We shouldn't block on missing values, as that might lead to very large blocks.
We should only block on known data.
Here we introduce a 'null-sentinel' which represent the string that stands for the NULL value. Defaults to the empty string, as that is most common in csv files.
We shouldn't block on missing values, as that might lead to very large blocks. We should only block on known data.
Here we introduce a 'null-sentinel' which represent the string that stands for the NULL value. Defaults to the empty string, as that is most common in csv files.