of APIs as being deprecated. > > > > I use these apis fairly frequently as a means to map back to my > > application's context. > > > > What are they being replaced with? I couldn't find an associated JIRA > > explaining how to upgrade my code. > > > > thanks, > > > > The context API has been replaced with the attachments API. The attachments > API provides two key improvements over contexts. Firstly, you can have > multiple attachments for a given object without them interfering with each > other. Secondly, it permits you to define the type of each attachment, > thereby allowing simpler options for memory management since you can use > reference counted pointers if you wish. > > The general pattern is: > > PN_HANDLE(FOO) // this is a macro that defines FOO to refer to a static > memory location guaranteed to be unique within the program > > pn_record_t *record = pn_xxx_attachments(xxx); > pn_record_def(record, FOO, PN_VOID | PN_OBJECT); > pn_record_set(record, FOO, value); > void *value = pn_record_get(record, FOO); > > For real use you'll probably want to wrap the get/set with accessors that > cast foo to/from a more specific type, e.g.: > > pn_foo_t *pn_xxx_get_foo(xxx); > void pn_xxx_set_foo(xxx, pn_foo_t *foo); > > I just noticed the pn_record_t API is missing doxygen. I'll add it to the > list for a doc blitz. I have a plane ride coming up soon, so hopefully will > have some time to churn out the missing doc pieces here. > > --Rafael > -- -K --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@qpid.apache.org For additional commands, e-mail: users-h...@qpid.apache.org
