That's right, migra does not compare dump outputs. I've found that the best parser of PostgreSQL dump files is PostgreSQL, and don't want to reimplement that.

That's right, migra does not compare dump outputs. I've found that the best parser of PostgreSQL dump files is PostgreSQL, and don't want to reimplement that. Hence, migra works directly on database connections. If you have a dump file you want to compare, you can just create a temporary database and load the dump into that first. There's an example in the docs of how to do this easily.