I am trying to import a trivial CSV to Postgres 8.4 database:
Here is a table:
CREATE TABLE public.sample (
a VARCHAR,
b VARCHAR
) WITHOUT OIDS;
Here is a CSV file sample:
"foo","bar, baz"
The query:
COPY sample FROM '/tmp/sample.csv' USING DELIMITERS ',';
Throws an exception:
ERROR: extra data after last expected column
CONTEXT: COPY sample, line 1: ""foo","bar, baz""
But, well, CSV parsing is not rocket science and I wonder if it is not possible to solve without reformatting the source CSV file.
Input comes from a 3rd party, I cannot change the format.
(I understand I could pre-process to change the delimiter before I import it.)
Final solution:
COPY sample FROM '/tmp/sample.csv' WITH DELIMITER ',' CSV;
Originally taken from https://stackoverflow.com/a/9682174/251311, and documentation page is http://www.postgresql.org/docs/8.4/static/sql-copy.html
Clearly, you have a CSV file while you try to import it as text format. For Postgres 9.1 or newer, use:
COPY sample FROM '/tmp/sample.csv' (FORMAT csv);
The default delimiter for CSV format is the comma (,
) anyway. More in the manual.
For PostgreSQL 8.4 or older:
COPY sample FROM '/tmp/sample.csv' CSV;