From: "Aaron Bannert" <[EMAIL PROTECTED]>
Sent: Wednesday, January 16, 2002 1:42 PM


> On Tue, Jan 15, 2002 at 05:39:59PM -0800, John Sterling wrote:
> > 
> > Long ago (maybe 3 years ago?) Manoj generated a patch for 1.3 which added
> > some widely used formats which were not implemented in the libc
> > implementation of strftime on win32.
> > 
> > Ideally we would continue to implement formats as necessary (or should we 
> > spend some
> > time implementing them all?).
> 
> John and I had some time this weekend to talk about this, and I definately
> think this is the right way to go. Unfortunately, I can't test this as
> I don't have a win32 box, but here is my +1 in concept.

+1 here as well, and since strftime() isn't even consistent on all Un*x 
platforms,
we would be best off doing this;

1. Define our set of supported strftime() escapes.

2. Create a bit of autoconf magic to format a string, repleat with all
   our supported escapes, and compare to our 'expected' result.

3. If so, HAVE_COMPLETE_STRFTIME and use the clib's implemention...

4. ... otherwise use the apr implementation.

I'll commit John's code, but want to see us start testing across the board for
our agreed-upon format tokens [however we define them].

Bill

Reply via email to