El dc 07 de 11 de 2012 a les 17:32 +0100, en/na Per Tunedal va escriure:
> Hi,
> thank you. I've read the Wiki and looked into the apertium-nn-nb.nb.dix
> file.
>
> Apparently, this is solved in a less transparent way in the nn-nb pair
> than in the examples in the Wiki.
It's less transparent because it is more complete. I think that
compounds work very similarly in sv, da, nn, nb so you could probably
just copy these paradigms and see how it goes.
> In the beginning of the dictionary,
> there are a lot of pardefs treating compounds, that I don't understand.
> Can anyone explain?
I can try.
> <pardef n="cp-both\Ø_LR_s\S__case" c="cp-L and cp-R. Analyse both -Ø-
> and -s- in compounds, generate -Ø- if cp-L">
> <e r="RL"><p><l></l> <r><s n="cmp"/></r></p></e>
This generates compounds without epenthesis.
> <e r="LR"><p><l></l> <r><s n="cmp"/><s
> n="compound-only-L"/></r></p></e>
This allows analysis of compounds without epenthetics. The
'compound-only-L' symbol is not output and is used to ensure that this
entry is only used when it forms part of the left side of a compound
boringhousebiscuit = boring+house+biscuit
Left Left Right
> <e r="LR"><p><l>s</l> <r><s n="cmp"/><s
> n="compound-only-L"/></r></p></e>
This allows the analyis of an epenthetic 's', again only when it is on
the left of a compound:
Left Left Right
infrastruktuurontwikelingsplan = infrastruktuur+ontwikel(s)+plan
epenthetic 's'---^
> <e> <p><l>-</l> <r><s n="cmp-split"/></r></p></e>
> <e r="LR"><p><l>s-</l> <r><s n="cmp-split"/></r></p></e>
This is for split compounds like
hargle- og barglewaffle
> <e r="LR"><p><l></l> <r><s n="compound-R"/></r></p></e>
This states that the unmarked "nominative" analysis should only be made
if this is the last part of a compound.
> <e> <p><l></l> <r></r></p></e>
This allows analysis and generation of unmarked "nominative" form.
> <e r="LR"><p><l>s</l> <r><s n="gen"/><s
> n="compound-R"/></r></p></e>
This states that a genitive analysis of 's' should only be made if the
word is the last part of the compound.
> <e> <p><l>s</l> <r><s n="gen"/></r></p></e>
This allows analysis and generation of -s genitive form.
> </pardef>
>
> The noun "kjempe" is advertised as possible to use in compounds, yet
> there is an entry for the adjective "kjempehøy" (= very high/tall). Why?
Although automatic compound decomposition is possible as a last resort,
it is still desirable to have entries. Automatic decompounding will
never be 100% accurate.
Fran
------------------------------------------------------------------------------
LogMeIn Central: Instant, anywhere, Remote PC access and management.
Stay in control, update software, and manage PCs from one command center
Diagnose problems and improve visibility into emerging IT issues
Automate, monitor and manage. Do more in less time with Central
http://p.sf.net/sfu/logmein12331_d2d
_______________________________________________
Apertium-stuff mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/apertium-stuff
