El dl 20 de 02 de 2012 a les 14:01 +0000, en/na Jimmy O'Regan va
escriure:
> On 20 February 2012 12:37, Francis Tyers <[email protected]> wrote:
> > Hey all!
> >
> > I just found some strange behaviour in the built-in lexical transfer in
> > apertium-transfer:
> >
> > Output from lt-proc -b | apertium-transfer -b:
> >
> > $ echo "direbech" | apertium -d . br-fr-biltrans | apertium-transfer -b
> > apertium-br-fr.br-fr.t1x br-fr.t1x.bin
> > ^Adj<SA><mf><sp>{^sans reproche<adj><mf><sp>$}$^punt<sent>{^.<sent>$}$
> >
> > Output from apertium-transfer:
> >
> > $ echo "direbech" | apertium -d . br-fr-tagger | apertium-transfer
> > apertium-br-fr.br-fr.t1x br-fr.t1x.bin br-fr.autobil.bin
> > ^Adj<SA><mf><sp>{^sans
> > reproche<adj><mf><3><sp>$}$^punt<sent>{^.<sent>$}$
> >
> > Bilingual dictionary entry:
> >
> > $ cat apertium-br-fr.br-fr.dix | grep 'sans.*reproche'
> > <e><p><l>direbech<s n="adj"/><s n="mf"/><s
> > n="sp"/></l><r>sans<b/>reproche<s n="adj"/><s n="mf"/><s
> > n="sp"/></r></p></e>
> >
> > Expansion of dictionary:
> >
> > $ lt-expand apertium-br-fr.br-fr.dix | grep 'sans.*repro'
> > direbech<adj><mf><sp>:sans reproche<adj><mf><sp>
> >
> > Anyone else come across this behaviour before ?
>
> The only thing that stands out is that the comparative macro sets sg
> when it sees sp, but that's to a variable that isn't used.
Yeah, that's weird, and if I comment it out I get the same results ...
> Otherwise,
> the macros and rules are the same as fr-es - do you get the same thing
> with 'cannois' in fr-es? (I don't, but it's the only example I could
> find of something similar).
$ echo "cannois" | apertium -d . fr-es-tagger | apertium-transfer
apertium-fr-es.fr-es.t1x fr-es.t1x.bin fr-es.autobil.bin
^adj<SA><mf>{^de cannes<adj><mf><sp>$}$^punt<sent>{^.<sent>$}$
$ echo "cannois" | apertium -d . fr-es-tagger | lt-proc -b
fr-es.autobil.bin | apertium-transfer -b apertium-fr-es.fr-es.t1x
fr-es.t1x.bin
^adj<SA><mf>{^de cannes<adj><mf><sp>$}$^punt<sent>{^.<sent>$}$
Same results for both modes.
> You could just check for sp and let nbr_sense_sp be <lit v=""/>, which
> should fix it, though without any illumination of _why_ it's
> happening. If you want to take bets on wild guesses, I'll put my money
> on weird voodoo with the comparative macro (and the butler in the
> library with the candlestick).
The weird thing is the difference in output between outside lexical
transfer and inside lexical transfer ... Looking at the code I would
have expected them to give the same results.
F.
------------------------------------------------------------------------------
Try before you buy = See our experts in action!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-dev2
_______________________________________________
Apertium-stuff mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/apertium-stuff
