Eana Eltu: Translator, Dictionary, API and putxìng.

'eylan na'viyä · July 21, 2010, 06:57:42 PM

fine, great work!

Just reporting that it looks like one translation is missing:
"aykeyä (key) Gesicht Bug. Please report that LEN hasn't been found. Plural Genitiv"
i guess it must be "leniert"

Tuiq · July 21, 2010, 11:08:50 PM

Uh, that's a problem. True. Because lenition isn't in the dictionary and that one is actually hardcoded. I suppose I have to add that manually.

Tuiq · July 23, 2010, 09:10:39 AM

The dictionary is finally useful - I fixed the bug where words had to be exact. As an example, searching for "kel" gives you all results now. Also, it can now show the IPA. The IPA in the SQL file is once again UTF8 only, no TeX in there anymore.

Muzer · July 24, 2010, 06:21:24 PM

Someone broke infixes. It's broken in the SQL file, so Tsim Apiak broke about 22 and a half minutes ago.

Tuiq · July 25, 2010, 05:06:25 AM

It broke when I changed the IPA behaviour. Fixed. SQL file recreated.

Tuiq · July 25, 2010, 06:10:11 PM

Eana is now capable of simple numbers, and in fact has lost its ability to parse almost every prefix/postfix. Will have to talk to Taronyu about that later.

Tuiq · July 31, 2010, 04:34:13 PM

The SQL generation is delayed until you hear something different from me.

Tuiq · August 05, 2010, 08:55:51 PM

OpenSource.

Sh4rK · August 06, 2010, 03:37:45 PM

W.O.W

Tuiq · August 14, 2010, 12:54:40 PM

The PDF creation is going to be replaced with a new system, allowing Taronyu to change the whole dictionary again without bothering me (not that I minded. Well I did sometimes *cough*).

These changes will also apply for translators; you have two new variables - PREFIRST and PREFIRST_FIND. Please fill these out as Taronyu did, but do not translate pre-first in PREFIRST_FIND. As you see right now in the Inflections-part, there is this ugly "in position 0" for non-English dictionaries (English had always been hardcoded). This is now possible for other languages as well. It depends on how you have translated it - but mainly, it will search for PREFIRST_FIND and replace it with PREFIRST. For example, in the German dictionary it's called "in Position pre-first" (because it's outdated it's possible that it is called 'in Position 0', as of this writing). So, PREFIRST_FIND is "in Position pre-first" and PREFIND would be "in der vor-ersten Position" or something like that - I'm not a linguist. Then, magical, "in Position pre-first" will be replaced with "in der vor-ersten Positon". This applies for all languages. If there are any problems, feel free to ask.

Also, the SQL generating will start again along with TSV/JMemorize, being updated daily. This update process will be open for all applications; this means that you can modify an existing module (they will be supplied on Github very soon) to create your very format. This could be CSV again (if you really want to mess around with that...), spellchecking things, binary databases. All modules have to be written in proper, clean Perl ("as clean as the other EE code"), they are not allowed to eat too much memory or time (with the database, there will a maximum of 10s per module) and they have to be approved by me. The good thing about this is that all these formats will be updated daily. Available in all current languages. All the time. Sounds good, eh?

Tuiq · August 14, 2010, 07:07:00 PM

Keep updated with these little eyecandies:

altering its color with the progress, it looks pretty awesome.

It's generateable for every language that is developed. The design will change soon I hope to look more appealing. Also, more sizes could be provided for different situations. But for now, this will do.

Edit: Added all languages.

Taronyu · August 14, 2010, 11:54:28 PM

Man, that's awesome. Where's the link for that?

Muzer · August 15, 2010, 06:33:04 AM

http://dvi.clonk2c.ch/navstat_LANGUAGECODE.png

(where LANGUAGECODE is the code for the language)

Tuiq · August 15, 2010, 06:04:43 PM

Added all images that were translated or are translated. Or even just created (I wonder what pl is doing - didn't they want to do their own "translate the dictionary" thing?)

Tuiq · August 16, 2010, 09:16:22 AM

Adding two example plugins for file output and added cronjobs as well as sample configuration file. Updated README.
Adding the magical picture generation and the used images.

Tuiq · August 17, 2010, 05:51:32 AM

Thanks to Muzer, the SQL generation works again. The format is the same.

Tuiq · September 01, 2010, 01:52:47 PM

Added RSS feeds (by muzer) and fixed various things.

Todo:
- Optimizing the SQL. Providing another version that makes use of the INSERT INTO () VALUES (), (), (); syntax and is zlib'd. The current database is shrinking to 137 KB (down from 1.2 MB), the performed could be a bit better (unzipped and zipped)
- Adding a languages table to the SQL, providing the native names for language codes and therefore allow applications to localize more.

Tuiq · September 03, 2010, 11:27:47 AM

50.
25.
10.
1.

RSS FEEDS.

Re-generated every day at 03:15 GMT+1 DST among the other stuff, the RSS feeds with the X latest words are generated. Made by Muzer.

Tuiq · September 10, 2010, 08:25:39 AM

Wake up son of mine, momma got something to tell you. Changes come. Life will have its way. With your pride, son. Take it like a man.

Dumped IDs in the SQL will be CHAR(40) instead of INT(11). They are SHA1 hashes of the internal-ID and the "mutation". IDs are now stable and won't change for a word.

Taronyu · September 10, 2010, 09:20:42 AM

I don't think I understand the actual meaning behind a single one of those sentences.