Failures in slnconstr.gi: Error, ModRat: for <r>/<s> mod <n>, <s>/gcd(<r>,<s>) and <n> must be coprime

[fingolfin]

The following input file triggers an error:
bug_O_C2_NonIsometric_-1_6_9.txt

with 2   3 field=9  5 FFFKF
Error, ModRat: for <r>/<s> mod <n>, <s>/gcd(<r>,<s>) and <n> must be coprime in
  LogFFE( det, Z( q ) ) / gcd / (d / gcd) mod ((q - 1) / gcd)
  at GAPROOT/pkg/recog/gap/matrix/slconstr.gi:1250 called from
SLCR.SLSLPbb( ri!.data, x, ri!.dimension
 ) at GAPROOT/pkg/recog/gap/matrix/slconstr.gi:2472 called from
slpforelement( ri )( ri, x ) at GAPROOT/pkg/recog/gap/base/recognition.gi:711 called from
SLPforElement( rifac, gg ) at GAPROOT/pkg/recog/gap/base/recognition.gi:733 called from
slpforelement( ri )( ri, x ) at GAPROOT/pkg/recog/gap/base/recognition.gi:711 called from
SLPforElement( KernelRecogNode( ri ), z!.el
 ) at GAPROOT/pkg/recog/gap/base/kernel.gi:55 called from
...  at bug_O_C2_NonIsometric_-1_6_9.txt:95
type 'quit;' to quit to outer loop
brk>

Perhaps @danielrademacher has some insights on this code?

[fingolfin]

This was originally reported by @TristanPfersdorff on gap-packages/ClassicalMaximals#111

[danielrademacher]

On my system the recognition fails but no error occurs (but I have an old version of the Recog package).

On the current master branch of recog I also get the error. So it has to do with some commit after Oct 7, 2021.

[danielrademacher]

It's not as easy as I thought to find the commit from which the error occurs. I may have also changed functions locally, which means that the error no longer occurs. I'll take a closer look at this.

It looks as if the recognition no longer fails and the function is called as a result. It probably wasn't like that in the past.