Create sql_serializer options for filtering data that gets serialized to allow for creation of smaller HAF databases
A fully flexible HAF server should keep all the blockchain data so that it can run any potential HAF application. But in many instances, a HAF server operator will only want run to a few specific HAF applications, in which case they should be able to configure the sql_serializer feeding their database to only write the data they need.
The first such example case we should concentrate on is the implementation of filtering that supports running a hafah app that only records account history for specific accounts. This will be a common usecase for exchanges, for example.
In this case, we will want to filter at least the largest tables created by sql_serializer (operations, account_operations, and transactions) which completely dominate the size of the base storage of a HAF database. We can probably stick to the same filtering specification used by the account history plugin: a list of account ranges, where a range is a lexicographic-specified range of strings.