[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Inconsistency in vlatai



On Wed, Nov 28, 2001 at 05:06:29PM -0500, Pierre Abbat wrote:
> lebymlatu                 : UNMATCHED : lebymlatu
> lecymlatu                 : Bad lujvo (y hyphen not required) :  lecymlatu
> ledymlatu                 : UNMATCHED : ledymlatu
> lefymlatu                 : UNMATCHED : lefymlatu
> legymlatu                 : UNMATCHED : legymlatu
> lejymlatu                 : Bad lujvo (y hyphen not required) :  lejymlatu
> lekymlatu                 : UNMATCHED : lekymlatu
> lelymlatu                 : UNMATCHED : lelymlatu
> lemymlatu                 : lujvo :  lemymlatu
> lenymlatu                 : UNMATCHED : lenymlatu
> lepymlatu                 : UNMATCHED : lepymlatu
> lerymlatu                 : UNMATCHED : lerymlatu
> lesymlatu                 : Bad lujvo (y hyphen not required) :  lesymlatu
> letymlatu                 : UNMATCHED : letymlatu
> levymlatu                 : UNMATCHED : levymlatu
> lexymlatu                 : UNMATCHED : lexymlatu
> lezymlatu                 : Bad lujvo (y hyphen not required) :  lezymlatu
> 
> All of these become well-formed lujvo when the 'y' is removed except 
> "lemymlatu", so why are some called bad lujvo and others unmatched?
> 
> phma
> 

It's just down to inconsistent displays by vlatai.  UNMATCHED is shown
when either the state machine jams before the end of the word (i.e. the
letters up to that point cannot be the start of *any* valid word form),
or if the state at the end of the word has no other word type associated
with it.  The "bad lujvo" case is actually detected as a separate case -
I could just as well have made that be shown as UNMATCHED.  I just
thought it more useful to give the user a hint as to why the word is
bad.  In the same way, attempted fu'ivla that fail the slinku'i test are
shown as such, rather than just saying UNMATCHED for those.

In your examples, note that cm, jm, sm, zm are valid initial consonant
pairs, whereas the others aren't.  If you run vlatai with the -v option,
you'll see the scan runs to the end of the word before jamming.  This is
quite common, because the next letter might be a consonant which would
make the word a valid cmene.

HTH

-- 
Richard.
---------------------------------+-------------------------------------
Richard P. Curnow                | C++: n., An octopus made by
Weston-super-Mare, UK            |  nailing extra legs on a cat.
http://www.rrbcurnow.freeuk.com/ |