[Snowball-discuss] a small 16 bit glitch

Jean Millo jean.millo at abdmf.com
Mon Dec 22 17:14:09 GMT 2008


Hello,

I just ported the stemmers to UTF-16 and found a small glitch.

I am running under Centos-5 64bits and use gcc as a compiler.
It is not enough to change the definition of "symbol" (in api.h) to 
"unsigned short", you have also to change "sb_symbol" (in libstemmer.h) 
to "unsigned short".
With these changes and the -w option in snowball everything looks great.

Best regards,
Jean Millo



More information about the Snowball-discuss mailing list