Re: EPICS database macro warning

[Wang Xiaoqiang <xiaoqiang.wang@psi.ch>]

Hi,

I actually use this 'bug' to selectively mask out certain parts of the 
database.

In the following example, if HS='#', the sequence selects "jog to 
limit". Otherwise, it "jog to home switch". Without this trick, I would 
need to create another record for this selection.

record(sseq, "INITSEQ")
{
...

    field (DOL2, "1")
    field (WAIT2,"Wait")
    field (LNK2, "TOLIMIT.PROC CA") # Jog to limit
$(HS)    field (LNK2, "TOHS.PROC CA") # Jog to home switch
....
}

If the 'bug' is useful, it probably becomes a feature.

Best
Xiaoqiang


On 01/31/2018 10:28 AM, Amien Crombie wrote:
> Hi
>
> I inadvertently stumbled upon this strange 'bug' in my ioc.db file
>
> I had the following record:
>
> record(subArray,    "$(uid):attenuator")
>      {
>              field(INP,        "$(uid):raw")
>              field(NELM,        "9")
>              field(MALM,        "9")
>              field(INDX,        "9")
>              field(FTVL,        "LONG")
>           #   field(FLNK,        "$(dd):getX")
>
>       }
>
> Now the field '*FLNK*' is commented out and I have deleted my macro 
> $(dd) in my substitution file.
>
> When I run my IOC, I get the following warning
>
> /macLib: macro dd is undefined (expanding string #    field(FLNK, "$(dd)")/
>
> /Warning: 'db/dbExample1.db' line 16 has undefined macros/
>
>
> So the comment (#) still evaluates the macro, even if the documentation 
> says it ignores the whole line. Of course if the macro exist it runs 
> without the warning.
>
> Not a critical issue but can be misleading.
>
> Thanks
>
> Amien