Sometimes I'd like to get row values with columnname indexes, sometimes by iterating over the row. The default (not setting the connection's row_factory) is to return a tuple, which gives the latter, but not the former. Setting row_factory to sqlite3.Row gives the former, but (surprisingly) not the latter. It's a surprise because back when pysqlite was not part of the Python distrib (I don't recall what version), its Row type was iterable. Any chance that bit of functionality could be put back?
