Treenode is a project mainly written in JavaScript, it's free.
Assemble HTML page from multiple sources
(experimental idea)
Given a meta-template which structures a HTML page in tree structure, fetch contents from local/remote service which only generates portion of HTML.
{ 'Document' :
[
{'Head' : { 'type' : 'http', 'url' : '192.1.1.10/gethead', 'insertHeader' : 1, 'insertPost' : 0 }
{'Body' : { 'type' : 'http', 'url' : '192.2.1.10/getbody', 'insertHeader' : 0, 'insertPost' : 1 }
{'Footer' : { 'type' : 'memcached', 'server': '192.5.5.0', 'key' : 'footer_version_2.1' }
{'Footer2' : { 'type' : 'local', 'path' : '/treenode/files/somefile' }
],
'prepend' : '<html>',
'append' : '</html>'
}Then,
<head>blah</head><body>hello</body>footer_version_2.1/treenode/files/somefileAfter that, the treenode will simply concat the result and return.
If there was an error(or didn't respond within X sec), template should define how to resolve it. Such as just ignore it, display alternertive text(error box, place holder), or just refuse to render whole thing and redirect to err page.
Because node.js could do all these asynchronously and some HTML pages are getting so complex, it might make sense to do this on complex pages, instead of having a single webserver always generating fully rendered HTML.
Fun weekend project :)