fakesnow

library0.11.9pypypi✓ verified 90d ago

Fake Snowflake Connector for Python. Run, mock and test Snowflake DB locally using DuckDB. Current version: 0.11.6. Release cadence: approximately monthly.

database testing aws

pip install fakesnow

github ↗package ↗changelog ↗

INSTALL

IMPORT

SIG · FAKESNOW

fakesnow

databasepythonv0.11.9

Install

10.0s avg

Import

2618ms

Disk

309MB

Pass rate

5/ 10

Env Coverage5 / 10

glibc

3.9–3.13

musl

3.9–3.13

Install & Compatibility

Where this runs

tested against v0.11.9 · pip install

no network on importno background threads

Install × environment matrix

Each cell = how many times install + import succeeded across repeated harness runs. Partial = flaky.
glibc = Debian/Ubuntu slim · musl = Alpine Linux

musl

py 3.10–3.910 runs

build_error

glibc

py 3.10–3.910 runs

installs and imports cleanly · install 10.0s · import 2.618s · 309MB

309MB installed

● package 309MB

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

fakesnow

✓ import fakesnow

✗ from fakesnow import connect

fakesnow does not expose a connect function; you use snowflake.connector.connect after fakesnow.patch()

fakesnow.patch

✓ from fakesnow import patch; patch()

✗ fakesnow.patch()

If you do 'import fakesnow' then you must call fakesnow.patch(), but it is clearer to import patch directly

snowflake.connector.connect

✓ import snowflake.connector; conn = snowflake.connector.connect(...)

✗ import fakesnow; conn = fakesnow.connect(...)

After fakesnow.patch(), snowflake.connector.connect is replaced with a fake version. Don't use fakesnow.connect.

Patch snowflake.connector, create a fake connection, run SQL locally.

import fakesnow import snowflake.connector fakesnow.patch() conn = snowflake.connector.connect( user='test', password='test', account='test', database='MYDB', schema='PUBLIC', warehouse='test' ) cur = conn.cursor() cur.execute('CREATE TABLE users (id INT, name STRING)') cur.execute("INSERT INTO users VALUES (1, 'Alice')") cur.execute('SELECT * FROM users') print(cur.fetchall())

Debug

Known issues

breakingPython 3.9 support dropped in v0.11.0

fix

Upgrade to Python >=3.10 or pin fakesnow <0.11.0

affects: >=0.11.0

breakingInformation schema column names now uppercase (v0.10.0)

fix

Update queries to use uppercase column names (e.g., 'TABLE_NAME' instead of 'table_name') or wrap in double quotes

affects: >=0.10.0

gotchafakesnow does NOT mock all Snowflake features. Unsupported SQL may raise errors or produce incorrect results.

fix

Check GitHub issues/README for known limitations before relying on fakesnow for complex queries.

affects: all

gotchaAlways call fakesnow.patch() BEFORE importing snowflake.connector if possible, or at least before any connection attempt.

fix

Run 'fakesnow.patch()' at the top of your test file or conftest.py, before any snowflake imports.

affects: all

Errors

Common errors & fixes

ModuleNotFoundError: No module named 'snowflake'

Missing snowflake-connector-python dependency

fix

pip install snowflake-connector-python

fakesnow.patch() not found

Incorrect import; fakesnow does not export patch if imported incorrectly

fix

Use 'from fakesnow import patch' or 'import fakesnow; fakesnow.patch()'

snowflake.connector.errors.DatabaseError: 000404 (42S02): Table ... does not exist

Unsupported SQL feature in fakesnow (e.g., certain DDL variations)

fix

Simplify the statement or check fakesnow's supported SQL subset (see GitHub README).

Upgrade

Version history

0.11.9latest on PyPI · released Jun 13, 2026

Audit

Dependencies

snowflake-connector-pythonrequiredfakesnow patches snowflake.connector.connect to return a fake connection, so snowflake-connector-python must be installed

duckdbrequiredUnderlying in-memory database engine used for SQL execution

Agent activity

18 hits · last 30 days

node

Resources

githubhttps://github.com/tekumara/fakesnow ↗

packagehttps://pypi.org/project/fakesnow/ ↗

changeloghttps://github.com/tekumara/fakesnow/blob/main/CHANGELOG.md ↗