Just for the hack of it, can I peddle a postgres version of a 'solution'?
Assuming a postgres instance, with module file_fdw installed, and your 5 text lines tab-separated saved under '/tmp/table.txt':
create server abc_server foreign data wrapper file_fdw; create foreign table file_table ( c1 text, i2 int, i3 int) server abc_server options ( filename '/tmp/table.txt', format 'csv', delimiter E'\t' + ); select c1, i2, i3 from file_table order by c1; select c1, i2 from file_table order by i2 desc; select c1, i3 from file_table order by i3 desc; select c1 , rank() over (order by i2 desc) , rank() over (order by i3 desc) from file_table order by c1;
Output (mimicking yours):
$ psql -qXf table.sql c1 | i2 | i3 ----+----+---- a | 23 | 11 b | 24 | 15 c | 16 | 19 d | 8 | 2 e | 40 | 41 (5 rows) c1 | i2 ----+---- e | 40 b | 24 a | 23 c | 16 d | 8 (5 rows) c1 | i3 ----+---- e | 41 c | 19 b | 15 a | 11 d | 2 (5 rows) c1 | rank | rank ----+------+------ a | 3 | 4 b | 2 | 3 c | 4 | 2 d | 5 | 5 e | 1 | 1 (5 rows)
Not entirely serious as a solution but I thought is was cute all the same: it is also the shortest till now, with only 3 lines :)
In reply to Re: manipulating a data table
by erix
in thread manipulating a data table
by v15
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |