pgsql: When enqueueing after-row triggers for updates of a table with a

Log Message:
-----------
When enqueueing after-row triggers for updates of a table with a foreign
key, compare the new and old row versions. If the foreign key column has
not changed, we needn't enqueue the trigger, since the update cannot
violate the foreign key. This optimization was previously applied in the
RI trigger function, but it is more efficient to avoid firing the trigger
altogether. Per recent discussion on pgsql-hackers.

Also add a regression test for some unintuitive foreign key behavior, and
refactor some code that deals with the OIDs of the various RI trigger
functions.

Modified Files:
--------------
    pgsql/src/backend/commands:
        tablecmds.c (r1.158 -> r1.159)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/commands/tablecmds.c.diff?r1=1.158&r2=1.159)
        trigger.c (r1.188 -> r1.189)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/commands/trigger.c.diff?r1=1.188&r2=1.189)
    pgsql/src/backend/utils/adt:
        ri_triggers.c (r1.78 -> r1.79)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/utils/adt/ri_triggers.c.diff?r1=1.78&r2=1.79)
    pgsql/src/include/commands:
        trigger.h (r1.53 -> r1.54)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/include/commands/trigger.h.diff?r1=1.53&r2=1.54)
    pgsql/src/test/regress/expected:
        foreign_key.out (r1.38 -> r1.39)

(http://developer.postgresql.org/cvsweb.cgi/pgsql/src/test/regress/expected/foreign_key.out.diff?r1=1.38&r2=1.39)
    pgsql/src/test/regress/sql:
        foreign_key.sql (r1.15 -> r1.16)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/test/regress/sql/foreign_key.sql.diff?r1=1.15&r2=1.16)