Delta Sharing Flask Browser

A lightweight Flask web app to explore Delta Sharing data resources — including shares, schemas, tables, metadata, and data previews — using only Python and Pandas (no Spark required). Why I created it? Sometimes it's not that user-friendly to explore what is available in delta shares and having a simple app to browse shares is helpful.

---

Features

• Browse all available Shares, Schemas, and Tables
• View table metadata (schema, version info, file stats)
• Estimate data size for tables via file aggregation
• Preview table data directly as a Pandas DataFrame (first 20 rows)
• Flexible profile management: upload, paste, environment variable, or file
• Works on any system with Python (Mac, Linux, Windows)

---

Requirements

• Python 3.9+
• Delta Sharing profile file (see below)

---

Setup

cd DeltaShareFlask
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

---

Providing Your Delta Sharing Profile

Authenticate using a valid profile file. You can:

• Place your profile at ./profiles/config.share, or
• Use the web UI at to upload/paste the profile JSON (stored as ./profiles/config.share).

Example profile config.share:

{
  "shareCredentialsVersion": 1,
  "bearerToken": "...",
  "endpoint": "https://.../api/2.0/delta-sharing/metastores/...",
  "expirationTime": "9999-12-30T23:59:59.480Z"
}

---

Running the App

python app.py

If 5000 is taken, use

PORT=5050 python app.py

Visit http://127.0.0.1:5000 in your browser to access the app.

---

How It Works

This app uses the Delta Sharing Python library. Key functions:

• SharingClient.list_shares()
  List all shares.
• SharingClient.list_schemas(share)
  List schemas in a share.
• SharingClient.list_tables(share, schema)
  List tables in a schema.
• SharingClient.list_files(table)
  Aggregate file sizes for data size estimation.
• SharingClient.get_table_schema(table)
  Retrieve table columns info.
• SharingClient.get_table_version(table) Current version of the table.
• delta_sharing.load_as_pandas(Table(...))
  Load a shared table into a Pandas DataFrame for preview.

---

Useful Delta Sharing Library Methods Available

You can use these (without Spark):

• list_shares(): List all shares
• list_schemas(share): List schemas within a share
• list_tables(share, schema): List tables within a schema
• get_table_schema(share, schema, table): Get table schema
• get_table_version(share, schema, table): Get table version
• list_files(share, schema, table): Get underlying Delta/Parquet files & estimate table size
• load_as_pandas(Table(...)): Load entire Delta/Parquet table to a Pandas DataFrame
• load_cdf_as_pandas(...): (if available) Change Data Feed to Pandas

See Delta Sharing Python API for more.

---

Additional Notes

• Size estimation uses list_files_in_table or similar API to sum file sizes; may show N/A if unsupported.
• Data preview uses delta_sharing.load_as_pandas limited to 20 rows for performance.
• You can manage multiple profiles conveniently using the UI.

---

Production Deployment

• For production, run with WSGI (e.g., gunicorn) and place behind a reverse proxy.
• Never commit credentials or secrets to source control.

---

License

MIT