Re: [EXTERNAL] RE: File having an array using new free form

On 2018-06-15 10:40 AM, dlclark@xxxxxxxxxxxxxxxx wrote:

OK, I understand. But, from the perspective of the next
programmer coming along, it is "tricking" the compiler because it is not
clear/obvious what is going on -- because it is not usual to use a prefix
with a period in it. Because, usually, the resulting I-specs are not
actually qualified. Here is my compile listing to demonstrate what I am
saying.


4 dcl-f MyFile disk extfile(*extdesc) prefix('CHE.')
...
8=I I 1 8 0CHE.IMLFCMPTID
9=I I 9 16 0CHE.CICOMPPTID
10=I I 17 24 0CHE.IMLFSEQTID

Sorry, I don't understand what you mean by "the resulting I-specs are not qualified". To me, having CHE.IMLFCMPTID in the I spec says that the I spec _is_ qualified.

But I agree that it's tricky, and maybe not an ideal way to code any more, now that we can just code the data structure on the I/O opcode without relying on I specs at all. Although it's maybe more a matter of tricking the next RPG programmer than tricking the compiler.

The support for PREFIX('DS.') was added in V5R1, one release prior to the support for allowing a data structure in the result field for I/O to externally-described files. This somewhat goofy F-spec PREFIX support allowed the I/O opcodes to read into and write from qualified data structures.