SQL/MX 3.2 Programming Manual for C and COBOL (H06.25+, J06.14+)

C/C++ Program Compilation
HP NonStop SQL/MX Release 3.2 Programming Manual for C and COBOL—663854-002
15-21
OSS-Hosted SQL/MX C/C++ Preprocessor
name of the SQL/MX C/C++ source file (for example, sqlprog.sql) without the
file extension.
-e
generates CHARACTER data types for date-time data types. This behavior is
compatible with NonStop SQL/MX Release 1.8. For more information, see INVOKE
and Date-Time and Interval Host Variables (SQL/MX Release 1.8 Applications) on
page 3-40.
-n
directs the preprocessor to automatically append a null terminator to all host
variable character strings before they are fetched into. Using the -n option does
have the potential to produce nonportable code. Moreover, if the -a option is used
together with the -n option, the -n option has no effect on VARCHARs.
-a
specifies that VARCHARs are to be translated into structures that contain a length
and character string. For details about using this option, see Generating Structures
Instead of Using Null-Terminated Strings on page 3-18 and Example: Using a
Structure on page 3-19.
This preprocessor option overrides the SQL/MX default VARCHAR, which is a
string with a null terminator. Moreover, if the -a option is used together with the -n
option, the -n option has no effect on VARCHARs.
-l list-file
is the name of the output list file that contains preprocessor error and warning
messages. The default is source-file.lst, where source-file is the name
of the SQL/MX C/C++ source file (for example, sqlprog.sql) without the file
extension.
-p
turns off the automatic generation of #line directives in the C/C++ output file,
disables source-level debugging, and shows the generated C/C++ code for
debugging purposes.
-o
overrides the use of Tandem floating point and uses IEEE floating point instead for
host variables. In addition, if used with invoked SQL/MP tables with a column of
type REAL, this option causes the invoked structure to be of type DOUBLE. For
more information, see INVOKE and Floating-Point Host Variables
on page 3-41.
By default, the c89 compiler (TNS/E targeted compilation) defaults to IEEE_float
and will invoke the -o option when it calls the preprocessor. If you want your
program to use Tandem_float, use the c89 -Wtandem_float option to compile
the mxsqlc-generated annotated source file.