Re: Named constants in DCL-S

On 2019-04-23 12:06 PM, Paul Bailey wrote:

Thanks Barbara.

The field CURR_1 in my example is defined by a file. (There is a DCL-F
earlier in the source). The RDi outline finds the definition fine, but the
resulting spooled file from the compile has the correct definition for
CURR_1 in an I-spec further down the source. Are DCL-F definitions done
after the DCL-C and/or DCL-S definitions? If so, is the only way I can
declare a zoned version of a packed file field to hard-code the field
definition?

The fields associated with files are defined by the generated I specs, so that would be too late for your definitions.

You could define an unnamed (or named) externally-described data structure to get the fields available earlier:

dcl-ds *n extname('MYFILE') end-ds;