onvistaFinanceDashboard

Local-first financial dashboard built with Python and Streamlit for monitoring a watchlist of onvista instruments.

Current status

This repository currently includes:

• watchlist management in the Streamlit sidebar
• adding instruments by ISIN, WKN, free-text name, or direct onvista URL
• local JSON persistence under data/
• disk caching for onvista resolution and chart data
• local in-app price charts rendered with Plotly
• a dedicated Streamlit settings page for layout and cache settings
• basic automated tests with pytest

The current app is intentionally small and local-first. It is structured so it can later be deployed as a hosted Streamlit app with minimal architecture changes.

Tech stack

• Python 3.10+
• Streamlit
• pandas
• httpx
• beautifulsoup4
• pydantic
• plotly
• kaleido
• playwright
• tenacity
• rapidfuzz
• pytest

Project structure

finance_dashboard/
  app.py
  requirements.txt
  README.md
  .gitignore
  data/
    watchlist.json
    settings.json
    cache/
    exports/
  pages/
    Settings.py
  src/
    config.py
    models.py
    storage.py
    providers/
      base.py
      onvista_resolver.py
      onvista_history_provider.py
    services/
      cache_service.py
      chart_service.py
      instrument_service.py
    ui/
      cards.py
      charts.py
      sidebar.py
      theme.py
    utils/
      logging_utils.py
      text_utils.py
  tests/
    test_chart_service.py
    test_models.py
    test_storage.py

Features

Dashboard

• local chart rendering inside the app
• period selector for 1D, 1W, 1M, 3M, 1Y, 3Y, 5Y, and MAX
• per-instrument error handling so one bad instrument does not crash the whole page

Watchlist

• add by ISIN
• add by WKN
• add by free-text name with selectable matches
• add by direct onvista URL
• remove instruments
• reorder instruments

Settings

• change dashboard grid columns
• change cache TTL
• change search result limit
• toggle HTML snapshot export preference

Persistence

• data/watchlist.json
• data/settings.json
• cached provider responses in data/cache/

Local setup

1. Create and activate a virtual environment

python -m venv .venv
.venv\Scripts\Activate.ps1

2. Install dependencies

pip install -r requirements.txt

3. Install Playwright Chromium

python -m playwright install chromium

4. Run the dashboard

python -m streamlit run app.py

The app will start locally, typically at:

http://localhost:8501

One-file VPS install for Ubuntu

This repository includes a single interactive deployment script:

./deploy.sh

The script is designed for Ubuntu VPS deployment and will ask for the values it needs during setup, including:

• domain or subdomain
• install directory
• Linux app user
• git repository URL
• branch
• reverse proxy choice
• whether Playwright Chromium should be installed
• whether ufw should be configured

Default recommendation

The script defaults to Caddy, which is the recommended choice for this project because it is simpler than Nginx and handles HTTPS automatically in the common single-domain VPS case.

The deployment script installs Caddy from the official Caddy APT repository on Ubuntu, so you do not need to preinstall it yourself.

Quick start on a fresh Ubuntu VPS

git clone https://github.com/AlexDashT/onvistaFinanceDashboard.git
cd onvistaFinanceDashboard
chmod +x deploy.sh
sudo ./deploy.sh

What the script does

• installs required Ubuntu packages
• clones or updates the repository
• creates a Python virtual environment
• installs Python dependencies
• optionally installs Playwright Chromium
• creates a systemd service for Streamlit
• configures Caddy or Nginx
• optionally configures ufw

Important before you run it

• point your domain or subdomain to the VPS first
• make sure you can use sudo
• for Nginx + Certbot, have an email ready for Let's Encrypt
• Caddy is the easiest default unless you specifically want Nginx

Streamlit pages

• Main dashboard: app.py
• Settings page: pages/Settings.py

Data files

data/watchlist.json

Example:

[
  {
    "display_name": "Oberbank Premium Strategie defensiv - I EUR ACC",
    "isin": "AT0000A20UD5",
    "wkn": "A2JH4L",
    "onvista_url": "https://www.onvista.de/fonds/OBERBANK-PREMIUM-STRATEGIE-DEFENSIV-I-EUR-ACC-Fonds-AT0000A20UD5",
    "instrument_type": "fund",
    "currency": "EUR",
    "source_label": "KVG"
  }
]

data/settings.json

Stores app-level UI settings such as:

• selected chart period
• dashboard column count
• cache TTL
• export preference
• search result limit

Running tests

python -m pytest -q

Packaging notes for Windows

This project is built for correctness and maintainability first, not packaging first.

For local use, the recommended run command is still:

python -m streamlit run app.py

If you want to package it later with PyInstaller, you will usually want a dedicated launcher script that starts Streamlit programmatically. That packaging wrapper is not included yet, so treat PyInstaller support as a documented next step rather than a finished feature in the current repository state.

Deployment notes for web

The app is already structured for later deployment as a web dashboard:

• no database dependency
• JSON-based local persistence
• provider logic isolated from UI code
• Streamlit-native interface

For hosted deployment, the main adjustments will likely be:

• replacing local JSON files with a deployment-safe storage option
• setting writable cache/export directories for the hosting platform
• installing Playwright browser dependencies on the target host if browser-based features are added later

Limitations

• this project depends on onvista page and API structures that can change over time
• screenshot-based chart fallback is not implemented yet
• chart image export is not implemented yet
• HTML snapshot export is only a stored preference at the moment
• this repository currently focuses on local charts, watchlist management, and app structure

Troubleshooting

streamlit is not recognized

Use:

python -m streamlit run app.py

Playwright browser is missing

Run:

python -m playwright install chromium

onvista search or chart fails

Possible causes:

• network timeout
• temporary onvista outage
• changed page structure
• changed onvista API shape

Try:

• refreshing the app
• searching the instrument again
• clearing cached files from data/cache/

Settings file fails to load

Delete the local data/settings.json and restart the app to regenerate defaults.

Notes

This repository contains a sample watchlist.json and default settings.json so the app can be started immediately after cloning.