Re: [css3-text] New Working Draft

"7.2. Emergency Wrapping: the ‘word-wrap’ property
[...]
break-word
  An unbreakable "word" may be broken at an arbitrary point if there
are no otherwise-acceptable break points in the line. Shaping
characters are still shaped as if the word were not broken, and
grapheme clusters must together stay as one unit.[...]"

Here I also suggest that contextually shaped characters should not
just keep their normal shaping, but the joining types should be taken
into account, to avoid breaking between joined character pairs, with a
higher precedence for disjoined characters.

Joining types here are what is currently standardized for Arabic, but
it could also apply to any script using a joined cursive style
(including Latin, Greek, Cyrillic). A problem appears : not all
scripts that can be rendered in a joined cursive style have joining
types defined, and the actual source for specifying such joining may
be found in fact within fonts, as substitution&positioning entries ;
however many fonts still use the same feature only to adjust the
kerning, without creating any junction or ligature, e.g. between "AV",
and many optional ligatures are still breakable, reversing the effect
of contextual shaping of its constituant characters. One good decision
would be to list only the scripts that are naturally joining. Are
there other scripts than Arabic to consider ?

Also for Latin, at least, the shaping of characters in a joined
cursive style is not preserved when wrord-wraps are occuring : the
hyphen is effectively inserted and letters on each side are shaped by
forcing them to their disjoined shaping. Consider for example the
shape of letter "j" when it joins with the following letter (most
often a vowel) or when it occurs at the begining (leading tip or not)
or end (closed curl or not) of word slice. Consider the shape of "r"
(its leading junction) after either "o" or after "a" or "t" or "l".

So I think that the current sentence is wrong : the presence or not of
the wrap affects the shaping and does not necessarily preserve it as
if there was no wrap.

Philippe.

Received on Wednesday, 20 April 2011 23:53:36 UTC