dev-gap/polycyclic: patch out a failing test

Upstream has already recommended that we disable the test suite for this
package, so we may be on our own here, but this one is again easy to
patch out and the rest of the test suite passes.

Signed-off-by: Michael Orlitzky <mjo@gentoo.org>

dev-gap/polycyclic/files/polycyclic-2.16-failing-test.patch

@@ -0,0 +1,28 @@
+From 019d573ddbade01304f62b9cd07b4e759e21398a Mon Sep 17 00:00:00 2001
+From: Michael Orlitzky <michael@orlitzky.com>
+Date: Tue, 19 Mar 2024 16:38:13 -0400
+Subject: [PATCH] tst/bugfix.tst: delete failing test
+
+https://github.com/gap-packages/polycyclic/issues/89
+---
+ tst/bugfix.tst | 4 ----
+ 1 file changed, 4 deletions(-)
+
+diff --git a/tst/bugfix.tst b/tst/bugfix.tst
+index 393390e..9293671 100644
+--- a/tst/bugfix.tst
++++ b/tst/bugfix.tst
+@@ -58,10 +58,6 @@ gap> NonAbelianTensorSquare(Image(iso*iso1));
+ Pcp-group with orders [ 2, 2, 3, 2, 2, 2, 2 ]
+ 
+ #
+-gap> # The problem with the previous example is/was that Igs(G)
+-gap> # is set to a non-standard value:
+-gap> Igs(G);
+-[ g1, g2*g5, g3*g4*g5^2, g4*g5, g5 ]
+ gap> # Unfortunately, it seems that a lot of code that
+ gap> # really should be using Ngs or Cgs is using Igs incorrectly.
+ gap> # For example, direct products could return *invalid* embeddings:
+-- 
+2.43.2
+

dev-gap/polycyclic/polycyclic-2.16.ebuild

@@ -22,7 +22,11 @@ PDEPEND="dev-gap/alnuth"
 # into them, upstream recommends disabling it:
 #
 #  https://github.com/gap-packages/polycyclic/issues/46
+#  https://github.com/gap-packages/polycyclic/issues/89
 #
-PATCHES=( "${FILESDIR}/${P}-hanging-tests.patch" )
+PATCHES=(
+	"${FILESDIR}/${P}-hanging-tests.patch"
+	"${FILESDIR}/${P}-failing-test.patch"
+)
 
 gap-pkg_enable_tests