On Wed, Oct 31, 2012 at 3:07 AM, Jeremy Nicoll - ml sourceforge <email@example.com>
The problem would come, I think, if one was writing for example a utility,
Mark Miesfeld <firstname.lastname@example.org
> Presumably, if you're writing the program you won't use variables with the
> same name as a column name. I just would never use stems in this
to handle any table, regardless of its column names.
In this use case, you should simply not use the stem format for the return. You should use the .directory format where there will be no conflicts, no need to worry about a collision with a variable name.
Stems are great, if there is nothing else available. But, in ooRexx there is something else available. If you are worried about stem tail names conflicting with variable names, don't use a stem. Use a .directory object.
I suppose one could write the code with varnames which have been given a
hopefully unlikely-to-occur name prefix, and/or include code that tests that
there's no conflict between unknown column names and those vars.
It doesn't make sense to me, to spend 30 seconds of time trying to do that, when you can just use a .directory object and be done with it. :-)