As a follow up to Furigana engine: modified version with various bugfixes, added features, I've tried to fix various issues which have cropped up, and posted the changes to https://gist.github.com/cyphr/6536814.

Changes are:

The mobile site is theoretically supported as well, but I don't know whether this script can be added to it.

P.S. I'm not as active here as I once was, so if anyone wants to take over the development of this script feel free. Otherwise, I'll try to maintain/add features to it as the community here decides - please post any requests here in this thread, or as a separate meta post with a link to the post in the comments and I'll see what I can do :)

  • 7
    – user1478
    Sep 21, 2014 at 7:42

2 Answers 2


This was a little weird:

enter image description here

But istrasci fixed it!

enter image description here

Here's the revision in question:

enter image description here


  • 1
    uh, yeah...I rolled out a fix (of sorts) for this issue to GitHub half an hour or so ago :) Basically "目覚め" with furigana "め・ざ・" containing character separators caused a <rb>め</rb> with a blank <rt></rt> to be created. I've made it so that blank or the same Furigana with character separation syntax outputs without Furigana, but you still have to have separators for each character separately as it is currently, e.g. "目覚める" can now be "め・ざ・め・る" or "め・ざ・・" but "め・ざ・める" won't work.
    – cypher
    Sep 29, 2014 at 22:09

I don't know if it applies to everyone, but I have a problem with full-width and half-width tilde characters: I can produce three of them, two full-width and one half-width.

The full-width tilde renders differently, loosely depending on whether a Japanese or Latin character follows or precedes it or doesn't:

Cf.「~で記入して」 and



(same tilde in both cases)

It's a small thing, but I've tried to put in ~から~まで (same tilde!) and failed.

The issue already is present in the input text field.

  • 1
    This should mostly be fixed in the updated version, as Japanese fonts will be used for both fullwidth Latin and Japanese (img: postimg.org/image/hurzsioyx). Previously, Japanese fonts weren't explicitly specified, and it used the OS/font rendering system to find a font to display glyphs that weren't found. I don't think there's much I can do about the input text field short of using a Japanese font, unfortunately like I said on chat I think that could be worse than the disease.
    – cypher
    Oct 2, 2014 at 8:48
  • Can you provide a screen shot? On my system on both SeaMonkey and Chrome they render the same way. But I think this could be related to East Asian ambiguous characters. Is considered to have ambiguous width in Unicode?
    – user1478
    Oct 3, 2014 at 3:37
  • @snailboat Actually, the difference is more visible on my system, so I updated my post with my own screenshot.
    – Earthliŋ Mod
    Oct 3, 2014 at 16:32

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .