Re: Identifying print page numbers in HTML documents

>  > CSS will probably be extended with more selectors but the issue of >
> readyness is important so using CLASS, NAME or ID is better than REL.  >
> 
> Does CSS provide a mechanism for matching, as a selector, any name
> attribute with a value of the form "IPP-[0-9]+"?

Attribute value exact matching is in the oven for the next version of
CSS. Attribute value pattern matching hasn't been discussed I think.

> Otherwise, class
> would be needed.

CLASS or ID value are possible selector today but no pattern matching
either.

One possibility is:

  <A CLASS=IPP ID=IPP-17 NAME=IPP-17>

That would allow CSS selection as in

 A.IPP {...}

or even
 A#IPP-17 {...}

Link could point at them using
 <A HREF=IPP-17>Go to Print Page 17</A>

Note that I used both NAME and ID in the target A because although
Cougar now supports ID as HREF, only NAME were supported before.

To elaborate on link A, we could have

 <A HREF=IPP-17 CLASS=L-IPP ID=L-IPP-17 TITLE="Link to Print Page 17">
	Go to Print Page 17</A>

Note that I used a different CLASS name, so that CSS can use different
selector for different rendering of link vs. target anchor.
(this is what was achieved with REL and REV in Murray's proposal)

I actually think using REL and REV is a better fit in theory, but CSS
of today does not support using them as selector, so it's another of
those "what is the timeframe of our solution?" issue. In this case
though, we could recommend both methods (CLASS/ID naming + use of
REL/REV), as the original IPP/BPP proposal recommends.

 
> Also, the format of the page number could be extended to permit a single
> letter prefix to precede the page number. Some documents, an example of
> which I am reading at the moment, contain multiple series of page numbers
> which need to be kept distinct. Thus, page numbers such as a31, h18 etc.,
> should be allowed, and similarly for Roman numerals, provided that one is
> careful in one's choice of available letters so as to avoid ambiguity.

Murray's proposal had IPP-[0-9]+" and "IPP-[ivxlc]", but in general I
think the convention should mention IPP (and BPP) as being the "class"
and the rest is free use (numeral, roman, letter).

Received on Thursday, 3 July 1997 07:04:55 UTC