Return to Answer

If you’re keeping the updates into a SQLite backend, you could turn the weekly update into a new table and compare it to the archived data with queries, before merging it.

Example of using SQL to find new additions to a table: https://stackoverflow.com/questions/2077807/sql-query-to-return-differences-between-two-tables

If a field in your DB stores the date of the transaction, you could just query all users who have had transactions in the past 18 months. Then the archive is just the full DB. Alternatively, you could query all users who haven’t, extract their data, then drop them. Updates are just any rows timestamped this week.

If you’re keeping the updates in a SQLite backend anyway, you could turn the weekly update into a new table and compare it to the archived data with queries, before merging it.

Example of using SQL to find new additions to a table: https://stackoverflow.com/questions/2077807/sql-query-to-return-differences-between-two-tables

If a field in your DB stores the date of the transaction, you could just query all users who have had transactions in the past 18 months. Then the archive is just the full DB. Alternatively, you could query all users who haven’t, extract their data, then drop them. Updates are just any rows timestamped this week.

If you’re keeping the updates into a SQLite backend, you could turn the weekly update into a new table and compare it to the archived data with queries, before merging it. (Example of using SQL to find differences between two tables: https://stackoverflow.com/questions/2077807/sql-query-to-return-differences-between-two-tables)

If you’re keeping the updates into a SQLite backend, you could turn the weekly update into a new table and compare it to the archived data with queries, before merging it.

Example of using SQL to find new additions to a table: https://stackoverflow.com/questions/2077807/sql-query-to-return-differences-between-two-tables

If a field in your DB stores the date of the transaction, you could just query all users who have had transactions in the past 18 months. Then the archive is just the full DB. Alternatively, you could query all users who haven’t, extract their data, then drop them. Updates are just any rows timestamped this week.

Mistakenly posted the answer to another question in this window. Deleted.

If you’re keeping the updates into a SQLite backend, you could turn the weekly update into a new table and compare it to the archived data with queries, before merging it. (Example of using SQL to find differences between two tables: https://stackoverflow.com/questions/2077807/sql-query-to-return-differences-between-two-tables)