Flakepipe

Flakepipe is a lightweight, reusable Python module for uploading datasets to Snowflake via secure staging.

It automates table creation, column normalization, and data ingestion using Snowflake’s PUT and COPY INTO commands – making it ideal for ETL pipelines, scraping workflows, and reproducible data transfers.

---

📦 Features

• Upload CSV files to Snowflake via PUT and COPY INTO
• Automatically create or truncate target tables
• Normalize column names to be Snowflake-compatible
• Support for prefix/postfix-based table naming
• Modular, testable, and production-friendly design

---

⚙️ Installation

Supported Python Versions

Flakepipe supports Python ≥ 3.8, and < 3.12.

Installing Flakepipe

To install globally via PyPI (recommended):

pip install flakepipe

To install locally from source (for development), use editable mode:

git clone https://github.com/geeone/flakepipe.git
cd flakepipe
pip install -e .

Alternatively, you can install only the dependencies (without installing the package itself):

pip install -r requirements.txt

💡 See the Dependencies section for version details.

---

🧩 Configuration

You must provide a config dictionary with Snowflake connection parameters:

config = {
    'SF_USER': 'your_user',
    'SF_PASSWORD': 'your_password',
    'SF_ACCOUNT': 'your_account',
    'SF_ROLE': 'your_role',
    'SF_DATABASE': 'your_database',
    'SF_WAREHOUSE': 'your_warehouse',
}

---

🚀 Usage

Upload a CSV file

from flakepipe.uploader import upload_csv_to_snowflake
from your_project.config import config

upload_csv_to_snowflake(
    config=config,
    file_path='output.csv',
    schema='MY_SCHEMA',
    table_name='my_table',
    truncate=True,
    overwrite=False,
    prefix='daily',
    postfix='20250705'
)

Truncate a table manually

from flakepipe.uploader import truncate_table
from flakepipe.connection import get_engine

engine = get_engine(config, schema='MY_SCHEMA')
truncate_table(engine, 'MY_TABLE')
engine.dispose()

---

🛠 Dependencies

Listed in requirements.txt, including:

• SQLAlchemy >= 1.4.15
• snowflake-connector-python >= 2.8.1
• snowflake-sqlalchemy >= 1.2.4
• pandas >= 1.4.0, < 2.1
• numpy>=1.21.0, <1.26.0

✅ Tested with pandas 1.4.x to 1.5.x
⚠️ Compatibility with pandas 2.x is expected but not officially guaranteed

---

📄 License

This project is licensed under the MIT License.

---

🤝 Contributing

Contributions, suggestions, and issues are welcome!
Please see CONTRIBUTING.md for guidelines.

---

🌍 Why Flakepipe?

Flakepipe was born from repeated use in real-world data workflows, especially in scraping and ETL projects.
Instead of rewriting boilerplate Snowflake ingestion code, this tool wraps best practices into a portable, documented, and open solution – freely available under an MIT license.

Whether you’re building one-off scripts or production pipelines – Flakepipe helps you do it right.