On 24/10/2011 22:39, Allan Kamau wrote: > On Mon, Oct 24, 2011 at 11:29 PM, Raymond O'Donnell <rod@xxxxxx> wrote: >> On 24/10/2011 20:23, Allan Kamau wrote: >>> Hi, >>> I have a tab delimited file with over a thousand fields (columns) >>> which I would like to import into postgreSQL. >>> >>> I have opted to import the entire record (line) of this file into a >>> single field in a table, one table record per file line. Later split >>> the contents of the field accordingly into an array for further >>> processing and decomposition. >>> >>> I am using the COPY command as follows >>> #COPY a.t(raw_data)FROM '/data/tmp/t.txt'; >>> >>> and it reports "ERROR: extra data after last expected column". >> >> I'd guess that there's a character in the input which PG is interpreting >> as a column delimiter. Use the WITH DELIMITER option to set the expected >> delimiter to something that won't appear in the input. >> >> Ray. >> >> >> -- >> Raymond O'Donnell :: Galway :: Ireland >> rod@xxxxxx >> > > Yes PG senses the tab and assumes it to be a field delimiter. I have > now used the '=' character as my "csv" delimiter after grepping the > file on '=' and seeing no lines were returned. > > Out of curiosity, what would be the correct use of "FORMAT text" > option in COPY command. The command below does give me an error as > indicated in my previous email. > > COPY a.t(raw_data)FROM '/data/tmp/t.txt' FORMAT text; I honestly don't know - it looks fine to me. The docs suggest that it'd give an error on pre-9.0 Postgres; but you said you're on 9.1... <shrug>. Ray. -- Raymond O'Donnell :: Galway :: Ireland rod@xxxxxx -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
