Re: [emacs-bidi] status? news?

[Roozbeh Pournader]

On Sun, 21 Jan 2001, Eli Zaretskii wrote:

> Could you please elaborate on these problems with next text lines?

With implicit bidi support (with LRM and RLM, but without LRE, RLE,
LRO, RLO, and PDF), you can first break the lines, and then apply bidi to
each line separately for reordering before the display.

With explicit support, you should first do bidi on the whole paragraph,
then break the lines, and finally reorder the text. This is considered
hard to do in many applications.

> HTML has additional bidi directives, which I imagine are easier to 
> support than the Unicode controls.  So HTML is probably in less trouble 
> than plain text.

HTML needs the same treatment. Let me list its bidi-related things here
(available at http://www.w3.org/TR/html4/struct/dirlang.html#h-8.2):

* "dir" attribute of block elements (<DIV>-like), which takes a value of
"LTR" or "RTL". It sets the exact equivalent of paragraph embedding level.

* "dir" attribute of inline elements (<SPAN>-like), which takes "LTR" of
"RTL", and is exactly equivalent to LRE-PDF or RLE-PDF.

* <BDO> element, which should have a "dir" attribute, and becomes the
exact equivalent of LRO-PDF or RLO-PDF.

* character references &lrm; and &rlm; which are the same as LRM and RLM.

* discouraging the direct use of LRE, RLE, LRO, RLO, and PDF characters in
a further W3 Note (http://www.w3.org/TR/unicode-xml/#Bidi).

For an HTML display engine, the only difference is that you may ignore the
explicit marks, but you should instead support the "dir" attribute and the
"BDO" element. The only difference may be that you're now sure that the
parser inserts the end tags (equivalents of PDF) for you, which is really
small.

For an HTML editor, the situation is worse. You should replace the
explicit marks (from any source) with appropriate markup.

--roozbeh