PL/PGSQL for permutations?

I've found a number of basic references for PL/PGSQL, but am looking for
something more complete. First question, is there available a *complete*
reference for PL/PGSQL? I'm using PostgreSQL version 7.2.3-5.73 (for
Redhat rpm packaging version), so it is version 7.2 I am concerned with
information on (this is a default version, so at least on this machine,
I'm not interested in updates in version).

The situation that leads me to ask for this is a simple need to select
two fields from a table, and generate a table that itself has two
fields, populated by every permutation of unique values from the first
table field pair. E.G., if I had in table 'one':
left  right
====  =====
a     b
a     c
b     d

...then I'd need a list of a, b, c, d, and produce a new table:
left  right
====  =====
a     b
a     c
a     d
b     a
b     c
b     d
c     a
c     b
c     d
d     a
d     b
d     c

This isn't hard, but I'm struggling with PL/PGSQL data types. I'd love
to just find a map or associative array type, and populate it from
queries, sort it, and generate a new array or map from a nested loop.
I'm not exactly sure how the PostgreSQL non-standard array type would
work for this from PL/PGSQL, if at all.

D. Stimits, stimits AT comcast DOT net