Creates Blob Retrieval Endpoint

[gbdubs]

The basic idea here is pretty simple: RMI has indicated to us that metadata is fairly nonsensitive, while the contents of the data (portfolios and reports) is hugely sensitive. My approach to this is to centralize the creation of download URLs in a single place (kind of the opposite of what we did for ADSCI's blob population logic). Additionally, we don't want to populate these proactively because RMI wants audit logs when people actually access the objects. Thus, here's my approach (which this PR starts to implement):

• Have a common download button vue Component, which takes in one or more BlobIDs as the inputs.
• When the download button is clicked, the component starts spinning (loading) it makes a request to our server for a download URL. The button is disabled, spinning during this request. It doesn't block page navigation.
• On the BE, we record (transactionally) audit logs, then generate the download URLs. The ordering ensures our invariant of only granting access when audited. This is the endpoint that this PR implements.
• Then, the download button will transition from spinning into a dial, using fetch and ReadableStream to increase the percentage done as the download (or downloads) complete.
• Finally the button will turn into a check mark and remain disabled to prevent accidental duplicate downloads.

Thoughts?

[bcspragu]

LGTM!