Fix regression tests

[olexandr-konovalov]

After #13, tests should be runnable

Under GAP's stable-4.10 branch there are two diffs:

########> Diff in /home/travis/build/gap-packages/sgpdec/gaproot/pkg/sgpdec/ts\
t/fl.tst:6
# Input is:
Range(IsomorphismPermGroup(flG));
# Expected output:
Group([ (1,2,3), (1,2) ])
# But found:
Group([ (2,3), (1,2,3) ])
########
########> Diff in /home/travis/build/gap-packages/sgpdec/gaproot/pkg/sgpdec/ts\
t/fl.tst:8
# Input is:
Range(IsomorphismPermGroup(Group(GeneratorsOfGroup(flG))));
# Expected output:
Group([ (1,3,2)(4,5,6), (1,5)(2,6)(3,4) ])
# But found:
Group([ (1,4)(2,5)(3,6), (1,3,2)(4,5,6) ])
########

and under the GAP master branch, there are three:

########> Diff in /home/travis/build/gap-packages/sgpdec/gaproot/pkg/sgpdec/ts\
t/fl.tst:6
# Input is:
Range(IsomorphismPermGroup(flG));
# Expected output:
Group([ (1,2,3), (1,2) ])
# But found:
Group([ (1,4)(2,5)(3,6), (1,3,2)(4,5,6) ])
########
########> Diff in /home/travis/build/gap-packages/sgpdec/gaproot/pkg/sgpdec/ts\
t/fl.tst:8
# Input is:
Range(IsomorphismPermGroup(Group(GeneratorsOfGroup(flG))));
# Expected output:
Group([ (1,3,2)(4,5,6), (1,5)(2,6)(3,4) ])
# But found:
Group([ (1,4)(2,5)(3,6), (1,3,2)(4,5,6) ])
########> Diff in /home/travis/build/gap-packages/sgpdec/gaproot/pkg/sgpdec/ts\
t/wreath.tst:19
# Input is:
ForAll(triples, function(t)
 local A,B,C,AB,BC,ABC,AB_C,A_BC;
 A := t[1];
 B := t[2];
 C := t[3];
 AB := Image(IsomorphismPermGroup(GroupWreathProduct(A,B)));
 BC := Image(IsomorphismPermGroup(GroupWreathProduct(B,C)));
 ABC := Image(IsomorphismPermGroup(GroupWreathProduct([A,B,C])));
 AB_C := Image(IsomorphismPermGroup(GroupWreathProduct([AB,C])));
 A_BC := Image(IsomorphismPermGroup(GroupWreathProduct([A,BC])));
 return IsomorphismGroups(ABC,A_BC)<>fail
        and
        IsomorphismGroups(ABC,AB_C)<>fail; end);
# Expected output:
true
# But found:
false
########

See the logs at https://travis-ci.org/gap-packages/sgpdec/builds/506023499.

The first two diffs can be fixed by making the test more robust and checking IdGroup of the Range instead of the exact group. The third problem in the GAP master branch requires a deeper investigation.

[egri-nagy]

Problems fixed, except the third one that fails in master. Temporarily I disabled the master branches for testing, as we need to focus on the stable release first. I leave this open, so I will remember to investigate it later.

[olexandr-konovalov]

If you will be making further changes to work under GAP stable-4.10, you may still be interested to ensure that they will not lead to even more failures under GAP master. So it's up to you, but I would not recommend disabling master branches for testing. if tests fail, you can inspect test logs to check why they fail, and you can merge pull request.