How do I get pandas.read_csv to interpret a null entry as integer 0?

Question

With pandas.read_csv, I am reading a CSV which includes null entries, e.g.,

col1,col2,col3
2,3,4
8,,9

read_csv interprets the null entry as a flo