double threat test cases, in hope of future implementation

[alain]

Testcases of double threat move, not yet implemented in GNU Go. (examples found on computer-go list or on KGS).

• Double ladder breaker, from Rémi Coulom and John Tromp.
• Life with miai move by Sanghyeon Seo.
• 3 Fuseki with ladder breakers (thanks to Min-u Kang).
• attack_either / defend_both , from comments in engine/reading.c

GNU GO task list, in paragraph Long term issues:

* A much improved combination module.  The combination module of
   today only finds combinations of threats to capture enemy groups.
   A more useful combination module would e.g. find combinations of
   threats to capture a group or enter opponent territory.  It would
   also be strong enough to find combinations of strategic moves and
   more indirect threats (a threat to a threat).  Possibly it could
   combine threats in AND-OR trees (DAGs?) that could be searched
   using ordinary tree search algorithms.  (Revision of combination.c
   is underway.)

And in Ideas:

 * Generalize representation and handling of threats so that we have a
   graph representation of threats that can be searched to see how
   different threats interact.

[alain]

move in K9 breaks both ladders, so save one, and win the game

[alain]

double threat to either capture or make a second eye. Winning move.

[alain]

fuseki with ladder, and ladder breaker on opposite corner

[alain]

invisible ladder (but well known)

[alain]

complicate fight (joseki?) and ladder breaker

[alain]

example from comments in engine/reading.c

[nando]

About miai-problem-1ply-ahead.sgf, this one can be solved at OWL level.

Something along these lines should be enough:

Pattern DCnn

.OX
XXO
.O*

:8,c, threatens_to_capture(A), value(35)

.CX
AAO
bO*

; olib(b)==1 && oplay_attack(*,A) && !defend(C)

(the olib clause is possibly superfluous)

[arend]

Alain, could you maybe add regression tests for these examples? (E.g. in strategy5.tst.) Ask if you are unsure how to do this.