Extend the range of precomputed simple groups from orders up to 10^18 to orders up to about 10^27

[hulpke]

(based on work of Stefan Kohl)

This is -- as basic first option -- simply extending the precomputed list of simple groups of "small" orders from 10^18 to about 10^27

Text for release notes

The range of SimpleGroupsIterator has been extended by 9 magnitudes

[coveralls]

Coverage increased (+0.03%) to 84.761% when pulling 734b395 on hulpke:additions into d1516f2 on gap-system:master.

[fingolfin]

In principle this is fine, but the documentation to SimpleGroupsIterator really should be adjusted to match the code change.

Ideally, a few more changes should be made (see my comment), but that's optional for this PR, and can be done by somebody else (e.g. me) in another PR.

[fingolfin]

Also: nice patch, thank you :-)

[fingolfin]

Nice! Minor (and optional) remarks added.

[fingolfin]

Optional alternative (should be double checked ;-):

Suggested change

	pos:=First([1..Length(SIMPLEGPSNONL2)],x->SIMPLEGPSNONL2[x][1]>=start);
	pos:=PositionProperty(SIMPLEGPSNONL2,x->x[1]>=start);

[Stefan-Kohl]

Why 10^27? -- This bound seems a bit arbitrary to me. -- Up to order 10^30, there are 2171 non-PSL2's, up to order 10^40, there are 22709 non-PSL2's, up to order 10^50, there are 298023 non-PSL2's and up to order 10^60, there are 4263813 non-PSL2's. -- Of course listing all the latter in the way you do it now would take hundreds of megabytes of text. -- But that's the reason for the iterator code, instead of using just a list.

[hulpke]

@Stefan-Kohl Yes, 10^27 is arbitary. I don't claim this is the best and end-of-all. If you want to offer your (better) general iterator code as PR, please do so, and I'd be happy to have my code replaced.