Shiba is an advanced Japanese dictionary app for Android, designed as a spiritual successor to the now-defunct Kantan. It supports a powerful search syntax that streamlines the process of identifying unknown kanji.

For example, take the word 躊躇 — a veritable mess of tiny components comprising two characters that aren’t even included in the kanji lists that native Japanese people would learn in school. To look this word up the traditional way, you would have to first identify all the radicals or graphemes in the first kanji, plug each one into a dictionary until the character is found, then repeat for each other one until you’ve spelled out the whole word.

With Shiba, all you need to do is identify a few graphemes from each character, plug them into the special search syntax, and let the search engine do the rest. In this example, we’ve identified 士 from the first kanji, and ⺾ from the second.

Unlike with a traditional paper dictionary, no special knowledge of kanji radicals is required — as long as you can squint at a character and identify some basic shapes, you can find anything.

Shiba also supports a wide variety of wildcards; the empty set [] will match any kanji, while a period (. or 。) will match any kana, a question mark (? or ？) will match any character at all, and an asterisk (* or ＊) will match 1 to many characters of any type. Even for searches with outrageously lenient patterns, results still come back in a matter of moments — searching for “[][][][]”, which will match any four-kanji compound (of which JMDict includes over 19,000), or even simply “*”, which will match everything; in both cases, results are returned in no more than a second on my Pixel 2.

I’ve been using Shiba in my own Japanese studies ever since the first prototype and I’m excited to share it with the world. Watch this space for further details!

Shiba is being developed in C# with Xamarin.Forms. Launching soon for free on the Google Play store.