Fix unexpected error in ConjugacyClassesSubgroups

This disables an attempt to use cyclic extension instead of lifting for
lattices of smallish groups. The issue is that CE returns data in a
slightly different format than required. This resulted in some commands,
e.g. `ConjugacyClassesSubgroups(SmallGroup(1344,11293))`, to enter a
break loop.

If tomlib is loded it's a mote issue, and thus does not merit further
investigation.

This resolves #4854

lib/grplatt.gi

@@ -845,11 +845,15 @@ InstallGlobalFunction(LatticeViaRadical,function(arg)
   makesubgroupclasses,cefastersize;
 
   #group order below which cyclic extension is usually faster
-  if IsPackageMarkedForLoading("tomlib","")=true then
+  # WORKAROUND: there is a disparity between the data format returned
+  # by CE and what this code expects. This could be resolved properly,
+  # but since most people will have tomlib loaded anyway, this doesn't
+  # seem worth the effort.
+  #if IsPackageMarkedForLoading("tomlib","")=true then
     cefastersize:=1; 
-  else
-    cefastersize:=40000; 
-  fi;
+  #else
+  #  cefastersize:=40000; 
+  #fi;
 
   makesubgroupclasses:=function(g,l)
   local i,m,c;

tst/testbugfix/2022-09-06-ConjugacyClassesSubgroups.tst

@@ -0,0 +1,5 @@
+# The following used to work in GAP 4.10, but was broken in GAP
+# 4.11.x and also 4.12.0.
+# See https://github.com/gap-system/gap/issues/4854
+gap> Length(ConjugacyClassesSubgroups(SmallGroup(1344,11293)));
+83