Issue: Documentation to be More Detailed, including Roadmap
Description
Current documentation (README.md) provides basic usage examples but lacks comprehensive details needed for contributors and advanced users.
Gaps in Current Documentation
1. Architecture Documentation
- No explanation of the data flow between components
- Missing description showing how parser → converter → writer interact
- LLM prompt design rationale is not documented
2. Development Guide
Missing:
- How to set up development environment
- Code style and contribution guidelines
3. Configuration Reference
Missing:
- Complete list of LLM config options
- Rate limiting parameters explained in README (30s cooldown between inferences)
- Environment variable reference (reference table)
- Checkpoint file format (purpose and how to load from it - can be by simply pasting into project directory?)
4. Proposed Roadmap (convert to diagram)
- Show already supported three criminal major acts (BNS, BNSS, BSA)
- Mention support for all other enactments through IndiaCode coming soon
- Mention integration with IndiaCode PDFs currently (egazette and custom PDFs in the long term)
- Add integration with SCR judgments supplied by the user in the short-medium term
- Add support for user uploaded CSV files in the short term
- Show support for Azure hosted models, huggingface to be added soon
- Mention integration with Indigo platform
- Mention integration with customised API server
Issue: Documentation to be More Detailed, including Roadmap
Description
Current documentation (README.md) provides basic usage examples but lacks comprehensive details needed for contributors and advanced users.
Gaps in Current Documentation
1. Architecture Documentation
2. Development Guide
Missing:
3. Configuration Reference
Missing:
4. Proposed Roadmap (convert to diagram)