It may be (imho) very usable to add the lazy parsing by rows.
Something like this:
table.first()
while not table.eof:
print(table.next())
My mdb-tables is too big and parser eats a lot of memory (
May be something like this already exists in your project...