Hi Daniel.

> In this case, we simply call FSADumpTool's main() method from LT.

I know. And I don't think it's such a good idea -- this really should
be application-specific. If LT produces those dictionaries then it'd
be better off writing its own compound file which would contain both
the info and the dictionary. You could then read both from such a file
(by positioning properly and using Guava's LimitedInputaStream, for
example) by means of:

https://github.com/morfologik/morfologik-stemming/blob/master/morfologik-stemming/src/main/java/morfologik/stemming/Dictionary.java#L107

Voila, no confusion.

> "Warning: FSA automaton without metadata file."
>
> =>
>
> "Warning: FSA automaton without metadata *.info file. The *.info file is
> expected in the same directory as the *.dict file."

Yes, sure -- this is not a big problem, if you provide a pr I'll
gladly merge it. The bigger problem seems to be that those .info files
need to be manually created by users (who are very unlikely to
understand what those magic constant mean)?

If your dictionaries use constant .info flags you could even get rid
of these files entirely and use an explicit dictionary constructor:

https://github.com/morfologik/morfologik-stemming/blob/master/morfologik-stemming/src/main/java/morfologik/stemming/Dictionary.java#L65

Just a few cents. Like I said -- I'll gladly change the message, but I
don't think it'll help all that much :)

Dawid

------------------------------------------------------------------------------
_______________________________________________
Languagetool-devel mailing list
Languagetool-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/languagetool-devel

Reply via email to