-Robin
--
http://singinst.org/ : Our last, best hope for a fantastic future.
Lojban (http://www.lojban.org/): The language in which "this parrot
is dead" is "ti poi spitaki cu morsi", but "this sentence is false"
is "na nei". My personal page: http://www.digitalkingdom.org/rlp/
I wrote a dumb little app that would basicalpy just open the webbrowser pointing at vlasislu.lojban.org/<word>. But that doesn't really meet the criteria for "non-suck". I usually just end up bringing up vlasisku from my bookmarks.
Maybe I should look into revisiting this app idea.
-Robin
Tried that with the stardict dictionaries from
http://www.lojban.org/tiki/Dictionaries,+Glossers+and+parsers ; same
thing, I can search on Lojban words but not English, let alone on
the body of the definitions. Perhaps there's something wrong with
the dictionaries.
> It supports dsl dictionary format. Most lojban dictionaries have
> been converted to dsl. There can be some bugs when viewing them in
> Fora, though. But I'm not sure. DSL dictionaries can be downloaded
> from here<https://sites.google.com/site/slovlacukt/file-cabinet> .
I grabbed ralju.7z, unzipped it, and tried to get ForaDict to read
them; it said there were no valid dictionaries.
Oh, they are compressed inside the compression. -_-
Fixed, trying again...
I've added both "bangu" and "eng-jbo_jbovlaste_glosswords"; I still
get no results for both "nest" and "house". Added a bunch more, no
help.
Of course, it also can't find "zdani", so I'd say I'm using it (Fora
Dictionary) completely wrong in some fashion. -_-
Got that installed with the DSL dictionaries. That *will* return
results for "house" and "nest", but it doesn't seem to have the
actual definitions in it; all it retuns for "zdani" in the jbovlaste
dictionaries, for example, is "zdani". -_-
No-one wants to work on jbovlaste, including me, but since you've
been so helpful, if you tell me which sort of export you're doing
and give me a bit more detail (i.e. and example of expected and
actual output), I'll see what I can do.
[snip various efforts]
... Aaaaaaaaaaaaaaaaaaaaaaah.
Retried with encoding "UTF-8" instead of "Unicode". Seems to work
now.
Which sure makes me wonder what "Unicode" was. -_-
Further experimentation reveals that it only finds *exact* matches
to the *entire* index field, or at least if there's a way to get it
to search more broadly I can't find it.
Fortunately, the oblique keywords and synonym lists seem to cover
everything, but a jbovlaste-gloss-word-only list would also be nice
just in case.
I'm now cleaning up the dsls, because some of them are UTF-16 and
some of them are UTF-8, which is a problem for me and I want to make
this available to other android users. Also the .dz compression,
for the same reasons.
I can't figure out what the encoding of
eng-jbo_basic-english-wordlist.dsl is at all; do you know?
jbo-jbo_jbovlaste.dsl seems to just be corrupted.
eng-jbo_rupnu.dsl and jbo-eng_rupnu.dsl translate almost everything
to "ru'u".
Feel free to review
http://teddyb.org/~rlpowell/media/public/android_dicts/README.txt in
case I missed or misunderstood something.
I used the URL you gave me, as far as I know. What file exactly
should I be downloading?
You *really* should be using the XML export; it's far easier for
programs to process. Does it have these same bugs?
I have no idea if you are or not; can you point me to the latest
version, please?
You're looking at the wrong page; look at
http://jbovlaste.lojban.org/natlang/en/apricot?bg=1
That's a deliberate design of jbovlaste, and has been like that
since the very beginning. Since Lojban words cannot be polysemous,
it makes no sense for a Lojban word to have two distinct meanings,
so a 1-to-1 mapping is enforced for dictionary production purposes.
Having said that, it has honestly caused more trouble than it's
worth, and it is unlikely that jbovlaste's successor, should there
ever be one, will work similarily. Unfortunately, no-one's actively
working on such a thing that I know of.
It might be possible to make an XML export that includes anything
with a positive number of votes, but it wouldn't be easy.
Yes, but the noralujv definitions are nigh unusable as written; you
need like a degree in noralujv to make use of them.
Also, they *all* are in jbovlaste, they just haven't all been
cleaned up enough to be useful, and hence have no votes, and hence
don't show up in the exports; see
http://jbovlaste.lojban.org/needed/ if you'd like to help
> "eng-jbo_phrasebook" can also be imported into jbovlaste but as
> examples are not exported... there will be a separate dsl file.
>
> The same can be said about bangu, rupnu, gugde if we import those
> definitions at once into jbovlaste using some script.
>
> It would be wonderful to have new fields in jbovlaste like
> "mnemonics", "antonyms".
Well, the source is available if you'd like to work on it.
> As for "eng-jbo_basic-english-wordlist" English and Basic English
> are different languages so we can also import that list into
> jbovlaste if "Basic English" is added to the list of possible
> languages ( http://jbovlaste.lojban.org/languages.html). Is there
> any need in
> http://www.lojban.org/tiki/tiki-index.php?page=basic+english+word+list&no_bl=y
> page ? Everything can be done in jbovlaste.
The reason for the basic english word list was *not* to provide a
dictionary; it was to act as a guide for words that need to be
created, that is, simple and obvious meanings/concepts/objects for
which we don't hav ewords yet.
Yes, I'm aware; I was explaining the historical motivation. Note
that I didn't write it.
> Actually what I need is not existing xml export.
> All that I need is xml dump of jbovlaste.
> Some time ago I asked you about sql dump but such dump turned out to be
> cumbersome for automated conversion.
>
> So what I need is xml dump.
> Or at least modification of existing mechanism that in addition exports the
> following fields:
>
> - glosswords
> - examples
> - etymology
> - vote infromation
>
> Sorry to mention but I'm not an expert in database management. Can't help.
And I don't have the time, and no-one else wants to touch it.
Sorry.
The files themselves worked fine.
-Robin
On Sat, May 21, 2011 at 02:25:31AM -0700, najrut wrote:
> https://sites.google.com/site/slovlacukt/file-cabinet
>
> Please check whether those files can be opened by Android apps.
>
> --
> You received this message because you are subscribed to the Google Groups "lojban" group.
> To post to this group, send email to loj...@googlegroups.com.
> To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
>
--
I'm still not happy with it because it *only* searches the word
fields, and it has to be a complete match; I'd really like a global
text search. Other than that, though, it's fine, and the
dictionaries najrut has provided are sufficient, used in
combination, that it works pretty well.
-Robin
The "simplified gismu" dict, taken from the smart.fm lessons, is broken. It has ~1700 entries, which IIRC is only a third of the complete list.
to pu benji ti fo lo mi me la.android. fonxa toi
mu'o mi'e.aionys.
-Robin
I think it's more an error in the conversion than it is a lack of the data, considering what's missing.
That said, I personally would like to have the full set, as I consider that list to be very useful, especially the sumti places.
to pu benji ti fo lo mi me la.android. fonxa toi
mu'o mi'e.aionys.
In particular, everything is in
http://teddyb.org/~rlpowell/media/public/anki_flashcards/smartfm.tsv
, although the [sound:...] stuff would need to be trimmed for most
uses.
-Robin
-Robin