#952 substructure searching bug with phenylhydrazine

[Jeff Janes]

I have two heterocyclic molecules which have no direct nitrogen-nitrogen bonds.

Both of them are being reported as superstructure matches against phenylhydrazine.

Both are matched when phenylhydrazine is spelled 'c1ccccc1NN'.

The second is found even when when it is spelled in canonical form, 'NNc1ccccc1'

obabel bad.smi -ocan -squery.smi

c1ccc2c(c1)scn2 bug1
O=C(c1ccc2c(c1)ncs2)Nc1cccs1 bug2
2 molecules converted

This is under both 2.3.2 and 2.3.3

Thanks,

Jeff