Yeah, I was thinking of adding a key for that, but running into serious issues with hotkey overlaps with Anki (since Yomichan and Anki have to share them... for instance pressing Ctrl+S in Yomichan saves the Anki deck). Beh, I'll work something out

Decided to stop doing Core2k and do a Yomichan deck to pull vocab from sources I'm likely to read. Took a bit to set it up (I had the unicode issues with the title and finally had to just make a separate, easy-to-grab file with the document re-named in it) but I pulled 33 words in like 15 minutes.

Thank you for creating yomichan! It's awesome.

This is awesome, you are awesome, and the time this saves me doing manual input is awesome.

I know someone's already mentioned it, but Cloze deletion of the Expression from the Sentence would sure be a spiffy feature!


Seriously though, add that and I'll name my first born after you. Hell, I'll name someone else's first born after you.

awesome

Hey Foosoft, thank you for looking into my earlier suggestion. May I dare one more? How about the ability to mouse over and look up the word you are reviewing IN Anki itself and not just words from texts opened with Yomichan? If the user added cards without using Yomichan (say through the importing of a .csv) you've got a lot of cards with blank definition fields waiting to be filled. It would cut out the middle man of going to Jim Breen's website since Yomichan brings Edict to Anki. Single click copying is brilliant! I recently compiled a vocab deck from a word frequency list that was floating around here on the forums and was copying them one at a time into Jim Breen. I have to say, it's tedious work with the old highlighting only the bit that you want, copy, paste method! Maybe it could have a button that just copies the parts speech and the definition since Anki takes care of the reading? Or heck, a separate button for each part of the entry, the word, the part(s) of speech, and the definition! I don't know... I need to learn to code so I can quit being a pest.

@firion1234

Thought about using this?

http://forum.koohii.com/showthread.php?p...#pid126791

I have used it actually. It's a pretty neat plugin, but since it is the equivalent of Jim Breen's text glossing tool there are occasions when it doesn't give you the whole definition you would have been provided had you looked the word up by itself in isolation. I have reason to think that this behavior on the part of the text glossing tool has been fixed considering the example I was going to cite for this post didn't yield the same results as last time ^_^' but I guess I'll still exemplify my previous findings.

word search: 影響 【えいきょう】 (n) (1) influence; effect; (vs) (2) to influence; to affect; to have an influence on; to impact; to have an effect on; (P); ED

VS

text glossing: 影響 【えいきょう】 (n,vs) influence; effect; (P); EP

Plus, a lot of the time I don't need anything other than the definition and the part(s) of speech and have to copy just the bit that I want. Highlighting and backspacing the parts I don't want is just as time consuming as highlighting and copying the bits I do.

When I used the WWWJDIC glossing and overture's plugin, this is what I get:

(n) (1) influence; effect; (vs) (2) to influence; to affect; to have an influence on; to impact; to have an effect on; (P); ED

In other words, my gloss results aren't cut off?

I'm using the original plugin, so I don't think it was ever a problem?

It looks like Yomichan is more prone to cut off the definition? e.g. 影響 gives (n,vs) influence; effect - Are you saying this is desirable or isn't desirable? Sounded like you wanted the whole definition, but then you don't want it because you'd have to backspace? Either way it seems like you'll either be adding or subtracting at some point, definitions aren't always going to be divided how you want or of the range you want.

I also use Lingoes/Stardict in Anki for when I don't feel like using definitions on cards (or don't have them).

If those two aren't enough, you've also got wrightak's Rikaichan mod. Either way, it's easier if you're talking about importing lists rather than selective handfuls to batch add and then (should you neurotically wish to do so ;p), backspace any unnecessary definitions (though might as well leave them, it's not like you have to read and factor them in whenever you study the card.)

By the way, ruby support is in Anki (see: http://forum.koohii.com/showthread.php?p...#pid130811) presumably via WebKit (since 2009 or so), so I'm wondering if importing the QtWebKit module could be used then, in Yomichan, to display ruby? If so perhaps Yomichan could read JNovel Formatter (or at the least one could use the regex in that thread to create basic html files from txts, if JNF's files are too complicated beyond the ruby markup).

Rather than you having to write some custom renderer thingy as part of the vertical text support.

Although I guess there'd be difficulty combining that QWebview thingy with whatever you're using to parse and display definitions?

Edit: Another possible use at some point might be to convert ruby in the text to rubified Anki fields when adding new cards.

One thing: thank you so much!!!

And a little bug:
The font color in the definitions window is white, the background yellow. It can't be read easily.

I'm on Ubuntu 10.04, Anki 1.2.7, Python 2.7

saru_yo Wrote:One thing: thank you so much!!!

And a little bug:
The font color in the definitions window is white, the background yellow. It can't be read easily.

I'm on Ubuntu 10.04, Anki 1.2.7, Python 2.7

Weird...it's not like that for me. Did you mess with the settings or something?

@nestor:
Importing webkit would definitely be nifty, however since it as a library not included in the default distribution of Anki, it would be kind of a pain for people to use it. They would have to install the same version of python that Anki uses, then install the webkit bindings. And since most people unfortunately use Windows this is all quite messy. Which is unfortunate, because dropping in a web control and telling people to go wild would be neat. There is QWebView control that is available for use but it's kind of lame. Anyway, I'll check out how feasible this all is.

@saru_yo
Colors show up the way they do for you because you have some sort of weird color scheme set up:

Code:

palette = QtGui.QApplication.palette()
    toolTipBg = palette.color(QtGui.QPalette.ToolTipBase).name()
    toolTipFg = palette.color(QtGui.QPalette.ToolTipText).name()

    html = u"""
        <html><head><style>
        body {{ background-color: {0}; color: {1}; font-size: 11pt; }}
        span.expression {{ font-size: 15pt; }}
        </style></head><body>""".format(toolTipBg, toolTipFg)

As you see, the definitions window uses the system default colors for the tooltip. Why it's white on yellow on your setup, I don't know

There's a QtWebKit4 .pyd and .dll file in Anki's lib folder, and a smaller .pyd in that folder's parent folder, is that what you meant? /n00b (I'm actually assuming Anki's QtWebKit files are how Anki allows for ruby markup on cards.)

orly... hmm, that would make things a lot simpler

And not to overload you, but once you get to thinking about rendering vertical text, might want to look into how that aozora-to-kindle converter uses platex to render it (assuming python could render it): http://forum.koohii.com/showthread.php?p...#pid129087

Excellent plugin! Really amazing. Thanks a lot FooSoft!

One suggestion (possibly already made?): How difficult would it be to get the expression of interest put into bold? Sometimes one comes across a lengthy sentence and it would be great to be able to focus the attention of the deck on the particular word with a bold font. Otherwise what I'm doing is manually polishing each card -- still a relatively minor task, but just thought I'd ask. Apologies if it has already been discussed (too lazy to read the whole thread hehe).

It wouldn't be difficult at all to bold the source word, but I'm not sure that is general enough. Like what if someone wants it to be italic? Underline? Different color? You see where I'm coming from?

Maybe the answer is just to provide additional escape markers for sentences, so something like:

%sb - sentence before word
%sa - sentence after word

So then you could do something like %sb<b>%e</b>%sa perhaps.

Edit: this would also allow for doing cloze deletion

%sb______%sa

Sounds good to me!
お願いします　

FooSoft Wrote:...Maybe the answer is just to provide additional escape markers for sentences, so something like:

%sb - sentence before word
%sa - sentence after word

So then you could do something like %sb<b>%e</b>%sa perhaps.

Edit: this would also allow for doing cloze deletion ...a

Awesome! One thing I'm wondering though is about sentences with more than one instance of the clozed word. If you add the %sb %sa values, would it still show the word if it shows up again later/earlier in the sentence?

For example, this sentence from the Wall Street Journal - Japan (sorry for the content, it's the first example I could think of): "この25年間で...福島第1原子力発電所は国内で最もトラブルの多い原発だったことが..."

I've added elipses to shorten it a bit, but the word "原発" shows up many times in the full sentence, 3 or 4 times, I think.

I definitely agree that being able to provide markers (correct term?) for the values is the right way to go. I tried to do %s-%e (clozed sentence = sentence - expression) to get cloze deletion originally, but that didn't work. Sorry if I'm asking for something crazy here, I don't know python or programming at all, I'm just really excited to get this tuned so I can really tear into some e-books.

Dur. Looks like I should check more thoroughly before I post something that doesn't actually include two clear instances of the same word in the sentence.

"すぐに修理申込みを行うお客様は「インターネット修理予約システムで申込む」をクリックしてください。"

There. If I want to cloze both "修理", would that be possible using the %sb and %sa values?

Clozing multiple items in a sentence is probably not going to happen because there is no reliable way of doing it (especially if you are clozing a single character that can repeat multiple times in other words). Basically the only way it can work correctly 100% of the time is if it clozes only the term in the original position.

This thread has grown rather elaborate, so I feel kind of ashamed returning to suich a trivial question : "why does the Anki plug-in crash on my machine ?". Hope this is still the right place to ask for help, though…

The point is that I had installed the plug-in some time ago and it did work. Then, by reading this discussion, I realized that a lot of progress had been made, therefore I decided to upgrade. I first moved the two files yomichan.py and yomichan.pyc, as well as the directory yomichan_plugin, from the plugin directory to my desktop, then proceeded to reinstall the extension from the Anki menu.
After quitting Anki, I got the following error mesage at relaunch :
___________________________________________________________________
Traceback (most recent call last):
File "ankiqt/ui/main.pyc", line 2679, in loadPlugins
File "/Users/jmm/Library/Application Support/Anki/plugins/yomichan.py", line 1, in
File "/Users/jmm/Library/Application Support/Anki/plugins/yomichan_plugin/yomichan_plugin.py", line 22, in
File "/Users/jmm/Library/Application Support/Anki/plugins/yomichan_plugin/core/reader.py", line 26, in
File "/Users/jmm/Library/Application Support/Anki/plugins/yomichan_plugin/core/reader_util.py", line 101
field: replaceMarkupInField(value, markup) for field, value in fields.items()
^
SyntaxError: invalid syntax
___________________________________________________________________

Not surprisingly, the extra option to open Yomichan did not appear in the Tools menu
Trashing the newly installed files and directory and moving back the older ones to their original location fixed the problem.
Is this a compatibility issue? I am now using Anki 1.2.8 "for old machines", which is the most recent one available for PPC Macs.

Can anybody help? I feel frustrated that I cannot try the new functions of this great add-on.

@jmignot
It's a compat issue but it should be easy to fix: http://forum.koohii.com/showthread.php?p...#pid127807

@FooSoft: thanks for pointing out the previous post in this thread.
Sorry for missing it. As I wrote, the discussion is getting long, and its content pretty intimidating…

Now, the plug-in loads properly, but I have another issue when I try to open the Preferences :

______________________________________________________________________

Traceback (most recent call last):
File "/Users/jmm/Library/Application Support/Anki/plugins/yomichan_plugin/core/reader.py", line 172, in onActionPreferences
dialog = DialogPreferences(self, self.preferences, self.anki)
File "/Users/jmm/Library/Application Support/Anki/plugins/yomichan_plugin/core/preferences.py", line 25, in __init__
self.setupUi(self)
File "/Users/jmm/Library/Application Support/Anki/plugins/yomichan_plugin/core/preferences_ui.py", line 89, in setupUi
self.formLayout.setWidget(2, QtGui.QFormLayout.SpanningRole, self.checkSearchGroupByExp)
AttributeError: type object 'QFormLayout' has no attribute 'SpanningRole'

______________________________________________________________________

Is there also already a fix for this one?

Looks like you are using an old version of Qt. It's failing in code that's generated by the pyuic4 tool. Unfortunately I don't have any way to test your config, but your best bet would be to take url=https://code.google.com/p/yomichan/source/browse/dev/preferences.ui]this file[/url] and run it through pyuic4 and output the result over the preferences_ui.py in the plugin.