Re: tlug: Kanji to Hiragana soft

[Rex Walters <rex@example.com>]

>>>>> Jim Breen writes:  (on 25 Oct 98)

> >> Eric S. Standlee asked:
> >> Is there a software package that will change kanji in to hiragana
> >> (furigana) for linux.  I need to take a large amount of Japanese text and
> >> filter kanji into hiragana for those who cannot read kanji well.
> 
> By now the main answers have been given. I think kakasi is probably the
> best for you, as it in-lines the kana. I haven't tried sumono yet, but
> juman/chasen is an excellent morphological analyzer, and fairly
> adaptable. I haven't run it under Linux, but it's fine with Solaris.
> 
> If a glossing is what you want (and this includes the kana expansions),
> you can try Hilofumi Yamamoto's "mailgloss" system (see:
> http://www.intersc.tsukuba.ac.jp) or my own WWWJDIC. The former uses
> juman with an EDICT match as a back-end. WWWJDIC uses EDICT searches
> alone.

This has *got* to be the most parochial, pedantic, jargonistic list I
subscribe to!  :-)

"Glossing" leaves me completely baffled.  "Morphological analyzer" is
too frightening to contemplate!  What do these programs *do*?

Kakasi sounds sufficiently useful that I will give it a spin (someday --
if I can ever quit pounding on the ^R key long enough to get through my
tlug backlog).

At the risk of offending the "show me the code" purists and asking for
something that I should probably write myself, I must say that I
actually miss the capabilities of your old jreader program.  The new
xjdic is lovely, but it's still comparatively painful for me to page
through Japanese text in one kterm and cut-n-paste unknown kanji to
another kterm running xjdic.  

My memory of jreader is vague, but if I remember correctly you could
advance the cursor phrase-by-phrase by hitting the space-bar, whenever
you came to an unknown kanji phrase another keystroke would bring up the
corresponding edict entry.  Wonderful for people with kanji reading
skills as poor as my own!

I've often toyed with the idea of extending the "less" pager to support
the "advance cursor by phrase" concept and add hooks for edict lookup,
but the payback simply isn't enough to justify the time required (as a
valid workaround already exists).  

"A place to stand" heck!  Give me enough *time* and I'll move the world!  

I really do find the "stop, find the mouse, hilite the unknown text,
paste into xjdic" process quite interruptive and impairing to my reading
comprehension.  A quicker way to look up just the kanji I'm looking at
RIGHT NOW would be really really nice.

I'm curious if you know of anything similar to your old DOS jreader
program that's currently available?  I'd really like to be able to pipe
occasional Japanese email or news posts to such a program.

Regards,
-- 
Rex
---------------------------------------------------------------
Next Nomikai: 20 November, 19:30 Tengu TokyoEkiMae 03-3275-3691
Next Technical Meeting: January, 1999 (details TBA)
---------------------------------------------------------------
Sponsor: PHT, makers of TurboLinux http://www.pht.co.jp