El dv 01 de 02 de 2013 a les 14:15 +0100, en/na Per Tunedal va escriure: > Hi, > why not simply trim the dictionaries on compilation? The complier could > just skip all entries that are not present in the bidix, couldn't it? > > What's important is that the binary dictionaries, i.e. the transducers, > are consistent, isn't it? > > The XML-dictionaries could be untrimmed, that is larger for one language > than the other or containing more than the translated words on both > sides. Words could be added in one dictionary at one time and in the > other some other time. When new translations are introduced in the bidix > the words might already be present in one or both monodixies. > > One advantage is that the development could more easily be distributed > among several contributors. They only have to know ONE of the languages > in the pair to contribute! And further, it's far easier to add a > translation in the bidix than choosing the right paradigm for a foreign > word. > > Wouldn't such a procedure make the development of a language pair much > simpler?
http://wiki.apertium.org/wiki/Automatically_trimming_a_monodix Fran ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_jan _______________________________________________ Apertium-stuff mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/apertium-stuff
