Slightly extend MinimalGeneratingSet for groups

[wilfwilson]

Currently, MinimalGeneratingSet only supports soluble groups. Cyclic groups are soluble, and so a group that is not soluble requires at least two generators. Therefore, in MinimalGeneratingSet, we can add a special case for groups that aren't soluable but already a generating set of size two.

So this behaviour:

gap> MinimalGeneratingSet(SymmetricGroup(5));
Error, `MinimalGeneratingSet' currently assumes that the group must be solvable.
In general, try `SmallGeneratingSet' instead, which returns a generating
set that is small but not of guarateed smallest cardinality

now becomes:

gap> MinimalGeneratingSet(SymmetricGroup(5));
[ (1,2,3,4,5), (1,2) ]

I'm wondering, do people think this is worth doing? And if so, have I gone about it the right way?

[codecov]

Codecov Report

Merging #1755 into master will decrease coverage by <.01%.
The diff coverage is 57.14%.

@@            Coverage Diff             @@
##           master    #1755      +/-   ##
==========================================
- Coverage   63.43%   63.43%   -0.01%     
==========================================
  Files        1000     1000              
  Lines      322760   322766       +6     
  Branches    13015    13038      +23     
==========================================
- Hits       204757   204756       -1     
- Misses     115361   115369       +8     
+ Partials     2642     2641       -1

Impacted Files	Coverage Δ
lib/grp.gi	84.13% <57.14%> (-0.01%)	⬇️
lib/queue.g	66.4% <0%> (-3.2%)	⬇️
hpcgap/lib/hpc/stdtasks.g	38.61% <0%> (-0.26%)	⬇️
src/funcs.c	71.01% <0%> (-0.15%)	⬇️
src/hpc/threadapi.c	34.36% <0%> (-0.1%)	⬇️
src/hpc/traverse.c	78.29% <0%> (+0.08%)	⬆️
src/lists.c	55.08% <0%> (+0.11%)	⬆️

[markuspf]

I think this is worth doing. Maybe @hulpke or @stevelinton have opinions on this too?

[fingolfin]

Looks good to me