home *** CD-ROM | disk | FTP | other *** search
-
- % Adjunct to "PrlgDigest".
- % ========================
-
- /*
- This monitor will not detect a data conflict in the collected clues
- files as they have been "Passed".
-
- Developed empirically from anomalies thus far encountered it may help
- where the user collects the clues.
-
- Addendum.
- ---------
- Should an imported logic problem be offered to "PrlgDigest" for analysis
- it can be expected that additional ad hoc utilities will be required. In
- this case collect any extra utilities under a prolog procedure called
- "do" and insert it in the new clues file that has been generated. See
- cluesfile "a" for an example of this technique.
- The new clues file can of course be loaded along with the "s" file.
- Hopefully it will then merge well. Outputs will then be seen, perhaps
- followed by an "Inadequate data" signal as the new file evolves.
- */
-
- monitor :-
- nl,nl,
- write(' "Monitor" present.'),nl,nl,
- fail.
-
- monitor :- nl,
- c(X,Y),
- i(X,Y),!,
- write(c(X,Y)),tab(5), % BUG detected
- write(i(X,Y)),nl,
- write('
- Cannot have both c(X,Y) and i(X,Y) with the same identities.'),
- nl,
- fail.
-
-
- % Don't link two members of the same group to each other.
-
- monitor :-
- c(X,Y),
- group(G),
- member2(X,G),
- member2(Y,G),!,
- write(c(X,Y)),nl,
- write('Two members of the same category are not to be linked.'),
- nl,
- fail.
-
- monitor :-
- group(G),
- i(X,Y),
- member2(X,G),
- member2(Y,G),!,
- write(i(X,Y)),nl,
- write(' Must not link two members of the same group.'),nl,
- fail.
-
- % We must now tell "monitor" that it has nothing further to do.
-
- monitor.
-
-