ge schreef:
On Friday 26 May 2006 04:01 pm, Simon Brouwer wrote:
I think that is more a question of implementation than a distinction
that does justice to the purpose of spell checkers vs. grammar checkers.
For example, in Dutch it is a pain that the OOo spell checker is not
able to check word groups such as "ad hoc", "'s avonds", "1 aprilgrap"
etc. whose parts are only valid in one combination or very few
combinations. This could be solved using a grammar checker but it could
also be done with a spell checker that is able to look at the previous
or following word.
Knowing the complexity of the spell checker I vote, that it should do spell
checking really perfectly and fast on a single word base, using two folded
affix handling, which is by no means trivial.
I also found, working with Daniel Naber's language tool, that lots of grammar rules are simply handling of easily misspellable word combinations, like "of cause" and "of course", that are clearly best done in a simple regular expression style grammar checker (without word list handling).
Your cases seem to me easily to be handleable with a simple regular expression checker, word list free.
What do you think?
If that is the easiest way to implement this, sure. It would be great if
the relevant data for the grammar checker would be easy to generate from
a list of such word groups.
--
Vriendelijke groet,
Simon Brouwer.
| nl.openoffice.org | www.opentaal.org |
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
