Hi, I'm using aspell 0.61 with the latest arabic dictionary. I can get it to run in pipe mode, for example:
.../tools/aspell-0.61-20071105: echo "فأحضرالزوجة" | ./aspell -a -d ar --dict-dir=./dicts/aspell6-ar-1.2-0 @(#) International Ispell Version 3.1.20 (but really Aspell 0.61-20071105) & فأحضرالزوجة 4 0: فأحضر الزوجة, فأحضر-الزوجة, فأحضرا لزوجة, فأحضرا-لزوجة However, for some words it returns a result that does not seem valid, according to the latest version of the manual (http://aspell.net/0.61/man-html/Through-A-Pipe.html#Through-A-Pipe): .../tools/aspell-0.61-20071105: echo "برسائلنا" | ./aspell -a -d ar --dict-dir=./dicts/aspell6-ar-1.2-0 @(#) International Ispell Version 3.1.20 (but really Aspell 0.61-20071105) + رسائل Note that this result begins with a plus sign + instead of & * or #. I believe this means that "rsA'l" is the root word of "brsAilnA" (since b+ and +nA are affixes), but this does not seem to be part of the usual aspell/ispell pipe api. Can someone point me to some documentation or clarify this functionality? Second question.... I can't get example-c to work with the arabic dictionary. Does anyone have example usage of example-c with a different character set and language (not english, non-latin)? I don't really understand why I need to specify a size and jargon before I can specify a character set (what do I set jargon to?). In any case, my examples are all utf8, which does not seem to be a valid character set for aspell. ./examples/example-c ar Error: The file "[...]/tools/aspell-0.61-20071105/lib/aspell-0.61/l-ar.cset" can not be opened for reading. ./examples/example-c ar huge blah utf8 Aborted Thanks, KP _______________________________________________ Aspell-user mailing list Aspell-user@gnu.org http://lists.gnu.org/mailman/listinfo/aspell-user