From: | "Dan Black" <fireworker(at)gmail(dot)com> |
---|---|
To: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: Charset conversion error |
Date: | 2006-05-30 16:04:12 |
Message-ID: | 27f606250605300904q1ea04a21m6336dc4c2927dc98@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
You are right. I was in a harry when wrote my last message.
I solved my problem by changing some source files. But It is not very good
because I have to do it every time when I need to update my database. And I
can't check all fields in all tables because it take a few days.
2006/5/30, Tatsuo Ishii <ishii(at)sraoss(dot)co(dot)jp>:
>
> It seems not.
>
> $ cat koi8.sh
> printf "SELECT '==\xb9=='"|PGCLIENTENCODING=WIN1251 psql -p 5435 koi8
>
> $ psql -p 5435 -c 'select version()' koi8
> version
>
> ---------------------------------------------------------------------------------------------------------
> PostgreSQL 8.1.3 on i686-pc-linux-gnu, compiled by GCC gcc (GCC) 3.3.220031218 (Vine Linux
> 3.3.2-0vl8)
> (1 row)
>
> $ sh koi8.sh
> ?column?
> ----------
> == ==
> (1 row)
>
> As you can see PostgreSQL 8.1.3 converts 0xb9 in Windows-1251 to a
> space of KOI8.
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
>
> > Yes, character converted to a non-space-char.
> > And there is reverse error. It's happened when I try to extract
> information
> > from database.
> > SQL Error: ERROR: character 0x8bbf of encoding "MULE_INTERNAL" has no
> > equivalent in "WIN1251""
> >
> > 2006/5/30, Tatsuo Ishii <ishii(at)sraoss(dot)co(dot)jp>:
> > >
> > > As far as I know the conversion table was not changed. Are you saying
> > > that Win-1251 0xb9 is converted to a non-space-char in the previous
> > > version?
> > > --
> > > Tatsuo Ishii
> > > SRA OSS, Inc. Japan
> > >
> > > > Previous version converted such characters all right. And there
> wasn't
> > > any
> > > > bogus ASCII spaces. But I looked at the KOI8 charset table and found
> out
> > > > that there is not equivalent symbol (0xb9) in this table.
> > > >
> > > > 2006/5/30, Tatsuo Ishii <ishii(at)sraoss(dot)co(dot)jp>:
> > > > >
> > > > > > I think you are right. But everything was alright before I
> updated
> > > my
> > > > > > database.
> > > > >
> > > > > Previous version converted such characters to ASCII spaces. So
> > > > > probably you have lots of bogus spaces anyway. If you think it's
> ok,
> > > > > then you could your own CONVERSION which behaves similar to
> previous
> > > > > version.
> > > > > --
> > > > > Tatsuo Ishii
> > > > > SRA OSS, Inc. Japan
> > > > >
> > > > > > So there are a lot of "incorrect" values in tables. And errors
> > > appear
> > > > > when I
> > > > > > execute "SELECT * FROM table".
> > > > > >
> > > > > > 2006/5/30, Tatsuo Ishii <ishii(at)sraoss(dot)co(dot)jp>:
> > > > > > >
> > > > > > > > Hi, I have an error after updating my database up to
> 8.1.4version.
> > > > > > > > "SQL Error: ERROR: character 0xb9 of encoding "WIN1251" has
> no
> > > > > > > equivalent
> > > > > > > > in "MULE_INTERNAL"'. "
> > > > > > > > My client program encoding is windows-1251 and database
> encoding
> > > is
> > > > > > > koi8.
> > > > > > > > What can I do to rectify the situation?
> > > > > > >
> > > > > > > It suggests that Windows-1251's 0xb9 cannot be converted to
> KOI8.
> > > You
> > > > > > > should check your Windows-1251 data.
> > > > > > > --
> > > > > > > Tatsuo Ishii
> > > > > > > SRA OSS, Inc. Japan
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > --
> > > > > > Verba volent, scripta manent
> > > > > > Dan Black
> > > > >
> > > >
> > > >
> > > >
> > > > --
> > > > Verba volent, scripta manent
> > > > Dan Black
> > >
> >
> >
> >
> > --
> > Verba volent, scripta manent
> > Dan Black
>
--
Verba volent, scripta manent
Dan Black
From | Date | Subject | |
---|---|---|---|
Next Message | SCassidy | 2006-05-30 16:05:42 | Re: UTF-8 context of BYTEA datatype?? |
Previous Message | John DeSoi | 2006-05-30 15:19:42 | Lossy character conversion to Latin-1 |