Solution for Initially Populating Solr

[crantila]

When Abbot is first installed on a server, HolyOrders loads Drupal data into the empty Solr instance. I didn't plan it that way, but it mostly works, and all the other solutions required manual intervention, so we'll go with this.

However, if this is our strategy for getting data into Solr, we need to be sure that it works. In particular, I'm concerned that not all the Chant resources will be picked up. Is there a better solution for this than doing a day-by-day update for every day since the Drupal database was created?

[crantila]

Idea: have Drupal export a list of the chant IDs contained within a particular source. So the initial data population will look like this:

1. Load the normal resource types, each type at once.
2. For each "source" resource:
   1. Ask Drupal for the IDs of the chants in that "source."
   2. Load the chants, one at a time, populating the database for issue HolyOrders: update robustness and failure safety #66 along the way.
3. Run the usual "which chants were updated today and yesterday?" style updated, to pick up any chants that were entered while we were doing the for-each thing.

[crantila]

Idea 2: HolyOrders will know to do the "initial population" thing for a resource type when its "most recent update" according to the updates DB (created for issue #58) is set to never.