Information exposure about contained resources

Given resources:

/foo/
/foo/bar
/foo/baz

where Agent has:

• Read access to /foo/
• Read access to /foo/bar
• No read access to /foo/baz

NSS responds to authorized GET requests to /foo/ which includes statements about /foo/baz (besides containment statements) eg:

<baz>
    html:Resource, ldp:Resource;
    terms:modified "2021-02-05T10:49:07Z"^^XML:dateTime;
    st:mtime 1612522147.974;
    st:size 12.

It'd be preferable to not expose types (besides indicating it is a container or a non-container eg. html:Resource), modified, mtime, size information about /foo/baz since agent is not granted read access to/foo/baz.

See proposed security consideration: https://github.com/solid/specification/pull/228

You may want to consider performance / response times if you follow-up on this change.

Start by tracing the authorized GET handling for the /foo/ container and the response serialization that exposes metadata for contained resources. Verify how read access is checked for /foo/baz, then ensure inaccessible resources retain only the allowed containment indication and do not expose types, timestamps, or size; check response-time impact as noted in the issue.