200 compound indexing

[michael-barbour]

closes #200

This should handle the active and passive sync between chemcurator_django and resolver.

• [x] Actively syncs unassociated compounds with resolver
• [x] Passively syncs newly added compounds with resolver
• [x] Passively syncs when a substance's association with a compound changes
• [x] SID CID uniqueness enforcement

Active sync:

To test this run manage.py sync. This will delete all records then add all substances and any solo compounds into resolver.

Passive Sync:

There are a stupidly high number of cases for this one. Broken down into rules, it should look something like this:

• "Paired" substances should have their substance and compound data represented in the identifiers. The ID will be the SID.
• "Solo" substances should be represented.
• "Solo" compounds will be represented where their CIDs as the ID and their compound data in identifiers.
• If a solo compound is associated to a substance, the solo compound record will be deleted and the newly paired substance will be updated.
• If a paired substance is "disassociated", the solo substance will be updated and a solo compound record will be added.
• Solo compound updates will update the solo compound record.
• Paired compound updates will update the paired substance record.

Constraint:

Saving a compound or substance with an already used sid/cid will result in an error. This validation only runs on create and not on update.