The Wave Dash in Japanese Punctuation
While trying to copy a printed vocabulary list into LibreOffice, I stumbled over a rather wide tilde character that was used as a placeholder as in
which means “to enjoy something” (namely, the thing represented by the tilde). Now I had quite a hard time typing “～” using Anthy under Ubuntu Linux and dug a bit into what this tilde character actually is.
Apparently, in Japanese punctuation the so-called “wave dash” (〜, U+301C) is used for ranges (5時〜6時, from 5 o'clock to 6 o'clock) or in places where in Western languages the colon, dash or brackets would be used.
Now as explained in the Wikipedia article about the Tilde character, when the wave dash character was imported into Unicode from JIS X 0208 character 1-33, the reference glyph was turned upside down.
This means that a JIS-compliant font should render the wave dash as
while a font that adheres strictly to the Unicode reference glyphs would have to render the same character as
which is obviously a problem. (However, this is mitigated by the fact that many fonts ignore the Unicode reference glyph and just render the correct wave dash character instead. For example, my default font in the browser displays the latter glyph, while my text editor's font displays the former.)
As a consequence of this glyph mismatch, the character “full-width tilde” (～, U+FF5E) is often used as a replacement for the wave dash, since the Unicode full-width tilde looks the same as the JIS wave dash. In particular, Windows will insert the full-width tilde U+FF5E when the user types a wave dash so that it “is generally difficult, if not impossible, for Windows users in Japan to type U+301C” (from Wikipedia).
Also, on Linux systems using Anthy, the full-width tilde U+FF5E will be inserted when “~” (normal tilde) is typed, as can be read – at least on Ubuntu systems – in the file
/usr/share/ibus-anthy/engine/tables.py (read about that on http://moritzmolch.com/349/some-secrets-of-ibus-anthy.html). However, on Mac OS X apparently the (correct) character U+301C is used.
So, it seems like using the full-width tilde character (instead of the wave dash) is quite ok for typing vocabulary lists. Just, unfortunately, it is apparently not possible to generate a simple tilde character (that would trigger the “～” insertion in Anthy) when using a keyboard layout where the tilde is a dead key ...