Re: update table from a csv file

On 12/27/2012 08:27 AM, Kirk Wythers wrote:
> I have been using COPY FROM to do a mass import of records from CSV files into a new database. I have discover however, a small number of records ( a few thousand) in one of the files that contain new data that needs to be added to the database, but on rows that have a primary key and have already been inserted (so I can't use COPY FROM because it violates the primary key).
>
> If the structure of the table is
>
> id data1 data2 data3
>
> and the structure of the CSV file is
>
> id data1 data2 data3
>
> and I need to update all the rows in data3 where the id = id.
>
> I have created a temporary table and used COPY FROM to load the update data into the temporary table. I seem to be stuck however. I thought I should be able to use the UPDATE command to update all columns and all rows the table.id = tmp_table.id
>
> Something like:
>
> UPDATE table FROM tmp_table WHERE table.id = tmp_table.id;
>
> Or am I completely off course?

No. Some questions though.

What version pf Postgres?
Is that the actual UPDATE statement, I see no SET?
Have you tried it?
If so and it failed what was the error?

>

--
Adrian Klaver
adrian(dot)klaver(at)gmail(dot)com