Restore old PostgreSQL 8.0 unicode dump to PostgreSQL 8.4

0
votes

I have an old backup of a PostgreSQL 8.0 database. PostgreSQL 8.0 has Unicode and plain backup options, and I used Unicode.

I tried to restore my backup using PostgreSQL 8.4 , which does not use Unicode format for backups. Rather, version 8.4 has UTF-8 backup encoding. When I restored the backup using version 8.4 it was restored, with some errors reported while restoring. But I found that some data was missing while I was checking the restore with my application.

Is there a way to restore data from an old Unicode format backup to UTF-8 format in PostgreSQL 8.4 without data loss?

1
postgresqlunicodeencodingutf-8
Please fix your grammar and add paragraph breaks, the details of your question are unclear. - o11c
ok sorry for that I need to know the procedure to restore postgresql 8.0 version backup file into postgresql8.4 version The problem is postgresql8.0 backup was encoded in unicode format and in 8.4 version It doesn't have the unicode format. 8.4 has utf-8 format default. I restored the plain backup using command prompt. but, coz of different encoding formats, the data was restored with some errors. Can you help me on this topic?? please. - HackoNe Naani
What's the exact text of the error message(s)? Also, is it really PostgreSQL 8.0, not something like Amazon Redshift? - Craig Ringer
error message is could not find function "headline_byname" in file D:/Program Files/Postgres/8.3/lib/tsearch2.dll I think there is no supported .dll files in 8.3 which are available in 8.0 version - HackoNe Naani
Did you try to find this error message with a search engine? Also, please find the instruction that causes this error message in your backup file. Post this instruction or the shortest possible script that reproduces the problem! - hagello

1 Answers

1
votes

The PostgreSQL developers have chosen to misuse the term Unicode for UTF-8. Or, as you can see on http://www.postgresql.org/docs/8.4/static/multibyte.html, UNICODE is an alias for UTF-8 in PostgreSQL.

So if all your assumptions are right, you try to restore UTF-8 as UTF-8.

I guess there is another problem anywhere.