[postgis-devel] sql.in.c do we like not like?

Sandro Santilli strk at keybit.net
Fri Mar 22 03:37:02 PDT 2013

On Fri, Mar 22, 2013 at 02:03:39AM -0400, Paragon Corporation wrote:

> To be fair to strk he did bring it up:
> http://lists.osgeo.org/pipermail/postgis-devel/2012-October/022217.html  
> and Mark -- you even responded :)
> http://lists.osgeo.org/pipermail/postgis-devel/2012-October/022223.html
> And I responded
> http://lists.osgeo.org/pipermail/postgis-devel/2012-October/022229.html

Great. Have you realized I made my part of trying to finding something
which is not a _C_ preprocessor ? It's called an "SQL Preprocessor" now.

We actually have multiple choices:

 - GPP, the Generic Preprocessor, packaged in apt-based system, already
        looked for by our ./configure
 - CPP, the C preprocessor, which mostly works fine when standalone,
        already looked up by our ./configure
 - CC,  the C compiler, in preprocessing mode, the worst choice but
        should work with -x c
 - SQLPP, a perl script written by someone somewhere (not packaged and
          not yet looked up by configure).

> But strk -- I think we may need to support MS VC in future so I don't want
> to make the future support more difficult than it needs to be.  I haven't
> tested on our Windows 2012 box yet, but that platform has a lot of changes
> that I fear may behave very differently from the other windows versions and
> may require us to use the MS toolchain.

Could you please just try manually modifying the SQLPP line in postgis/Makefile
to see if "-x c" fixes the issue ? I feel we're very close to the solution and
screaming for old behavior sounds like a purely fear-based way to deal with 
the issue.


More information about the postgis-devel mailing list