Implement CLI

[denisnazarov]

From the CLI google doc by @bigs :

• Given a (house-written) schema translator and a file/some files, a private key, and an “identity” (just a name for now), ingest a dataset into Mediachain

$ mediachain ingest tate ~/tate/artworks cole.crt +cole
Successfully ingested 1000 objects.

• Query for a list of Canonical IDs in the system (paginated)

$ mediachain list --page 1
Page 1 of Canonical IDs:
ffffffffffffffffffffffffff
eeeeeeeeeeeeeeeeeeeeeeeeee
...

• Given a Canonical ID, return the entire revision tree (as JSON--formatting unknown)

$ mediachain get ffffffffffffffffffffffffff
{
  …
}

• Given a blob ID, submit a revision of a piece of Metadata

$ mediachain revise ffffffffffffffffffffffffff updated.json
Updated ffffffffffffffffffffffffff.

• Given a Canonical ID representing a Person, find all works they authored

$ mediachain works dddddddddddddddddddddddddd
Fetching works for John Smith:
ffffffffffffffffffffffffff
Eeeeeeeeeeeeeeeeeeeeeeeeee

• Configure the path to the Mediachain server (dumped into ./.mediachain)

$ mediachain config 127.0.0.1:1337

[parkan]

Needs concrete output specs