SQLAlchemy cursors: AttributeError: 'tuple' object has no attribute 'items'

Question

I'm attempting to write an iterator to take in a query and export rows in batches as a list of dicts. Here's a complete example of what I'm working with:

class QueryStream(collections.Iterator):
    def __init__(self, conn_details, query, max_rows=None, batch_size=2000):
        # Initialize vars.
        self.engine = dst.get_connection(conn_details)
        self.query = query
        self.max_rows = max_rows
        self.batch_size = batch_size
        self.fetched_rows = 0

        # Create a database cursor from query.
        self.conn = self.engine.raw_connection()
        self.cursor = self.conn.cursor()
        self.cursor.execute(self.query)

    def next(self):
        try:
            if self.max_rows:
                if self.max_rows <= self.fetched_rows:
                    # Maximum rows has been fetched, so stop iterating.
                    raise StopIteration
                elif self.max_rows <= self.batch_size:
                    # Max rowset is small enough to be done in one batch.
                    batch_size = self.max_rows
                elif self.max_rows - self.fetched_rows < self.batch_size:
                    # On the final batch, must fetch only remaining rows.
                    batch_size = self.max_rows - self.fetched_rows
            else:
                # Get default batch size.
                batch_size = self.batch_size

            batch = [dict(row.items()) for row in self.cursor.fetchmany(batch_size)]
            if len(batch):  # 0 rows were returned, so we're probs at the end.
                self.fetched_rows += len(batch)
                print('Fetch {} rows so far.'.format(repr(self.fetched_rows)), file=sys.stderr)
                return batch
            else:
                raise StopIteration
        except StopIteration:
            self.cursor.close()
            self.conn.close()
            self.engine.dispose()
            raise StopIteration

The problem is this line:

batch = [dict(row.items()) for row in self.cursor.fetchmany(batch_size)]

throws this error:

AttributeError: 'tuple' object has no attribute 'items'

I was under the impression that a RowProxy object would be returned for every result in the result set, which supports dict-like operations (as is mentioned in this post), but it appears the results are just plain tuples. The SQLAlchemy docs are not 100% clear on what the expected type of the results from a cursor is, only offering this example of usage.

Question: Am I doing something wrong with my cursor usage? I need the results to be a list of dicts with column names as keys, but I don't know if that's possible without having RowProxy instead of tuples.

score 0 · Accepted Answer · answered Jul 31 '19 at 18:40

Well, it seems the problem has to do with using a raw connection:

        # Create a database cursor from query.
        self.conn = self.engine.raw_connection()
        self.cursor = self.conn.cursor()
        self.cursor.execute(self.query)

        ...

        batch = [dict(row.items()) for row in self.cursor.fetchmany(batch_size)]

was replaced with

        # Create a database cursor from query.
        self.engine = dst.get_connection(workspace_uuid, project_id)
        self.stream = self.engine.execution_options(stream_results=True).execute(self.query)

        ...

        batch = [dict(row.items()) for row in self.stream.fetchmany(batch_size)]

And all is well. I'm fortunately using a driver that supports stream_results (psycopg2):

stream_results – Available on: Connection, statement. Indicate to the dialect that results should be “streamed” and not pre-buffered, if possible. This is a limitation of many DBAPIs. The flag is currently understood only by the psycopg2, mysqldb and pymysql dialects.

SQLAlchemy cursors: AttributeError: 'tuple' object has no attribute 'items'

1 Answers1