Working with chunked files through the REST API

[David Chandek-Stark]

IMO It would be really helpful if the server could handle requests through the REST API more elegantly in the case of chunked content.

Currently the server returns a 404 on a GET of the content id of the chunked content -- because there is no actual file there -- but this response is somewhat misleading, since the content id has actually been used.  A status in the 3XX range (maybe 303 or 300) with a Location header would be more useful and also preclude unnecessary manual checks for the existence of a content manifest.

Additionally, it appears that there is no "direct" way to retrieve the "stitched" content through the REST API. Using the Retrieval tool is pretty inconvenient for single content ids (it doesn't seem designed with that case in mind). Manually "stitching" the chunked file is not terribly onerous, but it requires every client to implement some method of doing so, and it is at least a non-trivial operation (reading the manifest, getting all the chunks, verifying the full checksum).

What do you think?

Thanks,

David