Bracu RoutineParser

Bracu RoutineParser extracts class routine data from BRACU routine images.

It currently supports:

• Connect downloaded routines (screenshot)
• Connect screenshots
• Preprereg screenshots

The parser can return:

• Full routine entries (day, time, course, section, faculty, room)
• Minimal output (course code + section only) for autofill workflows

Features

• EasyOCR-based OCR pipeline (CPU-friendly)
• Handles multiple screenshot styles
• Detects and crops exam schedule section in mixed images
• Course+section-only parser for quick website integration

Project structure

• routine_parser/ - parser core logic
• examples/run_parser.py - full parser CLI
• examples/run_course_section_parser.py - minimal parser CLI

Quick setup

Python version:

• 3.11.11 (from .python-version)

python --version

Install dependencies:

python -m pip install -U pip setuptools wheel
python -m pip install --index-url https://download.pytorch.org/whl/cpu torch torchvision
python -m pip install -r requirements.txt

Usage

Full parser

python examples/run_parser.py image/image.png

Returns structured routine JSON including classes, meta warnings, and parser details.

Course + section only (recommended for autofill)

python examples/run_course_section_parser.py image/image.png

Returns:

{
  "course_sections": [
    { "course_code": "CSE330", "section": "06" }
  ],
  "meta": {
    "source_image": "image/image.png",
    "template_id": "light_grid",
    "warnings": []
  }
}

Notes and limitations

• OCR quality depends on screenshot clarity.
• Borderless/low-quality images may still need manual correction.
• Parser is tuned for BRACU routine formats and may not generalize to unrelated layouts.