On Feb 1, 2008 5:31 PM, Werner LEMBERG <[EMAIL PROTECTED]> wrote:
> > I tried to get this to work in XeTeX, and as per some examples in
> > the xCJK package, macros did not work.
>
> At start-up time, XeTeX is in UTF-8 character mode, and Japanese
> characters like the geta mark are treated as normal characters,
That is what I thought, but probably I am imagining something
incorrectly. I read the fontspec guide, and thought why do we now need
CJK if XeTeX can access the system fonts directly in UTF8 mode? I
changed the main font to a Japanese one and lo and behold, the Roman
letters were now in that font. However, XeTeX gave an error as soon as
I tried to enter Japanese characters also. Again, I cannot say whether
this is the fault of my installation or a misconception of mine. In
any case, I thought this might be a reason to require the CJK macros,
although I don't understand why in detail: from what point does the
XeTeX engine no longer handle Japanese characters if it can handle
them as normal characters in the input? Fontspec documentation
(v.1.14) example on page 21 indicates that it should "just work":
\fontspec{some Japanese font} ...Japanese characters...
The other thing that is still confusing me about XeTeX and CJK
together is what font names are required where. I'll have to check
this, but it seems the NFSS names no longer work for me at all for
setting the CJK environment, and the default Wadalab gets chosen
instead. The XeTeX and xCJK font commands however do as I expect and
this is how I am working successfully at the moment with XeTeX and
macros.
> without the CJK handling of UTF-8 active. If you look into xCJK.sty,
> you can see the line
>
> \AtBeginDocument{\XeTeXinputencoding "bytes"}
>
> at the very end; this means that only after \begin{document} CJK takes
> care of UTF-8 handling.
Excuse me, to clarify can you explain: does this mean that at begin
document the mode becomes "bytes" and that CJK requires "bytes" mode
to handle the CJK characters? I take it that "bytes" mode means
encoding-independent, as opposed to the default of UTF-8
interpretation of a sequence of bytes?
Regards, Gernot
_______________________________________________
Cjk maillist - [email protected]
https://lists.ffii.org/mailman/listinfo/cjk
