Hello,
Thank you for answering.
Németh László a écrit :
Hunspell with the 90 thousand words of the recent French dictionary is not
too slow for a single suggestion or spell checking a long document without
suggestions. For other tasks (automatic spell checking of long texts with
suggestions), remove or limit MAP definition,
In the previous dictionary, there were :
~ 700 affix rules
~ 90,000 entries
This generated 650,000 flexions.
Hunspell gives suggestions very quickly.
In the new dictionary, there is :
~ 15,000 affix rules
~ 60,000 entries
It generates over 2,000,000 flexions
Hunspell gives suggestions sometimes slowly, sometimes quickly enough.
and use
MAXNGRAMSUGS 0
in the affix file to disable n-gram suggestions.
Desactivating the n-gram suggestion does not improve significantly the
suggestion speed. I see no differences.
L'esperance (mistake) -- L'espérance
connaissoit (mistake) -- connaissait
Hunspell is slow on these two examples (one letter difference only).
However, it can be quick on others mistakes.
Removing MAP does not improve significantly the quickness either.
I'm afraid the affix file I created is too much complex.
Fortunately, I can reduce complexity of the affix rules by removing
flags which are not useful at the moment. That's better but still not
enough.
What's the impact of having 6 flags with 491 rules each, each rule
calling instensively other flags usually for prefixing add-ons?
Almost all the flags call other flags for specific prefixation.
I also use a lot the flag NEEDAFFIX.
> Other option is to use affixes to compress a large dictionary
> (~200-300 thousand words). There is a new tool in the Hunspell
> distribution for automatic affix compression, "affixcompress".
I'm going to try that. But I have to reinstall Linux first. :/
I'll give some news about it later.
Best regards,
Olivier
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
