Subject: Re: C-API Header diversion - msg#00017

Previous Message by Date: click to view message preview

Re: C-API Header diversion

Hi, Markus Schaber wrote: > I've attached a patch that should fix all differences between header and > implementation in the C-API. > > It compiles fine, but is not tested further. Actually, my code compiled fine, but the patch was broken. Strk, thanks for your note on IRC. I had accidentally fixed the GEOSGeom typo in geos_c.h instead of geos_c.h.in. And as geos_c.h is not removed in "make clean", only in "make distclean", I did not notice. Apply the following patch to the .diff file I sent (or correct geos_c.h.in manually). --- ../capi_unify.diff 2006-10-26 15:47:55.638568500 +0200 +++ ../capi_unify2.diff 2006-10-27 11:19:25.214624250 +0200 @@ -780,7 +780,7 @@ typedef void (*GEOSMessageHandler)(const char *fmt, ...); + -+#ifndef GEOSGEOM ++#ifndef GEOSGeom typedef struct GEOSGeom_t *GEOSGeom; typedef struct GEOSCoordSeq_t *GEOSCoordSeq; +#endif The resulting code compiles fine, and seems to pass "make check" on i386 and AMD64. But I get loads of messages like: > Trying with original input. > Did not run OverlayResultValidator as the precision model is not > floating Btw, it seems that "make check" does not include any C-API tests... HTH, Markus -- Markus Schaber | Logical Tracking&Tracing International AG Dipl. Inf. | Software Development GIS Fight against software patents in Europe! www.ffii.org www.nosoftwarepatents.org signature.asc Description: OpenPGP digital signature _______________________________________________ geos-devel mailing list geos-devel@xxxxxxxxxxxxxxxxxxxx http://geos.refractions.net/mailman/listinfo/geos-devel

Next Message by Date: click to view message preview

Re: C-API Header diversion

Hi, Strk, strk@xxxxxxxxxxxxxxx wrote: >> Apply the following patch to the .diff file I sent (or correct >> geos_c.h.in manually). > > Ok, can you send a full ChangeLog entry for this please ? -- snip -- Changed the C-API implementation to actually include the header that's included by the clients, this allows the compiler to check declarations and implementation for consistency. On this trip, some small differences (const, sign, bool) were fixed, and some const correctnes propagated into core. -- snap -- >> Btw, it seems that "make check" does not include any C-API tests... > > Right... we should really think about adding some. > I dunno if TUT will be appropriated for this though. Mateusz ? There's a geostest binary built in the capi directory, but I don't know whether that does any meaningful tests. HTH, Markus -- Markus Schaber | Logical Tracking&Tracing International AG Dipl. Inf. | Software Development GIS Fight against software patents in Europe! www.ffii.org www.nosoftwarepatents.org signature.asc Description: OpenPGP digital signature _______________________________________________ geos-devel mailing list geos-devel@xxxxxxxxxxxxxxxxxxxx http://geos.refractions.net/mailman/listinfo/geos-devel

Previous Message by Thread: click to view message preview

Re: C-API Header diversion

Hi, Markus Schaber wrote: > I've attached a patch that should fix all differences between header and > implementation in the C-API. > > It compiles fine, but is not tested further. Actually, my code compiled fine, but the patch was broken. Strk, thanks for your note on IRC. I had accidentally fixed the GEOSGeom typo in geos_c.h instead of geos_c.h.in. And as geos_c.h is not removed in "make clean", only in "make distclean", I did not notice. Apply the following patch to the .diff file I sent (or correct geos_c.h.in manually). --- ../capi_unify.diff 2006-10-26 15:47:55.638568500 +0200 +++ ../capi_unify2.diff 2006-10-27 11:19:25.214624250 +0200 @@ -780,7 +780,7 @@ typedef void (*GEOSMessageHandler)(const char *fmt, ...); + -+#ifndef GEOSGEOM ++#ifndef GEOSGeom typedef struct GEOSGeom_t *GEOSGeom; typedef struct GEOSCoordSeq_t *GEOSCoordSeq; +#endif The resulting code compiles fine, and seems to pass "make check" on i386 and AMD64. But I get loads of messages like: > Trying with original input. > Did not run OverlayResultValidator as the precision model is not > floating Btw, it seems that "make check" does not include any C-API tests... HTH, Markus -- Markus Schaber | Logical Tracking&Tracing International AG Dipl. Inf. | Software Development GIS Fight against software patents in Europe! www.ffii.org www.nosoftwarepatents.org signature.asc Description: OpenPGP digital signature _______________________________________________ geos-devel mailing list geos-devel@xxxxxxxxxxxxxxxxxxxx http://geos.refractions.net/mailman/listinfo/geos-devel

Next Message by Thread: click to view message preview

Re: C-API Header diversion

Hi, Strk, strk@xxxxxxxxxxxxxxx wrote: >> Apply the following patch to the .diff file I sent (or correct >> geos_c.h.in manually). > > Ok, can you send a full ChangeLog entry for this please ? -- snip -- Changed the C-API implementation to actually include the header that's included by the clients, this allows the compiler to check declarations and implementation for consistency. On this trip, some small differences (const, sign, bool) were fixed, and some const correctnes propagated into core. -- snap -- >> Btw, it seems that "make check" does not include any C-API tests... > > Right... we should really think about adding some. > I dunno if TUT will be appropriated for this though. Mateusz ? There's a geostest binary built in the capi directory, but I don't know whether that does any meaningful tests. HTH, Markus -- Markus Schaber | Logical Tracking&Tracing International AG Dipl. Inf. | Software Development GIS Fight against software patents in Europe! www.ffii.org www.nosoftwarepatents.org signature.asc Description: OpenPGP digital signature _______________________________________________ geos-devel mailing list geos-devel@xxxxxxxxxxxxxxxxxxxx http://geos.refractions.net/mailman/listinfo/geos-devel