On Fri, Jan 6, 2012 at 6:20 PM, Adrian Klaver <adrian.klaver@xxxxxxxxx> wrote: > http://www.postgresql.org/docs/9.0/interactive/sql-copy.html > > Search for > NULL Thanks Adrian. Without examples, it's hard to predict syntax. If the value after a pipe is missing altogether, I suppose the missing value is "\n" (newline). But this doesn't work: copy vl from 'data.txt' WITH DELIMITER AS '|' NULL '\n'; None of these work either: copy vl from 'data.txt' WITH DELIMITER AS '|' NULL \n; copy vl from 'data.txt' WITH DELIMITER AS '|' NULL \\n; copy vl from 'data.txt' WITH DELIMITER AS '|' NULL ''; The first two give errors, the third one throws the same missing value for column error. The data is stored like this: 123|big string here|189209209|US|2001-01-01 123|big string here|189209209|US|2001-01-01 123|big string here|189209209|US|2001-01-01 123|big string here|189209209|US|2001-01-01 But sometimes, the strings are: |big string here|189209209|US|2001-01-01 |big string here|189209209|US|2001-01-01 Or 123|big string here|189209209|US 123|big string here|189209209|US| So you see either the first column, which is the ID in a way, is missing so the "missing character" is probably a blank (''?). In this case I want COPY to just ignore this line. Or the last column is missing, where the missing character can be a newline I suppose? So how do I specify this in the COPY command so that it doesn't croak? If a line's ID is missing, it should ignore the line and go on instead of not doing anything by throwing an error for EVERYTHING! Thanks. -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
