Feature: paginated results

[rjzak]

Build a mechanism for search results to be paginated, or continued, so a client may get all the data available.

Thoughts:

• The search would return the results.
• The search would call count(*) on the search to get the total available.
• If the total is greater than what's returned, generate a random search identifier to store the search and the last file ID, so the next could follow-up from that file. Also store the user ID.
• Follow-up query would just be for this random search identifier. Ensure user's ID matches the ID of the creator of the paginated search.