[luatex] getting previous glyphs

[Norbert Preining]

Hi everyone,

I am trying to write some macros to convert Aozora (Japanese literature
site) formatted text to something I can feed through luatex.

In the Aozora format, ruby-annotations (the reading of Kanji written
on top of the Kanji) are written *after* the word, like in
	nnnKKK<uuuuu>
where 
	nnn
		are some *non* Kanji (Hiragana, Roman, etc)
	KKK
		are Kanji
	< >
		are Japanese angle brakets
	uuuu
		reading in Hiragana

TO make things worse, if there is the need to separate some Kanjis
from each other, then a | (in Jpanaese version) is added:
	KK|KKK<uuuu>
in this example all the K's are Kanji, but the ruby should only
be applied to the group of three K after the |

Now, luatexja, in particalur luatexja-ruby, provides a handy
\ruby{KKK}{uuuu} macro, but for getting the KKK I have to read
backwards somehow.

As we are already in luatex, do you have any idea how to achieve
this easily?

Thanks for any hint

Norbert

------------------------------------------------------------------------
PREINING, Norbert                               http://www.preining.info
JAIST, Japan                                 TeX Live & Debian Developer
GPG: 0x860CDC13   fp: F7D8 A928 26E3 16A1 9FA0  ACF0 6CAC A448 860C DC13
------------------------------------------------------------------------