RE: 'knife and fork' model

A "select all lines" selector could handle the prefix and suffix insertions below, e.g.,
 
<animation id="a1">
  <set attributeName="content" to="#"/>
</animation>
<cue select="//p/area::line(*)/prefix()" use="a1"/>
<cue select="//p/area::line(*)/suffix()" use="a1"/>
 
For speaker sensitive generated content, one could use a class or role attribute and use a complex selector predicate, e.g.,
 
<animation id="a1">
  <set attributeName="content" to="--"/>
</animation>
<animation id="a2">
  <set attributeName="content" to=">>"/>
</animation>
<cue select="//p[@role='mary']/area::line(*)/prefix()" use="a1"/>
<cue select="//p[@role='jane']/area::line(*)/prefix()" use="a2"/>

	-----Original Message-----
	From: Johnb@screen.subtitling.com [mailto:Johnb@screen.subtitling.com] 
	Sent: Wednesday, August 13, 2003 12:09 PM
	To: Glenn A. Adams
	Cc: public-tt@w3.org
	Subject: RE: 'knife and fork' model
	
	
	Glenn

		[GA.] In order to express authoring time line separators, we have included force line break as a presentation flow vocabulary item and break-after and break-before as style properties that induce the same effect.
		[JB> ] Might I suggest also a requirement for prefix-before in order to support the inclusion of symbols such as # (for music) and - or >> for speaker change? Or can this be handled in another way? - can't be in content if supporting soft layout! 
		

			[GA] : This model is supported by the Generated Content mechanisms of CSS2, e.g., using the :before and :after pseudo-element selectors in combination with the content property. However, there are some possible problems with using such a facility for anything more than very simiple content generation, e.g., one can't specify marked up content in the style sheet in order to correctly support generic internationalized text.
			
			
			Thus far, we have not entertained a requirement to support stylesheet based generated content. Would you propose we do so?
			 
			[JB> ] Yes, but I don't think you would need to include support for marked up content as generated content. Certainly in the soft layout scenario, I perceive a requirement for the generation of certain simple content as a result of UA decisions (specifically line breaking decisions)
			 
			Examples:
			 
			Original content is this. It is music lyrics - so by convention would have a # character prefix in UK subtitling / captioning market
			Scooby dooby doo where are you?
			

			
			So broken onto two lines it might look like this
			# Scooby dooby doo
			# where are you
			 
			BUT on one line it would look like this
			# Scooby dooby doo where are you?

			
			In the US (musical note not very representative of real glyph used)
			broken onto two lines it might look like this
			?Scooby dooby doo?
			?where are you?
			 
			BUT on one line it would look like this
			?Scooby dooby doo where are you?
			 
			I'd like to generate this content  - at least in the soft layout model - by application of style - not by inclusion in the content.
			 
			Speaker indication may also use a similar mechanism - but here the characters are included upon change of speaker within the same subtitle. So the following is spoken by two different speakers
			
			- Enjoy the dance last night?
			- Yes, did you?
			 
			>> Enjoy the dance last night? 
			>> Yes, did you?
			 
			A more interesting problem is that IIRC, the before: and after: selectors work on identified elements or ranges.
			Could this be made to work on the 'displayed lines' as created by a soft layout UA?
			 

	regards 
	John Birch 

	The views and opinions expressed are the author's own and do not necessarily 
	reflect the views and opinions of Screen Subtitling Systems Limited.