Built for a friend's personal investment workflow — automatically fetches and updates daily stock scan data so he can make informed trading decisions without manual data collection.
This project demonstrates full-stack desktop app development—from data fetching and database management to UI design and Windows packaging. Built with a focus on code organization, error handling, and user experience.
- Data Fetching: Automatically download and merge latest stock scan data from GitHub
- Database Management: Local SQLite storage with incremental updates and deletion tracking
- Advanced Filtering: Filter by file, date, column search, price range, and volume thresholds
- Sorting & Ranking: Sort by any column with dynamic indicators
- Excel Export: Export filtered results with formatted numbers
- Settings Persistence: Automatically save column widths, window size, and filter preferences
- Desktop Packaging: Builds to a standalone Windows
.exewith PyInstaller
- UI Framework: Tkinter with ttkbootstrap (modern theming)
- Data Processing: Pandas for manipulation and analysis
- Database: SQLite3 for persistent local storage
- Web Scraping: BeautifulSoup & Requests for data fetching
- Packaging: PyInstaller for Windows desktop distribution
- Language: Python 3.13+
pip install pandas requests beautifulsoup4 ttkbootstrappython app_github.pypython -m PyInstaller --onefile -w app_github.py
# Output: dist/app_github.exeMain UI
Filters and Export
app_github.py— Main application entry point (UI + logic)fetch_data.py— Data fetching and merging utilitiesAPP_GITHUB_ARCHITECTURE.md— Detailed system design and API documentation.github/workflows/— GitHub Actions for automated database updates
- Incremental Updates: Only new records are merged, avoiding duplicates
- Deleted Records Tracking: Maintains a history of deleted rows to prevent re-importing
- Runtime Data Persistence: Database files are written next to the executable, not in temp folders
- Error Handling: Comprehensive try-catch with user-friendly error dialogs
- Dynamic Schema: Automatically adds columns as new data arrives
MIT