RE: Use of SVG to describe EPUB region

>    <glyphs x="10 20 30" y="0" glyphids="3 1 2">abcd</glyphs>

Did we discuss that an n-to-m mapping dictionary would take care of this?
If that information is available the markup would become:
   <glyph-text x="10 20 30">3 1 2</ glyph-text>

> I took an action to think about this a bit more, but I would like the
> mechanism for that correspondence to be the same as for the feature that
> would allow bespoke graphics that are intended to be viewed as text (I think
> is the kind of use case you are referring to).  We didn't exactly discuss last
> week, but I think such a feature is important to have.  I haven't come up with
> a satisfying syntax for it yet however.

PDF already solved that problem. See chapter 10.8 in http://wwwimages.adobe.com/www.adobe.com/content/dam/Adobe/en/devnet/pdf/pdfs/PDF32000_2008.pdf
I see no reason that we can't reuse some of the same concepts (ie replacement text)

Rik

> -----Original Message-----
> From: Cameron McCormack [mailto:cam@mcc.id.au]
> Sent: Sunday, February 10, 2013 7:46 PM
> To: Richard Schwerdtfeger
> Cc: FUJISAWA Jun; public-svg-wg@w3.org
> Subject: Re: Use of SVG to describe EPUB region
> 
> On 7/02/13 1:19 AM, Richard Schwerdtfeger wrote:
> > I think what is also important is for ePub to use SVG 2.0. It would be
> > nice to state that in the content the role of this "bubble" is a
> > tooltip or if we define a better taxonomy for graphics in the future a
> "bubble".
> > These examples also require a means to expose the textual description.
> > as a scree reader user would be unable to process what is in the
> > bubbles, unless of course SVG text is used. Is there a plan to use SVG
> > text elements in the bubbles or will the text be drawn using strokes, etc.?
> 
> During the F2F last week, we discussed what might needed for a feature that
> allows specific glyphs from a font to be rendered, and to have character data
> in the document to be mapped to those glyphs.  Something
> like:
> 
>    <glyphs x="10 20 30" y="0" glyphids="3 1 2">abcd</glyphs>
> 
> has the problem that, if you are performing interactive selection of the three
> glyphs, you do not know how to map that to the "abcd" text that the
> element is claiming it corresponds to.
> 
> I took an action to think about this a bit more, but I would like the
> mechanism for that correspondence to be the same as for the feature that
> would allow bespoke graphics that are intended to be viewed as text (I think
> is the kind of use case you are referring to).  We didn't exactly discuss last
> week, but I think such a feature is important to have.  I haven't come up with
> a satisfying syntax for it yet however.

Received on Tuesday, 12 February 2013 18:38:01 UTC