On 10 November 2012 14:54, John Gilmore <[email protected]> wrote:

> As you know, compared strings must themselves be of equal length if
> anomalies like 'aa' > 'x' are to be avoided.
>
> Case also poses problems.  Are 'john', 'jOHn', 'John' and the like to
> match?  Sometimes this is appropriate, even necessary, sometimes not.

I understand there's many interesting algorithms to encode a table of
strings. The point is that I started with a table of 128 entries that
is indexed to get the right string. The table was getting a bit big,
and I realize I could get it down to less than half by mapping
identical entries to the same string. Using 128 pointers to ~50
strings seemed to be an easy way using literals.

Was just surprised the literals did not do the trick. Funny enough an
EX against a table like this would do it..
  LA  R1,=CL48'John'
  LA  R1,=CL48''
  LA  R1,=CL48'Bill'

When I feel creative, I guess I'll make them variable length and
generate them with a macro. That could save me even more than weeding
the duplicates out.

Rob

Reply via email to