Re: Using SQL to group records with similar elements

On Thu, Jul 21, 2011 at 4:35 PM, Alan Shore <ashore@xxxxxxxx> wrote:

In trying to explain my problem to one of my colleagues,
I may have inadvertently stumbled upon the answer.

SELECT * FROM Example WHERE
field1 in ('360', '160', '202', '400','118', ' ') and
field2 in ('360', '160', '202', '400','118', ' ') and
field3 in ('360', '160', '202', '400','118', ' ') and
field4 in ('360', '160', '202', '400','118', ' ') and
field5 in ('360', '160', '202', '400','118', ' ')

I can't believe it seems so simple.

I think it is that simple. However, it's still on the unwieldy side,
especially as you have to check each of 21 fields against the 21
values read in from the first record.

This is something I would pull out iSeries Python for. Its code would
look something like

f = File400('MYFILE', 'r', lib='MYLIB')
master = None
result = []
f.posf()
while not f.readn():
values = set(f.get()[1:])
if master is None:
master = values
if values <= master:
result.append(f.get())

After that runs, result will contain your desired records. This works
no matter how many fields MYFILE has (could be hundreds, if you want),
as long as the first field is the key and the remaining fields contain
the values to check.

John Y.