Closed GoogleCodeExporter closed 9 years ago
Thanks for bug report.
It is an assertion which indicates there may be problem with serialization or
paging mechanism.
What type of strings were inserted? Constant, random? Did it contained some non
ASCII characters?
Original comment by kja...@gmail.com
on 2 May 2011 at 1:20
It's a Slovak lemma dictionary ( form -> lemma ) , so probably more random than
constant. It does contain non ASCII characters both in the key and value, for
example "ľšťžáí" etc..
Original comment by Tomas.Ha...@gmail.com
on 2 May 2011 at 2:09
Neposlal by jsi mi ten slovnik? Nemuzu to replikovat.
BTW jsem z Cech :-)
Original comment by kja...@gmail.com
on 10 May 2011 at 10:37
Neposlal by jsi mi ten slovnik? Nemuzu to replikovat.
BTW jsem z Cech :-)
Original comment by kja...@gmail.com
on 10 May 2011 at 10:37
V prilohe posielam zozipovany slovnik v CSV. Prvu hodnotu v subore pouzivam ako
kluc, druhu ako hodnotu.Pri zapise som flushoval kazdych 1000 zaznamov a
maximalne sa mi podarilo ulozit 300k, ak som ich ulozil viac tak sa uz citanie
padlo na "java.lang.InternalError: bytes left: 31"
Original comment by Tomas.Ha...@gmail.com
on 11 May 2011 at 7:23
Attachments:
Fixed SVN rev 73.
HashBucket decompression was wrong. It was catched by assertion. It is fixed in
SVN, I expect new release 2.1 in a few days.
Original comment by kja...@gmail.com
on 13 May 2011 at 11:27
Original issue reported on code.google.com by
Tomas.Ha...@gmail.com
on 2 May 2011 at 11:31