On Mon, 7 May 2007 01:24:32 -0700 Mike Mattie <[EMAIL PROTECTED]> wrote:
> Hello, > > I was looking at Parrot_get_runtime_prefix as an example of getting > environment variable values the parrot way. The function looked > over-complicated so I tried a refactor. [snip] > > I am inlining the patch for discussion. It is based against > rev-18443 with a rebase of #41908 applied. If This API > change is acceptable I will rebase it against HEAD clean. > > Segway: I mentioned this on #parrot, jisom created a patch that > returns char* instead of STRING*. > > It passed the test-suite functionally, I tweaked it a little to pass > codingstd so it is not 100% verified. the Docs test is broken , I > haven't played with the docs yet so they are still out of sync with > this prototype of the patch. > > --- HEAD/src/library.c 2007-05-06 17:58:47.000000000 -0700 > +++ rev-18443/src/library.c 2007-05-07 01:10:27.000000000 -0700 [snip] While discussing the patch recently I noticed that it does change behavior in a fairly visible way. The value set in config hash now over-rides PARROT_RUNTIME. Good or bad I am not sure, but it does not seem to have a impact on the test-suite. Something to ponder over. [snip] Cheers, Mike Mattie - [EMAIL PROTECTED]
signature.asc
Description: PGP signature
