Closed GoogleCodeExporter closed 9 years ago
Oops, I uploaded incorrect versions of the tests. With the console output you
can see that the index key is present, and even get back the recid for the key,
but trying to get the value object returns null.
Original comment by keith.po...@gmail.com
on 6 May 2011 at 12:39
Attachments:
I tried Base64 encoding the byte arrays first to make sure there were no
non-ASCII characters being output, after reading the commentary for Issue 5.
Same problem.
Also, it's non-deterministic - the read fails on a different record each time,
so it's unlikely to do with the formatting of one of the records anyway.
The new version requires the Commons Codec jar for the Base64 encoding.
Original comment by keith.po...@gmail.com
on 6 May 2011 at 1:32
Attachments:
JDBM is fine. You used wrong method. On 'StoreReadTest.java' line 36 should be:
String s = index.getPrimaryValues(key).iterator().next();
Problem is that there are two methods with very similar name but different
meaning. One of them returns primary values for _secondary key_ other for
_primary key_. I will change method names in JDBM3.
Original comment by kja...@gmail.com
on 5 Jun 2011 at 1:57
What worked perfectly. Many thanks for setting me straight. This is a really
easy to use library by the way, great work.
I can proceed with my performance tests now. I don't suppose there's any
performance benchmark lying around ? One thing I've noticed so far is that
under Mac OS the writes are way faster than the reads which is very odd. It's
as expected on Windows. I'll have to get profiling...
Original comment by keith.po...@gmail.com
on 9 Jun 2011 at 2:04
Original issue reported on code.google.com by
keith.po...@gmail.com
on 6 May 2011 at 12:27Attachments: