To quote from the JDB Wiki:
There are three ways you can refer to columns in tables that are linked with foreign keys. · The canonical way is a list of referencing columns in the form: col.col...col . For example, in the sed <http://www.jsoftware.com/jwiki/JDB/Examples#sed> database, column popul in table r can be referenced from table t as enum.secnum.rnum.popul. Note that this method gives a unique reference. * A column can be referenced by name only if it is unique in the database, as in column popul in the sed database. * A column can be referenced in the form table.name. Note that the last two references work correctly only when there is a single route from the original table to the column, i.e. when no table has two columns both referencing the same table as foreign key. However, I have found that they work fine with non-unique paths and use the earliest defined path by default. I encountered this issue in a situation where I had non-unique paths to the same value, so I think I can completely ignore the fact that there are multiple paths (the example below has distinct values). Create__hd 'TFTABLE';'boo boolean;truth char' Create__hd 'ANDTABLE';'andls boolean,andrs boolean;landr TFTABLE' Create__hd 'ORTABLE';'orls boolean,orrs boolean;lorr TFTABLE' Insert__hd 'TFTABLE';< 0 1;'F',:'T' Insert__hd 'ORTABLE';< 0 0 1 1;0 1 0 1;0 1 1 1 Insert__hd 'ANDTABLE';< 0 0 1 1;0 1 0 1;0 0 0 1 Create__hd 'ANDORTABLE';'ls boolean,rs boolean;landrkey ANDTABLE,lorrkey ORTABLE' Insert__hd 'ANDORTABLE';< (],,~@<@]) 0 0 1 1;0 1 0 1 Reads__hd '* from TFTABLE' ┌───┬─────┐ │boo│truth│ ├───┼─────┤ │0 │F │ │1 │T │ └───┴─────┘ Reads__hd '** from ANDTABLE' ┌─────┬─────┬─────┬─────┐ │andls│andrs│landr│truth│ ├─────┼─────┼─────┼─────┤ │0 │0 │0 │F │ │0 │1 │0 │F │ │1 │0 │0 │F │ │1 │1 │1 │T │ └─────┴─────┴─────┴─────┘ Reads__hd '** from ORTABLE' ┌────┬────┬────┬─────┐ │orls│orrs│lorr│truth│ ├────┼────┼────┼─────┤ │0 │0 │0 │F │ │0 │1 │1 │T │ │1 │0 │1 │T │ │1 │1 │1 │T │ └────┴────┴────┴─────┘ Reads__hd '** from ANDORTABLE' ┌──┬──┬─────┬─────┬─────┬─────┬────┬────┬────┬─────────────┐ │ls│rs│andls│andrs│landr│truth│orls│orrs│lorr│lorrkey.truth│ ├──┼──┼─────┼─────┼─────┼─────┼────┼────┼────┼─────────────┤ │0 │0 │0 │0 │0 │F │0 │0 │0 │F │ │0 │1 │0 │1 │0 │F │0 │1 │1 │T │ │1 │0 │1 │0 │0 │F │1 │0 │1 │T │ │1 │1 │1 │1 │1 │T │1 │1 │1 │T │ └──┴──┴─────┴─────┴─────┴─────┴────┴────┴────┴─────────────┘ Reads__hd 'truth,TFTABLE.truth,landrkey.truth,lorrkey.truth from ANDORTABLE' ┌─────┬─────────────┬──────────────┬─────────────┐ │truth│TFTABLE.truth│landrkey.truth│lorrkey.truth│ ├─────┼─────────────┼──────────────┼─────────────┤ │F │F │F │F │ │F │F │F │T │ │F │F │F │T │ │T │T │T │T │ └─────┴─────────────┴──────────────┴─────────────┘ Has anyone played with non-unique paths in JDB and found any situations where it doesn't simply take the earliest defined path as default? Thanks, Jordan ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
