I have written a simple tool that decodes Japanese text to Heisig RTK keywords. Basically serves the same function that the reading tool on RevTK does. Information on it can be found here along with the actual script.
http://kd7yhr.blogspot.com/2009/07/new-heisig-tool.html


EDIT: This is now available as a tool within Anki, simply download RTKize from the plugin download section.

Please, Please make this into a plug in for anki

zazen666 Wrote:Please, Please make this into a plug in for anki

I used a small perl script to lookup kanji used in my sentence cards and add the keywords (and components) to another field in the deck, to be displayed with the answer -- saves having to manually look anything up later, and no plugin required.

zazen666 Wrote:Please, Please make this into a plug in for anki

Ok, I did a quick prototype and actually worked quite well. Need to speak with Damien though about inclusion into the Japanese bundle for Anki. See the link below for a screen shot.

http://kd7yhr.blogspot.com/2009/08/hacki...words.html

Thanks for pushing for this. Brian

Edit: Already emailed Damien, will update once I figure out what to do with regards to plugin.

Why must it be included in the Japanese plugin? You can just ship a supplemental plugin.

It doesn't make much sense to include it in the Japanese bundle. That bundle is for everyone studying Japanese in Anki, not for everyone using Remembering the Kanji. It should definitely be separate.

brianobush Wrote:

zazen666 Wrote:Please, Please make this into a plug in for anki

Ok, I did a quick prototype and actually worked quite well. Need to speak with Damien though about inclusion into the Japanese bundle for Anki. See the link below for a screen shot.

http://kd7yhr.blogspot.com/2009/08/hacki...words.html

Thanks for pushing for this. Brian

Edit: Already emailed Damien, will update once I figure out what to do with regards to plugin.

YES!!! I have been requesting this FOREVER. More than two years ago when I migrated to Anki from this site.

I can't wait for the plug-in. Thanks so much!

Great stuff!
Perfection (in my eyes) would be to have the information that is displayed when hovering the mouse over the kanji be read out directly from a specified field of the (most likely present) Heisig-model.
Like this one could also include the story and instant-edit stuff.
Really great stuff. And I believe that this has the potential to even grow greater with additional options like direct links to stroke order animation etc...
Brianobush is the hero of the day!

Jarvik7 Wrote:Why must it be included in the Japanese plugin? You can just ship a supplemental plugin.

Yes, I thought about it a bit and made sense that I do my own plug-in - It was late
I still have some unanswered questions, but are minor and mainly concern presentation (css).

Nice, why not just release the python script here so we can just drop it in the plug-ins folder for a manual installation?

RobotsAnger Wrote:Nice, why not just release the python script here so we can just drop it in the plug-ins folder for a manual installation?

Why not just download if from within Anki? I have added it as a plugin and you will find it named RTKize.

Hi there, got a little trouble to get it running. Used a test deck with sentences, added a tag "Japanese".
Anki gives the following error message:

Traceback (most recent call last):
File "ankiqt\ui\main.pyc", line 2401, in loadPlugins
File "C:\Dokumente und Einstellungen\Bj�Anwendungsdaten\.anki\plugins\RTKize.py", line 2068, in
UnicodeDecodeError: 'utf8' codec can't decode bytes in position 0-1: unexpected end of data

HerrPetersen Wrote:UnicodeDecodeError: 'utf8' codec can't decode bytes in position 0-1: unexpected end of data

I updated it to ignore decode errors, however there should be as the Kanji to keyword data is utf-8. I am running on Ubuntu and works fine - of course the author always says this.

That fixed it

Another little problem: For keywords made up of multiple single words (example: big dipper 斗) the tool only displays "big".

HerrPetersen Wrote:Another little problem: For keywords made up of multiple single words (example: big dipper 斗) the tool only displays "big".

Fixed, simple HTML generation bug.

It not working smoothly on my Mac. Just reviewed 50 cards, and holding the cursor over the kanji only showed the keyword about 5% of the time. (the other 95% nothing). (insert sad face here)

zazen666 Wrote:It not working smoothly on my Mac. Just reviewed 50 cards, and holding the cursor over the kanji only showed the keyword about 5% of the time. (the other 95% nothing). (insert sad face here)

It is probably the rendering engine and the CSS employed. I don't have a mac nor am I a CSS expert - thus you are on your own, however, if you have any patches I would be willing to take them, test them and redistribute fixes. On my system they are working great and have been a great help - not to gloat.

Since I have zero knowledge of CSS, patches, Fixes, etc, I will quietly weep and once again wait for someone to come along and fix this.

Thanks for the effort. I am sure it will be a big help for many. Hopefull someone with a mac will come along and fix it.

It works 100% of the time for me on OSX.

Perhaps you have a conflicting plugin.

Hmm. Updated Anki. No other plugs in. Still not working....

Did you check, if the field, where you want the plugin to work is labeled "expression". Maybe for your vocab-model that is the case, but not for your sentence model? That could explain the 5% wokring rate.

Is there an easy way to allow symbols such as á à ö ü etc for the keywords? I use some German keywords and also some Chinese keywords with pinyin (example: tiáo)

HerrPetersen Wrote:Did you check, if the field, where you want the plugin to work is labeled "expression". Maybe for your vocab-model that is the case, but not for your sentence model? That could explain the 5% wokring rate.

Hmm. so if I am using a different model (which I am) then the plug it wont pick it up?

I don't claim to understand how the plugin works but I think you have to make sure that
1.) The cards in question are tagged with "Japanese"
2.) The model(s) to which the cards belong must have an "expression" field.