db-derby-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Myrna van Lunteren" <m.v.lunte...@gmail.com>
Subject Re: order by
Date Wed, 01 Mar 2006 14:44:41 GMT
On 3/1/06, Michael Segel <msegel@segel.com> wrote:
> Ok...
> You said that the problem only occurs with your one result set.
> When you say that you're getting the data from a spreadsheet, how are you
> reading it in? Are you trying to access the raw XLS dataset? Or did you
> export the data as a .csv file?
> I don't believe it's a bug because of the fact that it only occurs in one
> data set, and that you could be introducing an error in a couple of
> places.
> (The data set itself for example.)(load program)
> > -----Original Message-----
> > From: news [mailto:news@sea.gmane.org] On Behalf Of tom
> > Sent: Wednesday, March 01, 2006 5:01 AM
> > To: derby-user@db.apache.org
> > Subject: Re: order by
> >
> >
> >
> > There still something is wrong. I recreated the database without
> > composite index, imported 1200 records without error, viewed
> > them and the programm was hanging at record 1100 without messages,
> > I killed the process after waiting 5 minutes and restarted
> > derby network server. I created a composite index and viewed the
> > data, now the sorting problem that I described at the beginning of this
> > thread appears again. pooh
> >

I agree with Michael, it still looks like there is some odd stuff in your
spreadsheet or your import program.

Maybe spurious characters (i.e. characters that are not interpreted by your
encoding and don't actually show up when you look at your spreadsheet), a
tab in a data field that gets interpreted as a new column/field by the
import program so you have an extra field the import program can't
handle...an occurrence of your field delimiter (if you have such a thing) in
a text string (same effect)...

In a past life in tech support I once had to take the 'binary' approach to
find exactly such a situation - there were no errors that helped... It was
labor intensive, bothersome, but got through the hurdle. I think there were
a couple of thousand rows.

I had to cut the dataset in sections, ignore the bits that worked, cut up
the bits that didn't work further, until I found the troublesome
record/records/entries. Then fixed those up in the original, and then  the
whole process worked.
Then I informed the customer to prevent whatever it was that caused the

Unless someone knows another way to do this?


View raw message