[gnugo-devel] more tests

[Paul Pogonyshev]

all these tests are around one position. they are related to badly
connected dragons (as usual :). i'm going to work on this problem and
will use these tests for checking.

test 1 (273) currently passes. i think it is needed to check if we
make gnugo to optimistic about cutting this dragon.

other three tests (274 - 276) currently fails and i hope to fix it
someday.

new file (regression/games/owl39.sgf):

(
;
FF[4]
GM[1]
SZ[9]
AB[de][ee][fe][ag][bg][cg][dg][eg][gg][gh][hg][hc][af][ac][bc][cc][dc][db][eb][fb][hf][hb][id]
AW[ge][gd][fc][ec][ed][dd][cd][be][bf][cf][df][ef][ff]
;
B[gc]
)


Index: owl1.tst
===================================================================
RCS file: /cvsroot/gnugo/gnugo/regression/owl1.tst,v
retrieving revision 1.11
diff -u -r1.11 owl1.tst
--- owl1.tst    12 Sep 2002 08:20:26 -0000      1.11
+++ owl1.tst    13 Sep 2002 19:56:37 -0000
@@ -38,7 +38,7 @@
 269 owl_defend D11
 #? [1 E9]*
 
-# tests 270 - 272 refer to badly connected dragon eyespaces
+# tests 270 - 276 refer to badly connected dragon eyespaces
 loadsgf games/owl37.sgf
 270 owl_attack A4
 #? [1 A5]*
@@ -50,6 +50,19 @@
 272 owl_defend H12
 #? [1 F13]*
 
+loadsgf games/owl39.sgf 1
+273 owl_attack E7
+#? [0]
+
+274 owl_threaten_attack E7
+#? [1 (G7 G4|G4 G7)]*
+
+loadsgf games/owl39.sgf
+275 owl_attack E7
+#? [G4]*
+
+276 owl_defend E7
+#? [G4]*
 
 ########### end of tests #####################