W3C home > Mailing lists > Public > www-style@w3.org > November 2003

Re: [CSS2.1] Text 16.6.2 Bidirectionality with Whitespace Collapsing

From: Tex Texin <tex@i18nguy.com>
Date: Wed, 05 Nov 2003 19:54:11 -0500
Message-ID: <3FA99BB3.576D86E4@i18nguy.com>
To: Najib TOUNSI <tounsin@yahoo.com>
Cc: fantasai <fantasai@escape.com>, www-style@w3.org, W3c I18n Group <w3c-i18n-ig@w3.org>, Richard Ishida <ishida@w3.org>

Najib,

As you know, the GEO group welcomes writing contributions and would be glad to
have you get involved, if not in this FAQ more generally. You might like to get
involved with the bidi section of the Authoring Techniques document.
http://www.w3.org/International/geo/html-tech/

With respect to the FAQ,
http://www.w3.org/International/questions/qa-bidi-space.html and your question:

> To summerise, is it worth to say that, in mixed texts,
> the natural bidi algorithm works until the author
> changes his text adding some punctuations characters?

You are of course right that the bidi algorithm influences the treatment of
neutral bidirectionality characters such as punctuation, and that this
influences where directional runs begin and end, and this may not be the
results the author wants. Having some FAQs discussing these issues from an
applied perspective or an authorship perspective would be great. I have seen a
lot of theoretical discussions; some experiential material would be helpful,
especially to those fearful of working with bidi.

I like the trick of using ] to distinguish parens and mirrored parens. I might
use that for a FAQ.

I think some more discussion and analysis is needed around bidi and authoring
stylesheets as well.
The recommendations to favor HTML markup over CSS for bidi and when to use
unicode-bidi:embed etc. could use more expansion, with real world examples and
discussions of design decisions and rationale, so that the more occasional
authors of bidi material understand proper coding of the pages.

If you or others can contribute it would be appreciated. Talk to the GEO chair,
Richard Ishida.

tex


Najib TOUNSI wrote:
> 
> This GEO FAQ
> (http://www.w3.org/International/questions/qa-bidi-space.html)
> deals with two things at the same time:
> - the collapsing spaces
> "if text is followed by white space inside an inline
> element that includes a dir attribute"
> - and with the may be unnecessary dir="ltr" attribute
> " Leaving out the attribute or the whole span element
> will also solve the problem."
> 
> In fact, the example shown may suggest that the dir
> attribute is not necessary at all. But suppose the
> absence of span.
> The whole page beeing rtl, if the HTML source is
>    arabic1 LATIN (W3C) arabic2
> you'll get (read from right to left!)
>    2cibara (LATIN (W3C 1cibara
> 
> The ")" after C in the source file, is taken as an end
> arabic "(" displayed in continuation of the rest of
> arabic text.
> It is the second latin word befor (W3C) which shows
> this phenomena.
> So the solution is
> <span dir="ltr">LATIN (W3C)</span>
> or
> LATIN (W3C)&lrm;
> (see http://www.w3c.org.ma/Tests/span-dir.html for an
> example)
> 
> In the FAQ example; the problem is a little bit more
> subtle.
> In fact, the () surrounding (W3C) looks the same after
> beeing inverted in Arabic. Just put (W3C] in this
> example to see [W3C) in browser
> (http://www.w3c.org.ma/Tests/span-dir2.html is an
> example for this situation)
> Note that in this latter case, TWO &lrm; are
> necessary: &lrm;(W3C)&lrm;
> 
> To summerise, is it worth to say that, in mixed texts,
> the natural bidi algorithm works until the author
> changes his text adding some punctuations characters?
> 
> So, it might be helpful to seprate into two FAQs the
> collapsing-spaces-problem (the actual qa-bidi-space
> FAQ) from the dir=ltr problem, which in this case has
> to do with the neutral punctuation character between
> ltr and rtl texts. I could write (or help in writing)
> this last one eventually.
> 
> Regards,
> 
> Najib Tounsi
> 
> --- Tex Texin <tex@i18nguy.com> wrote:
> >
> > I think I agree with these comments, but I am asking
> > the i18n group to comment.
> >
> > The FAQ that the GEO group wrote on this suggested
> > "remove all space before the
> > end tag of the inline element, or remove the dir
> > attribute (if appropriate)."
> >
> > Here it makes sense to remove unnecessary embedding,
> > not to remove explicit
> > overrides.
> > Also, the key impact is on spaces before end tags,
> > not beginning spaces.
> >
> >
> http://www.w3.org/International/questions/qa-bidi-space.html
> > tex
> >
> >
> > fantasai wrote:
> > > 16.6.2 Bidirectionality with Whitespace Collapsing
> > >
> >
> http://www.w3.org/TR/2003/WD-CSS21-20030915/text.html#q9
> > >
> > > # This is best avoided by using the natural
> > bidirectionality of characters
> > > # instead of explicit embedding levels.
> > >
> > > It's not always possible to rely on implicit bidi,
> > which is why we *have*
> > > embedding levels. You shouldn't discourage their
> > use like this. The problem
> > > is best avoided by keeping start and end tags
> > close to the text: by not
> > > putting spaces at the start and end of inline
> > elements.
> > >
> > > ~fantasai
> >
> > --
> >
> -------------------------------------------------------------
> > Tex Texin   cell: +1 781 789 1898
> > mailto:Tex@XenCraft.com
> > Xen Master
> > http://www.i18nGuy.com
> >
> > XenCraft                          http://www.XenCraft.com
> > Making e-Business Work Around the World
> >
> -------------------------------------------------------------
> >
> 
> __________________________________
> Do you Yahoo!?
> Protect your identity with Yahoo! Mail AddressGuard
> http://antispam.yahoo.com/whatsnewfree

-- 
-------------------------------------------------------------
Tex Texin   cell: +1 781 789 1898   mailto:Tex@XenCraft.com
Xen Master                          http://www.i18nGuy.com
                         
XenCraft		            http://www.XenCraft.com
Making e-Business Work Around the World
-------------------------------------------------------------
Received on Wednesday, 5 November 2003 19:55:09 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:25 GMT