On Jan 17, 2008, at 12:06 AM, Timo Paulssen wrote:
Yoav Nir schrieb:
coi rodo
jbovlaste has a distinction between "preferred only" valsi and
"all" valsi. There are, for example, 339 preferred works starting
with 'c' but 598 words in general that start with 'c'.
I have two questions:
1. What is the difference.
I think, that it's not "preferred words", but rather "preferred
definitions", as you can have multiple definitions in - for example
- english for a valsi, but using votes by the jbovlaste users, the
most fitting definition/translation is selected.
2. The XML export seems to export only the preferred words. How can
I get an export of all words?
If my assumption is right, then that wouldn't make much sense. At
least if you want to use the XML exported data for a dictionary or
lookup program.
I can only guess, but probably the code for XML export is based on
the code for the dictionary export, in which case it's the only
sensible thing to do to filter out non-preferred words.
- Timo
I thought so, but the "all" words don't contain duplicates. For
example, the word {ckupau} is in "all" words, with a good definition,
but it doesn't appear at all in the preferred words.
I think generating a dictionary without all these words makes it
limited.