Closed AMR-KELEG closed 3 years ago
Wow, bad bug. Could you upload a zip with the failing bre.automorf.bin? This is from apertium-bre revision https://github.com/apertium/apertium-bre/commit/1fd2fc710075e51889ddba3b974cfbd9abc91a51 ?
Wow, bad bug. Could you upload a zip with the failing bre.automorf.bin?
Sure, here it is. bre.automorf.zip
This is from apertium-bre revision apertium/apertium-bre@1fd2fc7 ?
Yes, it is.
Similar issue in the other direction:
$ cat cat.att
0 1 c c
1 2 a a
2 3 t t
3 4 @0@ +
4 5 @0@ n 0.004
5
4 5 @0@ v 0.0045
--
0 1 c c
1 2 a a
2 3 t t
3 4 @0@ +
4 5 @0@ n
5 6 @0@ +
6 7 s <pl>
7
$ lt-comp lr cat.att cat.bin
Warning: Multiple fsts in 'cat.att' will be disjuncted.
main@standard 15 15
final@inconditional 3 2
$ lt-print cat.bin
Error: empty set of final states
For the breton compiled dictionary
bre.automorf.bin
This is caused by the way Attcompiler deduces the type of the edge "https://github.com/apertium/lttoolbox/blob/master/lttoolbox/att_compiler.cc#L381"
I tried setting the type of all edges to word such that they are part of the main section. The lt-comp worked but the lt-proc is now entering an infinite loop on initializing the root state (Finding Epsilon closure).