A high-performance desktop GUI application for analyzing large log files (hundreds of MB to multiple GB) with real-time filtering, timeline visualization, error clustering, and report export.

This project demonstrates advanced PySide6 usage, efficient file processing, and production-minded GUI architecture.

• Efficient large-file handling

  • Memory-mapped file access (no full file loading)
  • Streaming index construction
  • Constant-memory random access to log lines

• Interactive filtering

  • Regex-based filtering (Python re)
  • Log level filtering (INFO, WARN, ERROR, etc.)
  • Time-bucket filtering via timeline interaction

• Timeline visualization

  • Histogram of log activity by minute
  • Click-to-drill-down into specific time ranges

• Error clustering

  • Normalizes log messages to group similar errors
  • Highlights recurring issues and patterns
  • One-click drill-down into cluster samples

• Export & reporting

  • CSV (spreadsheet-friendly)
  • JSON Lines (machine-readable, streaming-friendly)
  • HTML report (shareable summary + preview)

• Robust GUI design

  • Threaded workers for indexing, filtering, clustering, exporting
  • GUI updates restricted to the main thread
  • Custom QAbstractTableModel with lazy loading and LRU caching

The application is intentionally modularized for maintainability and clarity:

. ├── main.py # GUI, orchestration, signal wiring ├── log_file.py # File validation, memory-mapped access ├── indexing.py # Index construction and parsing logic ├── filtering.py # Filtering and error clustering workers ├── models.py # Qt table models (lazy data access) ├── export.py # CSV / JSONL / HTML export workers └── README.md

• Separation of concerns (UI vs. I/O vs. processing)
• Thread safety (workers emit signals, GUI updates via slots)
• Scalable performance (memory usage independent of file size)

• Python 3.10+
• PySide6

Install dependencies:

pip install pyside6

python main.py

1. Open a .log file

2. Apply filters (regex, log level, timeline)

3. Inspect clustered errors

4. Export results as needed

• For spreadsheets and analysts

• One row per log entry

• Columns: timestamp, level, message

• For automated pipelines and scripts

• One JSON object per line

• Stream-friendly (no full-file loading)

• Human-readable, shareable report

• Includes summary statistics and a preview table

• Designed to remain responsive even for large datasets

• Only plain-text .log files are accepted

• Binary files are rejected via content inspection

• Symlinks are refused for safety

• Permanent delete utilities are guarded and extension-restricted

• Memory-mapped file access (mmap)

• Offset-based indexing (byte-accurate line access)

• Custom Qt table model with lazy loading

• Thread-safe signal/slot architecture

• Defensive programming against invalid input

• Explicit painter lifecycle management

• Sparse or hierarchical indexing for very large logs

• JSON log format auto-detection

• Saved filter presets

• Bookmarks and annotations