Yomitan Settings
This page is taking longer than expected to load.
Due to a bug in Safari, it may be necessary to click the 
Yomitan
button in the browser bar to fully load the page
Dictionaries (# installed, # enabled)
Web browsers will sometimes clear stored data if the device is running low on storage space. This can result in the imported dictionaries being deleted unexpectedly, causing Yomitan to stop functioning. Enabling persistent storage tells the browser that the data should not be deleted in those circumstances.
On Firefox and Firefox for Android, the storage information feature may be hidden behind a browser flag. To enable this flag, open about:config and search for dom.storageManager.enabled. Setting its value to true should allow storage information to be calculated.
It may not be possible to enable Persistent Storage on Firefox for Android.
Chromium-based browsers should not need to enable this setting since the Yomitan extension has
the unlimitedStorage permission, which should prevent data deletion.[1]
- No grouping - Every definition will be listed as a separate entry.
- Group term-reading pairs - Definitions for the same term with the same reading will be grouped together.
-
Group related terms - Related terms that share the same definitions will be grouped together.
The Primary dictionary option should be assigned to a dictionary which contains related term information, and configuring the Secondary dictionaries will allow definitions for the related terms to be included from other dictionaries.
Not all dictionaries are able to be selected as the Primary dictionary.
Enabling this option will sort search results using a specific dictionary. This can be beneficial when using multiple dictionaries which may not have consistent sorting information.
Dictionary frequency data can be represented in one of two ways:
- Occurrence-based, where the frequency corresponds to a number of occurrences. Large values indicate a more common term.
- Rank-based, where the frequency value corresponds to a ranking index. Smaller values indicate a more common term.
The correct mode can be determined based on the contents of the dictionary; the Auto button attempts to auto-detect the correct value.
A keyboard modifier key can be used to activate text scanning when the cursor is moved. Alternatively, the No key option can be used to scan text whenever the cursor is moved, without requiring any key to be held.
More advanced scanning input customization can be set up by enabling the Advanced option and clicking Configure advanced scanning inputs.
zoom is used.
More…
The non-standard CSS zoom property interferes with the normal calculation of the pointer coordinates when scanning webpages. This property is discouraged from being used and its use is rare, but some webpages may still use it.
Enabling this option, which is on by default, will take the value of this property into account when scanning webpage content. It is currently put behind an option in case there are unforeseen negative side effects.
This setting does not have any effect in Firefox, as it does not implement the zoom property.
Rather than searching for the source text exactly, the text will only be required to be a prefix of an existing term. For example, scanning 読み will effectively search for 読み*, which may bring up additional results such as 読み方.
This will likely cause scanning and lookup to be slower, and the results may not be as relevant.
To enable text scanning when using devices with a touch screen, either the Touch inputs or the Pointer inputs option must be enabled. Touch inputs supports generic touches on a touch screen device, but do not distinguish between touch and pen inputs. Pointer inputs supports supports the detection pen devices, but may not work on all devices. If both options are enabled, Pointer inputs takes precedence.
The Pointer inputs option is only visible when the Advanced option is enabled.
By default, scanning text inside of an embeded <iframe> element will open
a new popup inside of that frame, which can sometimes be limiting due to the frame's size.
When this option is enabled, the popup will be created in the root of the page, if possible.
Note that when this option is enabled, there is a possibility that the extension can interfere with the underlying webpage, since it must send messages to the webpage in order to determine the correct position of the popup. This typically does not cause issues, but if anything unexpected happens, this option could be the cause.
The Default mode will position the popup relative to the scanned text. The Full Width mode will anchor the popup to the top or bottom of the screen and take up the full width of the screen, which can be useful on devices with touch screens.
Auto-scaling will scale the popup automatically based on the browser's zoom levels in order to keep the popup at a constant physical size, regardless of the zoom level. Zoom level uses the zoom level that is typically used on desktop browsers, and Viewport uses the zoom level that is typically used on mobile browsers.
Instead of showing definitions in a popup embedded into the webpage, a native browser window containing the popup content will be opened instead. This window will be shared across all tabs.
Yomitan is able to scan the sentence surrounding a term and parse individual words of the query on the search page. This information can also be added to Anki cards to provide additional context.
Yomitan's internal parser is a simple algorithm which can parse sentences using the installed dictionaries. It works by repeatedly detecting the longest term at the start of the sentence and advancing until no text remains. This provides decent results for the general case, but since grammatical structure is not taken into account, it can sometimes yield incorrect segmentation.
MeCab is a third-party program which uses its own dictionaries and parsing algorithm to decompose sentences into individual words. MeCab may provide more accurate parsing results than Yomitan's internal parser.
In order for Yomitan to use it, both MeCab and a native messaging component must be installed. A setup guide can be found here.
The following options are used during the translation process to create alternate versions of the input text to search for. This can be helpful when the input text doesn't exactly match the term or expression found in the database.
The conversion options below are listed in the order that the conversions are applied to the input text. Most of the conversions have three possible values:
- Disabled - This conversion will never be applied to the input text.
- Enabled - This conversion will always be applied to the input text.
- Use both variants - The translator will check the database for two variations: the raw input text and the converted input text. When multiple options use variants, the translator will search for combinations of the converted text.
"A dog" → search for "A dog","A do", "A d", "A"
or"A dog" → "A dog", "A"
Yomitan supports automatic flashcard creation for Anki, a free application designed to assist in remembering information. This feature requires installation of the AnkiConnect plugin.
Attempting to connect to Anki can sometimes return an error message which includes "Invalid response", which may indicate that the value of the AnkiConnect server address option is incorrect. Resetting it to the default value may fix issues that are occurring.
Notice for macOS users: If Yomitan has issues connecting to AnkiConnect, it may be necessary to adjust adjust some system settings. See this link for details.
The default address for a server on the same device is http://127.0.0.1:8765. If Anki is running and AnkiConnect is installed, clicking this URL should open a page showing the current version of AnkiConnect.
By default, duplicate checks are only performed for notes created with the same model. Enabling this option will check for duplicates across all model types.
A card is considered a duplicate if the value of the first field matches that of any other card. By default, this check will include cards across all decks in a collection, but this constraint can be relaxed by using either the Deck or Deck root option.
The Deck option will only check for duplicates in the target deck. The Deck root option will additionally check for duplicates in all child decks of the root deck. This allows adding cards that are unique for decks including a subdeck structure. For decks which don't have any parent-child hierarchy, both options function the same.
Audio files can be downloaded from remote servers when creating Anki cards, and sometimes these downloads can stall due to server or internet connectivity issues. When this setting has a non-zero value, if a download has stalled for longer than the time specified, the download will be cancelled.
AnkiConnect releases after around 2022-05-29 support a new note editor window which can be shown when clicking the View added note button. This can be tested using the buttons below. If an error occurs, Anki and/or AnkiConnect may need to be updated.
When coming across a word that is already in an Anki deck, a button will appear that shows the tags the card has. If set to Non-standard, all tags that are included in the Card tags option will be filtered out from the list. If no tags remain after filtering, then the button will not be shown.
Yomitan can watch the system copy clipboard for Japanese text and automatically search the text on the search page. This can be useful for using Yomitan to scan text from external applications.
This feature requires Yomitan to have clipboard reading permissions, and when this option is enabled, the clipboard will be frequently checked for new text.
Yomitan has two categories of keyboard shortcuts:
- Standard keyboard shortcuts are controlled by the extension, and can be added, removed, and configured to work on webpages that Yomitan functions on.
- Native keyboard shortcuts are controlled by the web browser, and function globally within the web browser or system-wide.
See yomichan-data-exporter for instructions to export your data from older Yomichan installations.
Google Docs now uses canvas-based rendering to display content[2] which prevents Yomitan from being able to scan text using the standard methods. Enabling this option will force Google Docs webpages to expose some additional text information which should allow Yomitan to still work.
Google has changed this compatibility implementation several times, and the changes do not seem to be announced or documented. Therefore, it is possible that this feature could stop working at any time the future without warning.
On supported browsers,
a popup's iframe element will be embeded inside of a container with a closed shadow DOM.
This container prevents scripts running on the underlying web page from being able to discover the iframe,
which helps avoid situations where the web page might try to modify or use the Yomitan popup for an unintended purpose.
When this option is enabled, the URL of the iframe element will be assigned by
directly changing the location of the iframe's internal document, rather than using
the iframe's src
attribute. This results in the extension URL not being exposed to the underling web page, and thus making it harder
to detect the presence of Yomitan.