Refactor FastNormalClosure and don't use it when kernel is abelian

[ssiccha]

• Refactor FastNormalClosure
• Don't use FastNormalClosure if kernel is abelian

[ssiccha]

I added a comment referencing Lemma 2.3.8 from "Permutation Group Algorithms" by Akos Seress, which reads:

Lemma 2.3.8. Suppose that G = <S> acts on H and let U ⊆ H . Let g be a random subproduct of S and u be a random subproduct of U . If is not closed for the G-action then Prob(u ^ g not in ) ≥ 1/4.

I addressed all comments and added two small commits:

• Add small optimisation to FastNormalClosure
• Fix documentation of findgensNmeth

[codecov[bot]]

Codecov Report

Merging #282 (e987ad8) into master (0401141) will increase coverage by 0.11%. The diff coverage is 100.00%.

:exclamation: Current head e987ad8 differs from pull request most recent head 01b29b1. Consider uploading reports for the commit 01b29b1 to get more accurate results

@@            Coverage Diff             @@
##           master     #282      +/-   ##
==========================================
+ Coverage   77.83%   77.94%   +0.11%     
==========================================
  Files          43       43              
  Lines       18385    18393       +8     
==========================================
+ Hits        14310    14337      +27     
+ Misses       4075     4056      -19     

Impacted Files	Coverage Δ
gap/generic/kernel.gi	89.74% <100.00%> (-0.13%)	:arrow_down:
gap/matrix.gi	94.19% <100.00%> (ø)
gap/projective/d247.gi	74.92% <0.00%> (-1.27%)	:arrow_down:
gap/projective/AnSnOnFDPM.gi	76.86% <0.00%> (-0.13%)	:arrow_down:
gap/projective/c6.gi	61.98% <0.00%> (-0.12%)	:arrow_down:
gap/base/recognition.gd	100.00% <0.00%> (ø)
gap/perm.gi	97.27% <0.00%> (+0.04%)	:arrow_up:
gap/projective/almostsimple.gi	65.94% <0.00%> (+0.09%)	:arrow_up:
gap/matrix/ppd.gi	93.27% <0.00%> (+0.11%)	:arrow_up:
gap/matrix/classical.gi	75.74% <0.00%> (+0.24%)	:arrow_up:
... and 2 more

[ssiccha]

Rebased onto master.