复合组件的资源解析器

Question

Could someone please explain what is responsible in MyFaces to figure out what is to treated as a resource/composite component. I am seeing this strange behavior where html tags like div,li etc are trying to be resolved as composite components. I am seeing the Classloader looking for {mycompositecomponentnamespace}/div.xhtml, {mycompositecomponentnamespace}/li.xhtml etc

In essence, my question is, how does myfaces know which tag needs to be treated as a resource lookup for a composite component, and which is to be ignored as a standard html tag

Answer 1

In few words, the algorithm just do what JSF spec says. If the associated namespace to a library has an composite component resource library, facelets algorithm scan for composite components following JSF 2.0 Resource Handler algorithm. In JSF 2.2, things were changed to include resource library contracts too. How MyFaces do it in code is part of its own implementation, but it follows the algorithm proposed on the spec.

Answer 2

MyFaces2.0.2 and 2.0.3 when trying to parse a Facelet using a SAXParser does try to look up each tag as a resource in the Classloader. I would have thought it would have the smarts to realize what needs to be looked up and what is a standard HTML tag.