with "LANG de_DE" being set words beginning with "-" are falsely accepted when "-" is defined as WORDCHAR. For example:
SET ISO8859-1
TRY esianrtolcdugmphbyfvkwäüößáéêàâñESIANRTOLCDUGMPHBYFVKWÄÜÖ-
WORDCHARS -
LANG de_DE
1
test
results with hunspell 1.1.5 in:
> echo -Test -test | hunspell115 -l -d testumlaut
>
while with hunspell 1.1.4 is was (which looks more correct):
> echo -Test -test | hunspell114 -l -d testumlaut
-Test
-test
>
when the "LANG de_DE" definition is removed also hunspell 1.1.5 says the "leading dash words" are incorrect.
Logged In: YES
user_id=726595
Originator: NO
I will remove special de_DE code for dash handling. Note: BREAK feature also permit leading dashes. Laci
Logged In: YES
user_id=726595
Originator: NO
Fixed. Using WORDCHARSBEGIN, -MIDDLE, -END will be better solution. Thanks, Laci