Streamlit ERD Viewer

Simple Entity-Relationship diagram generator and viewer for any database schema from your Snowflake account.

Can be installed right away as a Streamlit App in your Snowflake account. Or tested locally as a Streamlit web application.

ERD Viewer

Select one database and schema, change the theme and adjust the display options. Here is a typical diagram with table columns and PK-FK relationships between the tables:

Can show table entities only, with no column names:

Or full information about each table, with column names and their data types:

Script Generator

It generates as well the create DDL script:

Snowflake Connection

To connect locally from the web app to your Snowflake account:

• Install SnowSQL.
• Locate the ~/.snowsql/config file.
• Add a [connection.my_conn] section with the Snowflake accountname, username and password.

Local Streamlit App

To test and run the Streamlit application locally, on Windows, in VSCode or your own Python IDE:

• Clone this repository on your local machine, then open the IDE with the folder of this project. Make sure you have installed a Python version supported by Snowpark (v8, v9, or v10 at this moment).
• In a Terminal window, create a virtual environment with python -m venv venv. Switch to this new environment with venv/scripts/activate.
• Run pip install -r requirements.txt, to install the Snowpark and Streamlit dependencies.
• Call streamlit run app.py to open the web app in your browser. Type CTRL+C in the Terminal to stop the web app.

Streamlit App in Snowflake

To install and run this as a Streamlit app in your Snowflake account:

• To create an empty Chinook sample database, paste and run the scripts/chinook.sql script in a new worksheet. Or run snowsql -c my_conn -f scripts/chinook.sql from a Terminal window.
• In deploy.sql, change with the path to your app.py file. Then run snowsql -c my_conn -f scripts/deploy.sql, which will (re)create a STREAMLIT_ERD_VIEWER database, will load the Python file in a stage and will create a Streamlit app from this file. As an alternative, in a new database (and schema), create a new Streamlit app with the default template. Then replace the Python code with the full content of our app.py file.
• Test and run the Streamlit app online. Share a link to this app with other roles or users.
• To cleanup, drop the database with this Streamlit object. Or run snowsql -c my_conn -f scripts/cleanup.sql from a Terminal.