A tricky grammar

Hello,

[ GitHub seems to be borked at the moment, so I’m sending this to the
  list instead. I’ll make an issue and test in the test suite for it
  when (a) GitHub is back and (b) I have a spare moment. ]

This grammar (derived from one of the King's College grammars that
Michael pointed us to; thank you again, Michael):

S: A .
A: 'a', B ; 'x' .
B: 'b', A ; LDOE, A .
LDOE: M; 'l' .
M: 'm'; LDOE .

when presented with this input

amalx

Crashes both my parser and Steven's (sorry Steven!). Mine goes into an
infinite loop trying to work out the ambiguity of the forest. I imagine
something similar happens to Steven's as it never returns a result.

I'm not quite sure how this grammar differs from any of the other
ambiguities in the test suite, but it clearly does.

                                        Be seeing you,
                                          norm

--
Norm Tovey-Walsh
Saxonica