On 04/04/2009, at 2:57 PM, john skaller wrote:
After some debugging we have that this nasty horrible production action:
| OPEN SYNTAX basic_name_comma_list SEMI
@{
....
print_endline "OPEN SYNTAX HAS FINISHED, RETURNING RESULT TO
DYPGEN";
(0,sr,Snull),!bindings @ [Add_rules add_rules; Local_data
local_data; Relation prios]
}
is returning without throwing an exn. I think (not sure!) that this
means the Not_found
must be happening inside Dypgen code.
If this is so it is a bug in Dypgen: even if the data is wrong, Dypgen
should never throw
a Not_found (but rather some more informative Dypgen specific
exception).
Because the grammar extension is completed (from Felix's view point)
the next
possible error in Felix code would be in the user action of the first
parsed
production. That is of course possible but seems unlikely.. I have no
idea which
production that would be though.. almost all (if not precisely all)
the user code
is written in Scheme, i'm not sure if Scheme can throw a Not_found as a
result of a problem in the users Scheme code (but I don't think so!)
Hence, a high chance the problem is in the Felix constructed data
being given
to Dypgen and Dypgen failing to handle the bad data nicely.
It isn't possible for me to trace this problem any further without
editing Dypgen.
I will continue with code inspection. Erick is trying to get stack
trace working to find
where the seat of the throw is ..
--
john skaller
skal...@users.sourceforge.net
------------------------------------------------------------------------------
_______________________________________________
Felix-language mailing list
Felix-language@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/felix-language
