Created by: thejefflarson
I've added a --loosey
option to csvsql to optionally remove the column constraints.
This is particularly useful for very large csv dumps, where type-hinting can take an extraordinary amount of time. It allows the user to generate the sql with a subset of data like so:
$ head -n 100 some-large-csv.csv | csvsql --loosey --table some-large-csv
We've run into this problem many times with datasets on the order of hundreds of thousands of rows and many columns and having something like this would save time in tweaking the schema after the fact.