Re: (dsssl) Japanese and OpenJade

[Adam Di Carlo <adam@xxxxxxxxxxx>]

jany.quintard@xxxxxxxxxx writes:

> Anyway, if I have good results with ucs-2 unicode and koi8-r too,

Do you think I can get fuller demonstrations of koi8-r and ucs-2 and
sjis?  I am getting a lot of complaints from Russian users for
instance about difficult using jade/openjade and koi8-r -- if I could
provide and test with some complete examples (including requisite env
vars) then I could test this better.  Both for Debian and for OpenJade
in particular.

> With, sjis, I get good results too, but Openjade complains upon errors
> in the stylesheet :
> 
> openjade:/xo/.../sgml/sgml.dsl:71:44:E:invalid character after '#'
> (define (strip str #!optional (stripchars '(#\space #\&#RE #\U-0009)))
> (strip is one Norman Walsh's function)
> I supposec that this a result of the stylesheet being parsed using
> SP_ENCODING.

Uh.  Ouch.  I take it that '!' is remapped in sjis.  

Are you sure that's referring to Norm's stylesheet properly, e.g.,

  <!ENTITY docbook.dsl PUBLIC
        "-//Norman Walsh//DOCUMENT DocBook HTML Stylesheet//EN" CDATA DSSSL>

?  Perhaps try SDATA or NDATA?  (just a wild guess, shouldn't be necessary).


> I would be happy to get rid of this error, because I am afraid it will
> prevent the function to work, if needed.

Yes, that is scheme magic to denote optional arguments.

-- 
.....Adam Di Carlo....adam@xxxxxxxxxxxxxxxx<URL:http://www.onshored.com/>


 DSSSList info and archive:  http://www.mulberrytech.com/dsssl/dssslist