On Fri, Aug 9, 2019 at 10:33 AM Ebrahim Byagowi <[email protected]> wrote:
> Was looking inside libraqm code and saw the use of stdbool.h there, made
> we wonder why HarfBuzz went for hb_bool_t and whether we can go for it also
> or not.
>
Because back in 2006 even stdint.h was not portable-enough. You still see
the following block in hb-common.h:
#if defined (_SVR4) || defined (SVR4) || defined (__OpenBSD__) || \
defined (_sgi) || defined (__sun) || defined (sun) || \
defined (__digital__) || defined (__HP_cc)
# include <inttypes.h>
#elif defined (_AIX)
# include <sys/inttypes.h>
#elif defined (_MSC_VER) && _MSC_VER < 1600
/* VS 2010 (_MSC_VER 1600) has stdint.h */
typedef __int8 int8_t;
typedef unsigned __int8 uint8_t;
typedef __int16 int16_t;
typedef unsigned __int16 uint16_t;
typedef __int32 int32_t;
typedef unsigned __int32 uint32_t;
typedef __int64 int64_t;
typedef unsigned __int64 uint64_t;
#else
# include <stdint.h>
#endif
> If a migration isn't possible, most likely I'd guess, maybe having some
> compiler definition for using actual bool instead 'int', at least for
> clients that are embedding harfbuzz can trig some optimizations for them.
>
I don't think it makes any optimizations possible per se. I'd just forget
it...
We should review if we take any hb_bool_t* arguments in any API. If not,
it's probably safe ABI-wise to change.
>
> Thanks!
> _______________________________________________
> HarfBuzz mailing list
> [email protected]
> https://lists.freedesktop.org/mailman/listinfo/harfbuzz
--
behdad
http://behdad.org/
_______________________________________________
HarfBuzz mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/harfbuzz
