10ten Japanese Readerv1.28.0

What changed

• UXOptimized cursor lookup logic to avoid unnecessary processing at the end of lines.
• New featureIncreased maximum lookup length from 16 to 20 characters for longer entries.
• New featureAdded external links to Bunpro pages for vocabulary and grammar tags.
• PerformanceImproved lookup performance for long katakana sequences by bounding expansions.
• New featureAdded support for CJK extension characters in scanned Japanese text.
• UXAdjusted popup layering to ensure visibility over dialogs and top-layer content.
• UXFixed text lookup for shadow root children and adjacent inline content.
• PlatformResolved tap and double-tap interaction issues on iOS 26.
• UXFixed a bug where copying entries with hidden kanji omitted specific annotations.
• UXImproved visual readability for the blue theme.

Original release notes

• Avoid looking up text when the cursor is far off the end of the line.
• Increased the maximum length of text we look up from 16 to 20 characters so that longer entries such as 東京ディズニーランド・ステーション駅 are recognized.
• Made the Bunpro vocab and grammar tags link to their page on Bunpro, as the WaniKani tag already does.
• Bound the number of ー expansions generated while looking up text so that long runs of katakana no longer make lookups very slow.
• Recognize CJK extension characters in scanned Japanese text.
• Show the popup above dialogs, popovers, and other content in the top layer.
• Fixed text lookup for text nodes that are direct children of shadow roots, including scanning across adjacent inline content.
• Fixed tap and double-tap handling for the lookup puck on recent iOS 26 releases.
• Fixed copying an entry whose kanji headwords are all hidden dropping the "usually kana" annotation from the copied text.
• Fixed some readability issues with the blue theme.