CONFOR: tokey not excluding characters

[GoogleCodeExporter]

When the "Melitid" dataset provided by Charles Oliver Coleman is exported using 
CONFOR (or via the editor) the included characters ("character mask") differ 
from the version exported from the legacy version of delta.

When the dataset exported using the Open-Delta editor is run with the 
Open-Delta version of KEY, KEY has the following string in its output for the 
tabular key:

Characters included 7 11-233 235-236 238 240 242 244

Whereas the dataset exported using the legacy DELTA, run with the Open-Delta 
version of KEY has the following equivalent string in its output:

Characters included 7 11-12 14-28 31-34 36-63 65-88 90-114 116-129 131-175 
177-180 182-185 187-201 203-205 207-208 210-218 220-233 235-236 238 244

It appears that the Open-Delta version of CONFOR is not excluding numeric 
characters without key states defined, or mulitstate characters with a large 
number of states.

Original issue reported on code.google.com by chris.fl...@gmail.com on 9 Jan 2013 at 1:30

[GoogleCodeExporter]

The original CONFOR excludes characters with > 20 states during KEY 
translation.  The new CONFOR now does the same.

Original comment by chris.go...@gmail.com on 10 Jan 2013 at 3:45

[GoogleCodeExporter]

Original comment by chris.go...@gmail.com on 10 Jan 2013 at 3:45

• Changed state: Fixed
• Added labels: Fixed-1.1