Alerts are bad. I keep getting them from this plugin when I don't have ankiconnect started, even though I have it selected in my settings. Any page I go to now throws alerts.

I think the ordering for words and names with the prepackaged dictionaries is not ideal. Take the word 男 for example. If you have both dictionaries installed and turned on, it shows いのまた (a name) before おとこ (a common word). It could potentially be fixed by separating results from different dictionaries into separate sections/tabs. Or something more intelligent to intermix them, such as at least showing common words first.

Hi, in JMDICT the priority tags are attached to kanji and reading elements separately "because on occasions a priority is only associated with a particular kanji/reading pair". It seems Yomichan is showing a priority tag for a kanji/reading pair if either the kanji or the reading has that tag. It would be more accurate to JMDICT's intention to show the tag if both the kanji and the reading have it.

My main use case here is looking up a priority kanji word which has one priority reading and at least one non-priority reading, and trying to decide which reading is more likely. (At the moment they all appear as priority.)

Using 0.997 on Fedora.

Right now, yomichan has trouble detecting some words with furigana in ruby tags. When there is no okurigana, this seems to works fine. For example, 東京とうきょう can be detected. However, when there is okurigana, it does not work. For example, 起おきます cannot be detected.

EDIT: Apparently this happens whenever the entire word isn't together, not just with okurigana. For example, trying to put the furigana above specific kanji such as 漢かん字じ causes it not to be detectable by yomichan.

Just say hello and remind that Ankiweb interface was changed.
It does not allow cross domain XHR, changed interface from "GET" to "POST" .

(Hey Foosoft, just posting a few requests) The individual definitions (glosses) from JMDict shown aren't grouped by senses. In JMDict, they are, because each group of glosses is a cluster of synonyms. How about displaying them grouped the same way, to simplify learning? I'd love for this to be output as {glossary-list} in AnkiConnect. Example of how grouped glosses might look:

練る； 煉る

1. [v5r,vt]  to knead; to thicken into a paste (stirring over a flame)
2. [v5r,vt]  to polish (a plan, etc.); to refine; to elaborate; to work out
3. [v5r,vt]  to train; to drill; to exercise
4. (練る only) [v5r,vt] to gloss (silk); to soften; to degum

If I understand correctly, Alt + End should go to the last entry in the popup. Instead it opens the Yomichan options in the chrome://extensions tab.

I just saw #7 which seems to be something related to the firefox support. Is there any plan for firefox? It will be great to have this on firefox.

Currently you are displaying furigana across the entire word like 漢字かんじ.
If possible, it would be better to display them like 漢かん字じ, where it's clear which furigana belong with each kanji. Related to that, okurigana probably shouldn't get furigana. This is very obvious with the work 必ずしも.
The popover currently shows 必ずしもかならずしも.
Instead, it should probably show 必かならずしも.

I know this may not be possible depending on what the data you get looks like, but I figured I'd put in the request it case it was possible.

I just encountered this today. I don't know if the audio source is simply wrong if it this has to do with how the audio is being retrieved.
Take the following word: 留守番電話
For 留守番電話, the audio for 留守録 plays.
This also gives the result 留守番, which actually plays the audio expected for 留守番電話.
And when searching for 留守録 (the wrong audio from the first one), it plays とめどない.

It now seems that the online Sanseido dictionary is not going down.

I am wondering if it would be possible to include Sanseido definitions in the same way Rikaisama does.

This is an important issue to me and perhaps to other readers. For convenience I am pasting here a post I made on the Kanji Koohii forums explaining why the issue matters.

With Rikaisama/Sanseido back in action, at least for the present, I thought It might be helpful to FooSoftさん to tell him whether we would be switching back to RIkaisama and if so why.

In my case I will and for one simple reason. Sanseido. I like the brevity of the definitions. With Yomichan currently I get a big definition with a lot of dictionary-esque text plus bits of English I don't want (even using the Japanese EPwing dictionary exclusively). Of course I can edit this but it all adds time and complication to the card-making process. And I still miss Sanseido's concise, boiled-down definitions.

As it happens I just explained in this thread why I need very concise definitions.

The problem is the dictionary, not the plugin, so if it were possible for Yomichan to have the same arrangement with Sanseido that Rikai has, it would be solved.

In addition to this of course, the means of obtaining an EPwing Japanese dictionary are complex, dubious (obviously I can't recommend it on KawaJapa for example), and may not last. It also doesn't work on 32-bit systems.

I don't know if it is possible, but I will open an issue on GitHub because to me it is not quite make-or-break but certainly enough to make me return to Rikai.

I tend to use an IME in the search, even though it converts romaji to hiragana. This generally works, but not when the word has a double consonant. For example, when I try to type 学校, when I get to 「がｋｋ」 in the IME, it is converted to 「がっk」from the romaji replacement. Because of this, the IME can't detect the rest of the word as being associated with the first part, and it can't convert the word to 学校.

Hi @FooSoft :
Just happened to know such wonderful chrome extension for foreign language studying.

User Experience:

I am English learner (as Chinese native speaker). I followed IR (incremental reading) practice to write down new English word with sentence as context while visiting news webpage like BBC, CNN. usually I will manually input these information into Anki for reviewing later. So, you can not imagine the stunning feeling 👍 I had when I first time knew this tool. It saved me lots of time to copy and paste, and just focus on what they called another IR(immersion reading). after I finished reading, all cards were there already in Anki!

Personal Practice:

so, I want to use and convert it in my Learning English process. By ignoring extra Japanese learning module, like Kanji-dict, Rule, in your code. the edict part, including [express,reading,tags, glossary], can be perfectly mapped to [word, pronunciation, tags(part of speech) and definitions] and meet my requirement.

My Problem:

the only problem is that Japanese and Chinese is character based language, but English is word based. so, when I tried to mouse over select English word, it only workable when I put mouse point exactly above the first letter of that word. by the way, the "Scan length" in option is also calculated by Japanese character/English letter.

I tried to fork this project and made some changes on client.js and source-range.js, it's kinds of workable by detecting white space between words , but it not sounds like more general solution with well behavior coding skill (I never wrote JS before). so, just wondering if there would have more general option as listed below,

Feature Requested:

1. add one option item in back-end option page to indicate character or word based language.
2. based on above item 1, can mouse over select one by one character or word by word. of course, the scan length also could be character or word based.

Other Suggestions (or imagination):

1. [express, reading, tag, glossy and sentence] with Anki connect is core and general function for all language studying, so, it could be as basic package for all language learner. I have used readlang and LWT (Leaning with Text), all those tools need extra effort to import text and integrate with anki. Yomichan is the best without any extra effort. just hope it can be used by more peoples. as for other module, like Kanji, Rule, Katakana, Hiragana(and wonderful! de-inflection function), maybe it can be de-coupled from core function as addons.
2. when build my English-Chinese dict by following your edict format, noticed you already dumped these dict data in sqlite format in compile.py. As you may know json data is very memory consumed in Chrome(the extension itself sized up to 45M also, and Chrome memory allocation can be over 1G) so, by leveraging anki connect as a bridge to access local sqlite dict db, it may save both extension size and Chrome memory. Also, it also could be configurable in the future to change dict(not embedded in extension itself) by following your standard dict format.
3. just like Anki addon "awesome TTS", don't know if it's possible that audio link can be pre-configured as service provider and select-able. understood it's not just as simple as usual query link, like http://www.online-dict/?q={word}, but if it could be extended by coding level to provide more online audio service provider in progress, it's also highly appreciated.

jisho.org includes information about kanji frequency in newspapers. With 本 for example:
"10 of 2500 most used kanji in newspapers"
http://jisho.org/search/%E6%9C%AC%20%23kanji

Some info in advance: I'm using Firefox 53.0 with Yomichan 1.1.14 under Windows 10

Basically, I went through this process a few times:

1. I was getting annoyed at yomichan popping up on English text,
2. I'd go into the options and uncheck "Search alphanumeric text"
3. yomichan would stop popping up AT ALL,
4. I'd go into the options again to reenable the option,
5. but yomichan still wouldn't show up at all
6. I'd live with that for the time being
   --- later on
7. I'd restart firefox and yomichan would work again

This process repeated itself a few times as the "Search alphanumeric text" option was always checked when I restarted... Just recently I figured Yomichan works as I want it to if I have the option unchecked when I restart. Before making this issue I tried to reproduce the problem a few times and... I couldn't do it consistently. I was able to reproduce it once deliberately and then I couldn't anymore. I'm unsure if there are other reasons as well why yomichan could stop popping up (could that happen with other options as well?), but this is what I've done so far and what I'm aware of.

Hi, in this thread, people are talking about automatically finding and marking the target word in the sentence, for cards that are already made.
http://forum.koohii.com/thread-14276.html

But when Yomichan is making new cards, it already knows exactly where the target word is in the sentence, so would be able to mark it without any guessing.

I'd like to suggest using <span class="targetword"> or similar, so the appearance is defined in the card CSS, and people can change their mind later about how and whether to display this marking.

Allow for a setting that automatically shows word meanings on hover without the need to hold shift+hover. This would allow for the rikai-like functionality most are already used to, as well as give support to devices without mouse support (touch devices, MS surfaces, etc.).

Envisioning this as a toggleable setting.

The new button for going back to a word from a kanji no longer works. I tried restarting my browser and disabling all other extensions, but this didn't make a difference. I am using Chrome version 57.

As discussed on #23

Hi, looking at this page for example,

https://www.jw.org/ja/%E5%87%BA%E7%89%88%E7%89%A9/%E6%9C%AC/%E8%81%96%E6%9B%B8-%E7%89%A9%E8%AA%9E/2/%E3%83%AD%E3%83%88-%E5%A6%BB/

The text seems to have hidden spaces that appear when copy and pasted, and Yomichan gets caught on them. Mainly it fails at deconjugating past and te-form verbs. I think it was working fine in 0.994 and broke in 0.995 (but maybe something changed on the website).

Continued from this: FooSoft/yomichan-import#7

It would be great if there was an option to ignore alphanumeric dictionary entries.

The yomichan extension can get annoying when I type into a text box and hit Shift while my cursor happens to be hovering over an English word contained in an EPWING entry.

Yomichan doesn't seem to work on posts on Facebook (Shift + mouseover) for me, does it work for anyone else? Not sure if it's because Facebook uses some funky CSS rules, or catches the Shift key or something. Is there a generalisable solution?

Hi, if I change the connection method from AnkiConnect to Disabled and back to AnkiConnect, it forgets all the AnkiConnect settings. I would expect them to come back how they were before. (Testing with 1.0.1 on Windows.)

This extension breaks i-a-n's chrome-image-uncenterer, probably because the CSS is getting overwritten since the image loads on the side, then pops into the middle. A solution would be to not do anything to pages with just an image, or not do anything when Yomichan is turned off.

With Yomichan enabled:

With Yomichan disabled:

Also, I've tested it with all of my other extensions, and Yomichan is definitely the culprit.

There are several sites, like google translator toolkit, where different parts of the page are broken up into iframes. It would be nice to be able to check text in all the frames on sites that still use them.

Would it be possible to add an option to add the card as suspended? I usually have a few extra modifications to make to my cards before I actually want to see them.

I don't know if this is on the extension's end or the source of the audio, but most words are not producing audio. This includes common words like 日本.

Here is a sample word that does work: 思い出.

If there is ruby text on each character in a word, the parser only works on them one at a time. Can we please have the option to let it ignore anything in rt/rp tags?

First of all, thank you for porting this extension for Firefox. Overall it works fine but there are some bugs and room for improvement.

1. I use Yomichan for Chrome and you could toggle the extension on/off by clicking the extension icon but Firefox doesn't have this functionality. This feature also disappeared from the Chrome version of Yomichan, I believe it was available pre-x.x.16?
2. Search form scan works really wonky in Firefox, same as in Chrome. Looks like another string of text is created on top of the text in the search form, here's how it looks like: http://i.imgur.com/oGXdKyt.png
   Selecting text becomes a nuisance with strings overlapping each other. Rikaichan doesn't have this issue.
3. If the audio for a word is missing, you get a long long speech "this audio is unavailable, it will be added shortly blah-blah". It should probably be changed to the Chrome's short sound cue.
4. Would it be possible to ship Firefox extension with RikaiCake dictionary? Link to dict: https://addons.mozilla.org/en-US/firefox/addon/rikaicake-jp-en-dictionary/
5. If possible, some sort of built-in word search/lookup a-la Rikaichan would be nice to have.

Hi,

This is a great extension for chrome. Thank you very much for providing it!
This is a very specific request but possibly it is a larger issue related to the dictionary.

I have this sentence: "キーイメージサイズのサムネ画像を取得し、取得後画像入れ替え。"
When looking up "取得" Yomichan suggests "とりえ" as first and "しゅとく" as second result.

But I believe it should be "しゅとく" and that "とりえ" would normally be written as 取り得, though I am likely wrong :-)

Making a long story short, the request is to somehow improve the listed order of suggestions to match the most common reading of the word.

I'm using Windows 10 and it looks like Alt + ESC is a system level key combination. Because of that, it doesn't clear the search as it should.

If I'm scrolling the Yomichan popup with the mouse scroll wheel and hit the top or bottom of the list, it switches to scrolling the outer page and the popup closes when I didn't mean it to.

I think it would be OK to leave the popup open for any page scrolling.

(Using 0.994 in Fedora, and 0.991 in Windows as it hasn't updated yet.)

Hi!

Steps to reproduce:

go to http://www.asahi.com/ and press shift over various words; yomichan works as expected
press ctrl + s to save the page locally
open the saved page, press shift over various words, yomichan doesn't show anything

Is this expected? Let me know if you need any additional detail.

I find some of the audio files to be too loud. If possible, adding a setting to adjust the volume of the audio files would be very helpful.

I searched for 日本 using the dedicated search page. When I clicked a kanji from the results, I got the following error:
"Error: TypeError: Cannot read property 'text' of undefined"

Starting this tomorrow and would like some feedback.

Firstly, I would like to make this feature compatible with both sentence and vocab cards. I'm thinking I will have to add a checkbox to chrome settings, which will look for the vocab word anywhere in the Reading field. If so it will treat the card as existing, enabling a red X for that vocab entry in the dictionary popup. As you can imagine you might have multiple cards with that vocab word, so multiple cards could be failed that match the Kanji and Kana in the cards Reading. Can I enforce that the user must have a field Reading to use this feature? Or do I need to allow the user to specify a field name from a dropdown of available fields?

Approach:

1. Setup trigger to send fail message on a new red X icon, which will be to the right of the smallest plus for each entry. X will be grayed out if the card does not exist, but will still be present. X will turn red when card exists.
2. In AnkiConnect, mark the matching card as due today, return success if card was found and successfully marked.
3. Give some sort of user feedback on how many cards were failed. Would like some idea how to approach feedback. Small popup in the bottom right hand corner that shows a list of Expressions failed, and then fades out after half a sec?

Can I release this in just Chrome to start?

I believe some of the recent changes broke behavior related to highlighting text. I have the option “Select matched text” turned on, so when I scan and a match is found, the text is highlighted. Previously, the highlighted text was focused on, so doing Ctrl+C would copy the selected text. In version 1.1.8 (not sure about 1.1.7), this functionality no longer works, perhaps because the popover is being focused on instead.

I personally use this to copy text and paste it into jisho.org to see secondary information that may not be included in yomichan.

This is a relatively minor issue, but on the field/value settings, selecting a value on the value dropdown replaces any previous value.

The problem is that users may sometimes want to have more than one value in a field. This is in fact possible by typing in the extra values, but that is not clear to an inexperienced user. Perhaps some way of making it clearer might be helpful. Could be no more than a little extra text in the information section above the grid

(Hey Foosoft, just posting a few requests) Sometimes I want to find out what a single kanji means (e.g. in a name or coined compound or unusual/old word). Rikaisama/chan allows switching to kanji lookup by hitting a hotkey, which is handy for such situations. Would be nice if Yomichan could do this too.

Hi, currently the "add to Anki" icons look the same whether the card is already added, or if there is no response from Anki yet. This can be confusing when skimming over words quickly. I would like to suggest using a different icon for the two situations.

(Hey Foosoft, just posting a few requests) Sometimes (kanji/hiragana) words are written in katakana for emphasis (don't have a recent example unfortunately), and Yomichan doesn't translate them. It would be nice if it had a setting/mode to do this. Essentially, Yomichan could match katakana text against the readings of dictionary words.

If I scroll down in a Yomichan popup, close it, and look up another word, the scroll position is kept. I would expect a new popup to start at the top regardless of what the previous one was doing. (Using 0.994 in Fedora, and 0.991 in Windows as it hasn't updated yet.)

Searching dictionaries is very useful, but less useful since switching keyboard layouts is required. It would be nice if it was not necessary.

(This one's a nice-to-have)
One thing I miss from Rikaisama is the frequency rank that it can output for each word. This helps me decide whether I should export the definition to Anki (I tend not to export uncommon words). Maybe Yomichan could potentially consider displaying this in the future, though the work might be slightly involved. Rikaisama uses the 5,000+ novel (Innocent?) corpus, while Aedict3 uses the Kamermans and Matsushita frequency lists.

Yomichan already displays frequency tags like P, news, ichi, etc. Those are useful and I always export words with those tags to Anki. But most words don't have them, so having specific frequency ranks would be helpful for more advanced learners. For example, I now add words from Aedict3 when they're in the top 20,000 by Matsushita rank.