Support for SQLite

[kennethreitz]

Support for SQLite

[kennethreitz]

I haven't used the DB-API directly myself. What would the integration look like?

[MichaelMartinez]

Using the Sqlite standard lib is straightforward: here is some code I wrote while scraping Living Social for my wife. 😈 It is not the prettiest code out there, but it works... and a reason why a simple method to do this with Records will be highly requested.

# Start the DB
def init_db():
    con = sqlite3.connect("Living_Social.db")
    cur = con.cursor()
    cur.execute(
        '''CREATE TABLE IF NOT EXISTS deals (id INTEGER PRIMARY KEY, name TEXT, price_before REAL, price_after REAL, url TEXT UNIQUE, scrapetime text )''')
    con.commit()
    cur.close()
    con.close()


def save_to_db(deal_info):
    con = sqlite3.connect("Living_Social.db")
    cur = con.cursor()
    # Had to use OR IGNORE due scrape pulling down more than one listing
    # Using the URL as the Unique identifier because it contains the deal ID and we need it anyway
    cur.execute("INSERT OR IGNORE INTO deals (name, price_before, price_after, url, scrapetime) VALUES (?,?,?,?,?)", \
                [deal_info['name'], deal_info['price_before'], deal_info['price_after'], deal_info['url'], deal_info['scrapetime']])
    con.commit()
    cur.close()
    con.close()

As far as integration into records is concerned... no idea. 😕

[lucianmarin]

You can simply check DATABASE_URL string and use sqlite3.connect() instead of psycopg2.connect().

[kennethreitz]

It looks like the parameterization style is different. I'd have to take a closer look.

If we're going to support multiple databases, I think using SQLAlchemy as our driver would be the simplest choice.

[kennethreitz]

Currently, this library plans to support Postgres–only. That may change, but not now.

[kennethreitz]

cough support for sqlite has been added!

[kennethreitz]

✨ 🍰 ✨

v0.4.0 includes support for all major databases. Enjoy!