This is, as I see it, two problems. The first one is to make the the domain point to a specific root in the path. That is if we have multiple domains served by the same CQ instance. Then we need to map the content root to the domain, both forward and reverse mappings. That is done something like this for the domain/site about.mysite.org that is stored in CQ under /content/mysite/en (Not actually tested).
Forward <?xml version="1.0" encoding="UTF-8"?> <jcr:root xmlns:sling="http://sling.apache.org/jcr/sling/1.0" xmlns:cq="http://www.day.com/jcr/cq/1.0" xmlns:jcr="http://www.jcp.org/jcr/1.0" jcr:mixinTypes="[cq:ReplicationStatus]" jcr:primaryType="sling:Mapping" sling:internalRedirect="[/content/mysite/$2.html$3]" sling:match="(.*)about\.mysite\.org\.[0-9]*((?!(/)).*)\.html([\?|/|#].+)?$"/> Reverse <?xml version="1.0" encoding="UTF-8"?> <jcr:root xmlns:sling="http://sling.apache.org/jcr/sling/1.0" xmlns:cq="http://www.day.com/jcr/cq/1.0" xmlns:jcr="http://www.jcp.org/jcr/1.0" jcr:mixinTypes="[cq:ReplicationStatus]" jcr:primaryType="sling:Mapping" sling:internalRedirect="[^/content/mysite/]" sling:match="/"/>
Some explanation on the regular expression in the forward. We use the (.*) infront of the about part of the domain name, since then we can use it on all environments, like qa-about and stage-about. That is the first capture group, $1, that we don't use.
The next one is the .[0-9]*. That is when we use different ports. Then comes the ((?!(/)).*) . This is the magic second capture group, $2, that catches the path. This one we append in the internal redirect. The last capture group is ([\?|/|#].+) . This one grabs and forwards the query parameters or sling parameters. In this example, I only allow this to be done on the ,html extension. Everything else just uses the normal path.
The reverse is much easier. I map a specific path, the /content/mysite, to the /. Maybe you can make the reverse mapping append your domain, but I don't think so...
This is the first part of your question.
The second one is a bit trickier... All the output of image URLs are done relative. The jsp parser does not know that we want all images to start with static.mysite.org. We have two options now. Either we make our own, domain aware, adaptive image, and have that append our static domain to all images. Or we make our own adaptive image, and use the externalize api. I would use the later, since then I can have non-static paths on the author environment and static on the publish.
Hope that this gives you some more info.
/Ove
