Some CSV files have issues with them. Below is a code snippet to save the information in a csv file into a duckdb database, ignoring all the ‘broken’ rows.
# pip install duckdb
import duckdb
# create context manager to connect to duckdb database
# a duckdb database will be created if there is not one already
# with that name / filepath
with duckdb.connect("database.duckdb") as con:
# create table statement
# rows with errors in will not be loaded into duckdb database
make_table = '''
CREATE TABLE raw_bronze AS
SELECT
*
FROM read_csv('my_faulty.csv', ignore_errors=True);
'''
con.sql(make_table)
# context manager will handle the closing of the connection
# to duckdb database so no need to close it explicitly