Re: Time Constant Errors

I think I may have discovered one reason for  the "over excitement" of LPEX.

I tried out the sample of code (below) in a new source member. My usual practice is to hit ENTER a few times to give me some blank lines (head room) before coding starts. When I pasted in the sample, all sorts of errors were highlighted and it appeared that LPEX thought I was coding RPGLE source with specification letters in column 6. In other words, the **FREE directive was being ignored.

Since the **FREE was not the first line of code, I deleted the leading blank lines, ("head room"). The LPEX errors remained in the current edit session. However, if I saved the source and re-opened - no errors.

If I pasted the sample with _no_ leading blank lines, (i.e. **FREE as the very first line of code) - no errors.

Kinda makes sense, (**FREE must be the first line). However, LPEX should be more amenable if leading blank lines are removed within the edit session.

Curious.
Brian.

On 26/01/2018 16:41, Mark Murphy wrote:

RDi - 9.6.0.1
IBM i - 7.1

I have the following code

**free
ctl-opt Option(*SrcStmt : *NoDebugIo: *NoUnref)
DatFmt(*ISO) TimFmt(*ISO)
DftActGrp(*No) ActGrp(*New)
BndDir('JCS')
Main(JCUPDFLCYR);

...

dcl-s RunTime Time Inz(t'02.03.00');

...

The time literal causes the following errors in the LPEX editor:

Multiple markers at this line
- RNF3429E Initial value is not valid; parameter is ignored.
- RNF0306E The Time literal is not valid.
- RNF0607E Text is skipped to recover from the previous error.

It does compile properly though.

If I change my default time format to *HMS, and the time literal to
t'02:03:00', then RDi is happy (it also compiles properly). But I prefer to
keep my dates and times in *ISO format internally.