[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Re: [edict-jmdict] EDICT and JMdict not updating since july 22



Hi Kim,

I have to confess that I haven't updated the examples database
used on jisho.org since april. But the format is fast enough
to do a daily import of, so I'll add that in the next update.

Well a daily import is probably a bit optimistic but a weekly
one would be sensible.  Do you do a check of the date stamp
to see if it's changed or not before downloading?

I'll also add you to the credits on the about-page.

I wouldn't bother, although I'm not going to stop you.

Sure. How would you prefer to have that implemented? A link
to the suggestion form in WWWJDIC (i presume that the ID
numbers each sentence has are the sequence they appear in the
file) or something custom?

There's nothing special about the WWWJDIC form - and I don't
know how Jim handles ID's (whether they are preserved or not).
As long as it has a) The sentence(s) to be commented on and
b) Space to type corrections / comments then it's fine.

Come to think of it I should probably link to the word
correction page in WWWJDIC as well.

>  2. I assume you are using some sort of auto-parsing to get
>  the word links from the examples. May I suggest that you
>  would be better off using the keywords line?

I wrote the importer script over a year ago so I had to go
back and check, and I actually use the keywords line. But I'm
not taking the {} field into account ... but that was a quick
fix. So the next time I update the server (about once a month)
it should link everything.

Incidentally the keywords in the B line are in the order they
actually appear in the example.  Somebody really enthusiastic
could use that to prevent the wrong characters being
highlighted when there are ambiguous matches.

For example in the following the は is after さん so it
shouldn't get mixed up with the first half of はい.

「はい、ありません」とジョーダンさんは答えた。	
"No, I don't," said Mr Jordan.	
はい 有る{ありません} と[2] さん は 答える{答えた}

Best,

Paul