- From: Kipp Howard <khoward@courtlink.com>
- Date: Fri, 16 Nov 2001 08:10:51 -0800
- To: html-tidy@w3.org
Hans Deragon [mailto:deragon@aqiii.org] wrote: > I want to build a template driven system for processing html files. > New tags are placeholders and instructions for how to manipulate the > html files. > > And I want these tags to be in xml format. I then pass the > html files > through jtidy > to make them xml compliant. Then, using DOM functionnality, I can > manipulate > easily the different html and resolve their dependencies. I > then create > a new > XML structure resulting from the merging of many files, and use this > structure > to generate the final static page (I might pass this final > static page > again through > jtidy). > > Its not the most efficient way of doing, but I am not seeking most > efficiency CPU > or memory wise. I am seeking for the most efficient use of existing > code, keeping > maintenance easy, etc... Using jtidy, I have no parsing to do! It > provides me > the DOM document of my HTML pages (hopefully with my tags > intact). Then > I only have to play with the Nodes, replacing them, deleting > them, etc.. > to get > the final structure. FWIW: I haven't necessarily done the above but we do add some tags to HTML files that give us some additional information about the files. We do this by adding <div> and <span> tags with very specific attributes so that these <div> and <span> tags are not confused with existing tags. This means we do not have to have any new tags that tidy does not support. Here is a short example: <div title="myConfigInfo"> <span title="color">red</span> <span title="size">medium</span> </div> Then when we are using the DOM/XSLT, we just select these elements with: //div[@title='myConfigInfo]/span[@title="color"] Just thought this might help make your job a bit easier by not having to modify how tidy works.
Received on Friday, 16 November 2001 11:11:59 UTC