Paul Burba <paulb@softlanding.com> writes:
>> > #ifdef AS400
>> > #define SVN_UTF_UTOE_XLATE_HANDLE "svn-utf-utoe-xlate-handle"
>> > +#define SVN_UTF_ETOU_XLATE_HANDLE "svn-utf-etou-xlate-handle"
>>
>> We already have one of those in libsvn_subr/cmdline.c
>
> Yes, but how does that help me in io.c? Unless you mean these should be
> made public...should svn_utf.h have all of these?
>
> #define SVN_UTF_NTOU_XLATE_HANDLE "svn-utf-ntou-xlate-handle"
> #define SVN_UTF_UTON_XLATE_HANDLE "svn-utf-uton-xlate-handle"
No, those are only used in one file.
> #ifdef AS400
> #define SVN_UTF_UTOE_XLATE_HANDLE "svn-utf-utoe-xlate-handle"
> #define SVN_UTF_ETOU_XLATE_HANDLE "svn-utf-etou-xlate-handle"
> #endif
I'd suggest putting SVN_UTF_ETOU_XLATE_HANDLE into a libsvn_subr
header, it doesn't need to be visible outside libsvn_subr.
>> I feel a bit guilty about suggesting that since I seem to be asking
>> for all your patches to be re-written :-/
>
> Heh, I'm not worried about that, I just want it done correctly. I see 3
> possible options to solve this:
>
> 1. Tweak the existing patch to avoid the deadlock issue and call it a day.
>
> 2. Figure out how to make apr_file_t and spawn() work together.
>
> 3. Move the patch entirely within hooks.c. Just add a second definition
> for run_hook_cmd() along with the existing implementation within a #ifndef
> AS400...#else...#endif block.
>
> FWIW I like option 3, it is certainly the smallest patch and introduces no
> new APIs.
3 is fine as far as I am concerned.
--
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Tue Mar 14 20:32:04 2006