Enable the automatic generation of a number of different sitemap and related mechanisms from existing HTML navigation constructs by specifying a minimal amount of semantic information.
There is currently no way to encode in a machine readable way which navigation link elements pointing to locations in a site should be considered for inclusion in a sitemap.xml, cache manifest or related artifact, additionally information is not currently available which would enable specification of priority weighting or frequency of change.
Currently accurate sitemap.xml files, HTML sitemaps and cache manifests are written manually, or created by supplying information generator services/sites. This is rather disconnected from our sites, and not always as accurate as we would like.
Examples of what could be generated if we had such information:
- HTML sitemap for usability/accessibility
- HTML5 cache manifest
Countless sites contain HTML navigation constructs which already detail all/most of the information required to build an HTML sitemap or cache manifest, and parts of the information required to build a sitemap.xml file.
Typically these constructs leverage link elements which conveniently expose an 'href' attribute. Add a way to specify 'priority' and maybe 'changefreq' and enough would be in place for machines to build a sitemap.xml.
Useful metadata could be generated by machines or browser plugins with both leveraging a simple microformat.
Per the microformats The microformats process: