Matrix groups over ZZ (and possibly other rings) weirdness

[lgoettgens]

Describe the bug
The trivial matrix group of dim 1x1 behaves differently depending on the base ring.

To Reproduce

julia> GQQ = matrix_group([matrix(QQ, 1, 1, [1])])
Matrix group of degree 1
  over rational field

julia> GZZ = matrix_group([matrix(ZZ, 1, 1, [1])])
Matrix group of degree 1
  over integer ring

julia> describe(GQQ)
"1"

julia> describe(GZZ)
ERROR: MethodError: no method matching isfinite(::ZZRing)

Closest candidates are:
  isfinite(::DecFP.Dec128)
   @ DecFP ~/.julia/packages/DecFP/Cud5F/src/DecFP.jl:540
  isfinite(::RealFieldElem)
   @ Nemo ~/.julia/packages/Nemo/tzyHK/src/arb/Real.jl:476
  isfinite(::Missing)
   @ Base missing.jl:101
  ...

Stacktrace:
 [1] is_infinite(x::ZZRing)
   @ Nemo ~/.julia/packages/Nemo/tzyHK/src/Infinity.jl:105
 [2] describe(G::MatrixGroup{ZZRingElem, ZZMatrix})
   @ Oscar ~/code/julia/Oscar.jl/src/Groups/GAPGroups.jl:2443
 [3] top-level scope
   @ REPL[45]:1

julia> small_generating_set(GQQ)
1-element Vector{MatrixGroupElem{QQFieldElem, QQMatrix}}:
 [1]

julia> small_generating_set(GZZ)
ERROR: MethodError: no method matching isfinite(::ZZRing)

Closest candidates are:
  isfinite(::DecFP.Dec128)
   @ DecFP ~/.julia/packages/DecFP/Cud5F/src/DecFP.jl:540
  isfinite(::RealFieldElem)
   @ Nemo ~/.julia/packages/Nemo/tzyHK/src/arb/Real.jl:476
  isfinite(::Missing)
   @ Base missing.jl:101
  ...

Stacktrace:
 [1] is_infinite(x::ZZRing)
   @ Nemo ~/.julia/packages/Nemo/tzyHK/src/Infinity.jl:105
 [2] small_generating_set(G::MatrixGroup{ZZRingElem, ZZMatrix})
   @ Oscar ~/code/julia/Oscar.jl/src/Groups/GAPGroups.jl:641
 [3] top-level scope
   @ REPL[47]:1

This boils down to is_finite(::ZZRing) not being implemented. But I think adding this just fixes the symptom here, not the underlying issue. Instead,

Oscar.jl/src/Groups/GAPGroups.jl

Line 2443 in 615af44

if G isa MatrixGroup && is_infinite(base_ring(G))

and

Oscar.jl/src/Groups/GAPGroups.jl

Line 641 in 615af44

if G isa MatrixGroup && is_infinite(base_ring(G))

and other similar places should have a workaround for is_finite not being available for base_ring(G).

Thanks to @flyingapfopenguin for making me aware of this issue.