[antlr-interest] Token lookup crashing program.

[doctlo-antlr at yahoo.com]

A couple of things I've observed.
The input is line oriented ( which should be apparent from the
definition of EOL and BEOL ).

After some basic data, a set of lines is sent which is parsed 
exactly the same. Then another set is parsed in the same way.

basic_dt1
basic_dt2
basic_dt3
...
<type1>
type1_1
type1_2
type1_3
...
</type1>
<type2>
type2_1
type2_2
type2_3
...
</type2>
final fatal

Three observations about the error:
1) It happens while parsing type2 data, about 3/4 of the way from the 
     end. When I remove the last line correctly parsed it still fails,
     but when I remove six ( but not five ) lines, that clears things 
     up.
2) If I delete all but one piece of type1 data, the program runs OK.
3) If I delete all the type2 data  after the last correct 
     line parsed, the program runs correctly. But if I keep the next 
     line of type two data, the program fails again.


--
Thad