Re: Embedded SQL compiler at Linux

On 16 May 2002 15:32:48 GMT, "Patrick L. Nolan" <pln_at_razzle.Stanford.EDU> wrote:

>In comp.databases.oracle.misc Ole Streicher <ole-usenet-04_at_gmx.net> wrote:
>
>> The "real" big problem is now that is doesn't work with
>> mathcalls.h. One file gives the error
>
>> Syntax error at line 54, column 1, file /usr/include/bits/mathcalls.h:
>> Error at line 54, column 1 in file /usr/include/bits/mathcalls.h
>> __MATHCALL (acos,, (_Mdouble_ __x));
>> 1
>> PCC-S-02201, Encountered the symbol "f" when expecting one of the following:
>> ; , = ( [
>
>> this error is repeated several times (on different lines of
>> mathcalls.h), and then the compilation process crashes with an
>> "segmentation fault".
>
>I experienced this a couple of days ago with Oracle 9i. When I
>used the proc option parse=none, all this stuff went away. I wish
>I knew what this means.

Its been a while since I worked with Embedded SQL (Pro*C). The problem is that the preprocessor is trying to parse your SQL against a database. Setting parse=none tells the preprocessor not to try to verify your SQL. The MATHCALL errors are strange, but the PCC-S errors are normal if you haven't set you compile environment up correctly. Received on Fri May 17 2002 - 05:41:02 CEST