Snowflake Spreadsheet Uploader

This a demo showcasing various capabilities of Streamlit applications built on Snowflake:

• Directly upload data from spreadsheets (.csv, .xlsx) into Snowflake tables
• Visualisations
• Data & file validation
• Caching / Memoization

Setup

Prerequisities

You will need the following:

• Snowflake account. We highly recommend creating a trial account over any production accounts.
• Snowflake role with CREATE DATABASE and CREATE TABLE privileges. If using a trial account, the SYSADMIN role is sufficient for this exercise.

If running this on your own machine, you will need either:

• Python3.8
• Docker

Code and deployment has been tested and verified on Gitpod

Please Note

The Streamlit code creates a few objects within your Snowflake account:

• A Database: SNOWFLAKE_LAB_ONLINE_RETAIL
• A Table: SNOWFLAKE_LAB_ONLINE_RETAIL.PUBLIC.TRANSACTIONS

Set up Credentials

This sets up the connection from your Streamlit app to your Snowflake account. Make a copy of the included ./.streamlit/secrets_example.toml file:

cp ./.streamlit/secrets_example.toml ./.streamlit/secrets.toml

Substitute each value with values applicable for your Snowflake demo account

Running your Streamlit Application

Option 1: Python on Docker

Simply build & run the image with the following:

docker build -f Dockerfile -t snowflake-csv-uploader .

# Run
docker run -it \
    -p 8501:8501 \
    -v "$PWD":/app \
    -w /app \
    snowflake-csv-uploader

Assuming everything is run and configured correctly, you will be given a URL link to the app running.

Option 2: Python

(Optional, but recommended) Create a virtualenv

python3 -m venv ./snowflake-csv-uploader-env
source ./snowflake-csv-uploader-env/bin/activate

Install required packages:

pip3 install -r requirements.txt

Run your app

streamlit run streamlit_app.py

Assuming everything is run and configured correctly, you will be given a URL link to the app running