RE: DOCBOOK: Attributes for text direction and language

["HAMILTON,DICK (HP-FtCollins,ex1)" <dick_hamilton@hp.com>]

> Thank you very much for your reply. You are right, but the problem
> is that numbers are neither LTR nor RTL intrincically, and are
> interpreted either LTR or RTL depending on the direction of the
> text preceding them.
> 
I just took at look at the unicode bidirectional algorithm
as described in: http://www.unicode.org/unicode/reports/tr9/.
It is very complex and I don't claim to understand it in its
full glory, but there are codes that provide explicit directional
information, and the handling of numbers is covered in some
detail in the algorithm.  I can't say for sure if it will handle
what you're doing, but since Hebrew was explicitly considered in
the development of the algorithm, I think there's a good chance
it will do what you need (if browsers and stylesheets use the
algorithm).
>
> I asked what I asked because I'm planning to convert my documents,
> mostly in the field of (Hebrew) linguistics, from Word format to
> DocBook XML. Though DocBook may not be originally meant for
> linguistic documentation originally, I thought (and still think) it
> might be usable for this purpose without customizing its DTD.
> 
I think you're right and that you should be able to do this
without customizing the DTD.

You may want to check out the unicode organization's web site.
They have a lot of information and at least one archived mailing
list that might give you some information if you decide to pursue
a unicode approach.

Good luck,
Dick
>
> -----Original Message-----
> From: Tsuguya Sasaki [mailto:ts@ts-cyberia.net]
> Sent: Monday, June 17, 2002 2:45 PM
> To: DOCBOOK
> Subject: Re: DOCBOOK: Attributes for text direction and language
> 
> 
> > It's been a while since I looked at this, so take
> > it with a grain of salt, but I think that if you
> > use Unicode (or UTF8), the codeset itself provides
> > the information you need to render right to left
> > and left to right text, including some characters
> > that act as cues for cases where there might be
> > ambiguity.  There is a technical report from the
> > unicode consortium that discusses this:
> >
> > http://www.unicode.org/unicode/reports/tr9/
> >
> > While that seems like the cleanest way to handle
> > RTL and LTR text, I don't know whether browsers
> > and/or stylesheets use this information.
> 
> Thank you very much for your reply. You are right, but the problem
> is that numbers are neither LTR nor RTL intrincically, and are
> interpreted either LTR or RTL depending on the direction of the
> text preceding them.
> 
> So things can become quite complex and messed up, for example,
> when you try start an LTR paragraph with RTL text chunks followed
> by numbers; it's interpreted as an RTL paragraph, numbers come to
> the right of the RTL text chunks, which is the correct visual order in
> RTL paragraphs.
> 
> I asked what I asked because I'm planning to convert my documents,
> mostly in the field of (Hebrew) linguistics, from Word format to
> DocBook XML. Though DocBook may not be originally meant for
> linguistic documentation originally, I thought (and still think) it
> might be usable for this purpose without customizing its DTD.
> 
> Tsuguya Sasaki
> http://www.ts-cyberia.net/
> 
> PS: Please reply only to the list. If you send a reply to me 
> and to the
> list, I receive two copies of the same message.
> 
> 
>