Ben,

What's so ugly about them? In the old days, that was all we had. It's funny, 
back when we ran Universe, I-types were faster at doing the same job than 
S-types. Now, on Jbase (which also supports both), ther S-types are faster. Go 
figure...

Regards,
Charlie Noah

[EMAIL PROTECTED] writes:
Chuck wrote
> Specifically, I'm looking to pull a single Multi-Value
> ( like EXTRACT(@RECORD,5,5,0) ) using an A-type

This is only for Universe, not for Unidata,
because Unidata doesn't have the Pick-style "A" types.

Using an "A" (or "S") type of attribute definition
in Universe, you can extract a single value by doing
a translate back into the primary file.

It's ugly but it works.

Replace "a" and "v" with the real attribute number
and value number.

>ED DICT filename FIELD-a.v
New record.
----: I
0001= A
0002= 0
0003=  
0004=  
0005=  
0006=  
0007=  
0008= Tfilename;X;;a,v
0009= L
0010= 20
0011= 
Bottom at line 10.
----: FI
"FIELD-a.v" filed in file "DICT filename".

There is also an alternate syntax, with the
value number appended to the translate opcode,
in which case line 8 would look like

0008= Tfilename;Xv;;a

For example, you could get value 9 of attr 35 of the INV file
by doing either
   TINV;X;;35,9
or
   TINV;X9;;35

If possible, avoid this form and just use a simple "I" descriptor,
or, if you're on SB+, just use /FD to define field number "a.v"
(for example, field number "35.9").
-------
u2-users mailing list
[EMAIL PROTECTED]
To unsubscribe please visit http://listserver.u2ug.org/

Reply via email to