- From: Douglas Rand <drand@sgi.com>
- Date: Mon, 13 Jan 1997 09:51:13 -0500
- To: "Joel N. Weber II" <nemo@koa.iolani.honolulu.hi.us>
- CC: www-style@www10.w3.org
Joel N. Weber II wrote: > > I'm trying to write the HTML parser in GNU E-scape to not directly use > the name of the tag wherever possible, and encode all the information in > the style sheet. > > Other than certain exceptional tags, this information can be encoded > almost entirely with CSS1 and with tag attributes. However, there's no > way to know if I have something like an <IMG> tag, for which there is no > corresponding </IMG> tag, or something like <PRE>, where there is a > corresponding </PRE> tag. > > I'd like to add a new property. I'm thinking about calling it > `container', with possible values of `container' and `non-container'. No, this would be a big mistake. The information about what tags require a closing tag, which are implied and which are just elements (what's the SGML nomenclature for this?) is contained in the DTD. No simple declaration in CSS1 will fix this. You need to create a real parser of some kind. You can do pretty well with a few boolean variables per tag, better yet is to use the DTD to understand what tags are valid at what levels so you can perform adequate error recovery. Doug -- Doug Rand <drand@sgi.com> (508) 567 - 2217 Silicon Graphics/Silicon Desktop http://reality.sgi.com/drand Disclaimer: These are my views, SGI's views are in 3D
Received on Monday, 13 January 1997 09:51:31 UTC