Re: macro substitution

[jbk@aps.anl.gov (Jim B. Kowalkowski)]

More about macro substitution:

Nicks writes:

>FROM npr@jach.hawaii.edu Mon Feb 12 14:02 CST 1996
>
>1. I don't think the current syntax precludes the ommission of a
>   reserved word such as `set' or `define' - just an embedded equals sign
>   may be sufficient. Then the amount you type must be about the same.
>
>2. I believe the current dbLoadTemplate is really just a host side tool
>   since it writes to stdout. Am I correct? Thus you don't really have to
>   worry about the number of file opens if you are just wanting to provide
>   the same functionality. Basically, I thought the conclusion was that if
>   you have many files and use ftp you would always use dbAsciiExpand on
>   the host for efficiency reasons.

I disagree, dbLoadTemplate() is useful running on the IOC.

>
>3. dbLoadAscii could always buffer its file handles... (No, I don't think
>   I like this either - it isn't really realistic if you are using ftp on
>   VxWorks).

This seems very bad.

>
>4. I agree to supporting both formats. However, I think the syntax that is
>   chosen should be consistent, and the semantics such that:
>
>file gizmo_database.db
>{
> i1{ pre=part1,parm=big,   alarm_limit=20 }
> i2{ pre=part2,parm=medium,alarm_limit=30 }
>}
>
>is identical to (changing my syntax to match yours):
>
>pre=part1
>parm=big
>alarm_limit=20
>include "gizmo_database.db"
>
>pre=part2
>parm=medium
>alarm_limit=30
>include "gizmo_database.db"
>
>i.e. the `file' statement is sort of an implied for loop.
>
> Now the user needs to maintain 300 files.  Tell me if there is something
> I am misunderstanding.
>
>No, I think you had it absolutely right.
>

I do not like this at all.  Remember, we have many users already set up
to use the other format.  I think it is unreasonable to ask them to
change to a format that increases the maintainence of their files and
possibly reduces readability of substitution variables.

Jim