Pierre Lewis wrote:
> Just checking the Book here, I understand from table 3-5 (p. 3-17)
> that both PS and LS are block separators (B), ie. treated the same
> wrt BIDI.
>
> The rest of the description seems to indicate that BIDI processing
> is done on blocks. So I assume that ones starts anew after either
> PS or LS.
I think this is right, but didn't want to say so with great
definiteness. But I reiterate my former point: if PS and LS
are block separators in Unicode plain text, then <P> and <BR>
are block separators in Unicode HTML text.
Thus (under the usual convention that a-z are type L and
A-Z are type R):
ABC<BR>DEF should look thus:
CBA
FED
and abcABC<BR>DEFdef with overall type R should look thus:
CBAabc
defFED
and with overall type L thus:
abcCBA
FEDdef
--- if this is wrong, what I mean is: just as if the <BR> sequence
(U+003C U+0042 U+0052 U+003E) were an explicit U+2028. Making this
equivalence is, as I said, a higher-level protocol with respect to
plain Unicode.
-- John Cowan cowan@ccil.org e'osai ko sarji la lojban
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:34 EDT