What are the tradeoffs of reusing a cursor vs. creating a new cursor?

In cx_Oracle (or O开发者_开发知识库racle in general), is it possible to allocate a cursor for each query, or to reuse a cursor across several queries.

def getSomeData(curs):         # case 1: pass in a cursor, which is generally
    curs.execute('select ...') #         reused across queries
    return curs.fetchall()

def getSomeData(conn):         # case 2: pass in a connection,allocate
    curs=conn.cursor()         #         a cursor for this query
    curs.execute('select ...')
    return curs.fetchall()

Of course, both approaches return the same data.

What are the tradeoffs between the two approaches? Is one particularly more or less efficient? Are there any potential pitfalls to reusing a cursor over many queries?

You can re-use a cx_Oracle cursor as much as you like, no problem. If you're executing thousands of small queries in a small space of time, you might see a slight performance improvement by re-using the cursor, but I doubt it.

I will sometimes create new cursors, and other times re-use an existing one, depending on whether it makes the code easier to read and understand.

For example, if I have a variety of procedures that need to access the database, I might pass an Oracle connection object around, or a cursor created on that connection.

Whatever makes your code more readable and easier to maintain is what I'd go for.