I'm facing the ORA-04091 with a trigger I'm trying to write.
After reading Google on the topic, I now understand that it is
not allowed to have the trigger code refer to the table from
which the trigger was activated.

So, I'm facing a design issue here, and would appreciate any
help. Simplified, I'm having two tables, parent and subsidiary.
Each row of subsidiary refers to a row in parent (many-to-one;
there can be several subsidiary rows referring to a single parent row).

What I attempt to do with the trigger is to remove the parent row
when there are no more subsidiary rows pointing to it (so, when
subsidiary is removed, check which was the parent for this subsidiary,
and are there any subsidiaries left with the same parent -- and if
not, remove the parent along with the subsidiary).

This doesn't sound like anything extraordinary or uncommon -- but
I don't seem to find the correct pattern for implementing this.
Any pointers for solutions?

Thanks,
--
Wolf  a.k.a.  Juha Laiho     Espoo, Finland
(GC 3.0) GIT d- s+: a C++ ULSH++++$ P++@ L+++ E- W+$@ N++ !K w !O !M V
         PS(+) PE Y+ PGP(+) t- 5 !X R !tv b+ !DI D G e+ h---- r+++ y++++
"...cancel my subscription to the resurrection!" (Jim Morrison)

set up a before statement trigger, initializing a collection of id's
after each row: buffer the subsidiary id
after statement: loop through your collection, issue your select
statement, and deal with the parent accordingly
Classical solution , available on http://asktom.oracle.com

--
Sybrand Bakker
Senior Oracle DBA

-----------------------------------------------Reply-----------------------------------------------