Implement Schema Migration API endpoint

[dmitrizagidulin]

One of the major user complaints with Riak Search (and Solr in general) is the inability to re-index existing documents when changing a Search schema.
The only way to perform schema migrations currently (to re-index all existing documents) is to manually write scripts to "touch" every object in an index (that is, read an object and write it back immediately) -- scripts that are usually powered by a streaming list keys or "select all documents" query.

Given that Explorer keeps cached lists of keys (that can be refreshed via either streaming list keys call or, better yet, something like issue #129 - key list via search query or issue #127 - throttled background key listing), we would make users' lives dramatically easier if we provided a push-button ability to do schema migration:

1. Generate cached key lists for all the buckets in a given bucket type that has a search index, if not yet present.
2. Set up a long-running Explorer job (throttled, connection-pooled and multi-threaded) to iterate over those key lists and "touch" the objects (read-then-write).
3. Since the key lists are generated beforehand, provide polling capability to determine job progress on the client side.

Possibly only implement this in Dev mode. (Needs discussion, though).