Thanks guys. I know what I will spend the night, doing ;). > On Oct 27, 2014, at 16:33 , Tom Lane <tgl@xxxxxxxxxxxxx> wrote: > > Chris Ruprecht <chris@xxxxxxxxxxx> writes: >> I can export the data out of that database into flat files just fine, but then I try to import the data to Postgres, I'm getting errors like this: > >> ERROR: invalid byte sequence for encoding "SQL_ASCII": 0x00 >> CONTEXT: COPY attachments, line 14: "58025 1 cl\Cert.r 10 M04P'15A415).($-H87)4:6UE+$-(05)!0U1%4BQ)3E!55"!I5&EM92!)3E1% M1T52'$585$523B!7..." > >> I tried LATIN1, SQL_ASCII, UTF-8, nothing works. I even tried to make the data type 'bytea', no luck. I'd love to have a "NO-CONVERSION" option on the copy command that just takes what ever bytes come along and doesn't try to interpret them. > > That's hardly possible, considering you're expecting COPY to recognize > field and record boundaries. What you probably need to do here is declare > the column as bytea and then write some sort of preprocessing script that > converts the binary data into hex-encoded form (basically \x followed by > hex digits, if memory serves, but check the description of bytea in the > PG manual). > > regards, tom lane -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
