Question

A simple example:

import sqlite3, datetime, csv
import pandas.io.sql as sql
from dateutil.parser import parse

my_db = 'test_db.db'

connection=sqlite3.connect(my_db,detect_types=sqlite3.PARSE_DECLTYPES|sqlite3.PARSE_COLNAMES)
cursor = connection.cursor()

cursor.execute('''CREATE TABLE test_table (Id  INTEGER PRIMARY KEY, Date DATE);''')

date_str = '1/1/2011'
date_parsed = parse(date_str)

cursor.execute('INSERT into test_table (Id, Date) values(?,?)',(1,date_parsed))

cursor.execute('SELECT * FROM test_table')

yields:

ValueError: invalid literal for int() with base 10: '01 00:00:00'

I’m simply trying to have the SQL db return my dates in datetime format so I can then perform filtered queries on them.

I’ve already read this related post for reference:

SQLite date storage and conversion

Answer 1

Inside your CREATE TABLE, you’re using Date DATE

CREATE TABLE test_table (Id  INTEGER PRIMARY KEY, Date DATE)

The problem with this is that that tries to map to datetime.date which your format isn’t, if you change your format to use a TIMESTAMP, then it works correctly and tries to make it a datetime.datetime…

CREATE TABLE test_table (Id  INTEGER PRIMARY KEY, Date TIMESTAMP)