Faker for Snowflake!
This repository originally described a Snowflake External Function which invoked the Faker python library in an AWS Lambda function, directly from a Snowflake query. This older codebase and information can be found under the 'external_function' branch of this repo.
As of mid 2022, Snowflake now offer Python support natively on platform. This means that the new way to invoke Faker is as simple as:
create or replace function FAKE(locale varchar,provider varchar,parameters variant)
returns variant
language python
volatile
runtime_version = '3.8'
packages = ('faker','simplejson')
handler = 'fake'
as
$$
import simplejson as json
from faker import Faker
def fake(locale,provider,parameters):
if type(parameters).__name__=='sqlNullWrapper':
parameters = {}
fake = Faker(locale=locale)
return json.loads(json.dumps(fake.format(formatter=provider,**parameters), default=str))
$$;
You can use pretty much any of the standard providers listed here.
It scales really nicely, meaning you can generate large quantities of fake data in seconds.
Read the new Medium article for further info and examples.
- James Weakley (Omnata)
Generate 5000 fake names, in the US English locale:
select FAKE('en_US','name',{}) as FAKE_NAME
from table(generator(rowcount => 5000))
100 dates between 180 days ago and today::
select FAKE('en_US','date_between',{'start_date':'-180d','end_date':'today'})::date as FAKE_DATE
from table(generator(rowcount => 50));
Generate 100 fake profile attributes (name, job, company), and extract them from the returned variant column:
with FAKE_PROFILES as (
select FAKE('en_AU','profile',{'fields':'name,job,company'}) as FAKE_PROFILE
from table(generator(rowcount => 100))
)
select FAKE_PROFILE:company::varchar,FAKE_PROFILE:job::varchar,FAKE_PROFILE:name::varchar
from FAKE_PROFILES
