conflict-collection 0.0.2

Data collection from Git merge conflicts.

conflict-collection

Data collection toolkit for Git merge conflicts: structural types, social signals, and similarity metrics.

conflict-collection builds on conflict-parser to turn raw in-progress merges into rich, typed records suitable for research, analytics, or ML.

---

✨ Features

• Classify and enumerate merge conflict cases (modify/modify, add/add, delete/modify, etc.)
• Produce structured frozen dataclasses with per-side contents & resolution metadata
• Build canonical 5‑tuples (A,B,O,M,R) for dataset curation
• Extract social / ownership signals (recency, blame composition, integrator priors)
• Compute a 3‑way anchored similarity ratio between two resolution candidates
• Fully typed (PEP 561) & test‑covered

---

📦 Install

pip install conflict-collection
# or with documentation extras
pip install conflict-collection[docs]

---

🚀 Quick Start

from conflict_collection.collectors.conflict_type import collect as collect_conflicts
from conflict_collection.collectors.societal import collect as collect_social
from conflict_collection.metrics.anchored_ratio import anchored_ratio

# 1) Enumerate conflict cases after a merge produced conflicts
cases = collect_conflicts(repo_path='.', resolution_sha='<resolved-commit-sha>')
print(len(cases), 'cases')
print(cases[0].conflict_type, cases[0].conflict_path)

# 2) Capture social signals
signals = collect_social(repo_path='.')
for path, rec in signals.items():
	print(path, rec.ours_author, rec.owner_commits_ours)

# 3) Similarity metric example
O = 'line1\nline2\nline3'
R = 'line1\nX\nline3'
R_hat = 'line1\nY\nline3'
print('anchored ratio =', anchored_ratio(O, R, R_hat))

---

📚 Documentation

Full docs (usage guides + auto-generated API reference) are published with MkDocs & mkdocstrings:

https://jinu-jang.github.io/conflict-collection

Local build:

pip install -e .[docs]
mkdocs serve

---

🧩 Data Models

Model	Purpose
Typed Conflict Cases	Frozen dataclasses per conflict archetype
Conflict5Tuple	Canonical (A,B,O,M,R) capture
Social Signals	Ownership & recency metrics per file
Anchored Ratio	Algorithmic similarity between two edits

---

🔬 Testing

git clone https://github.com/jinu-jang/conflict-collection
cd conflict-collection
pip install -e .[dev]
pytest -q

---

🤝 Contributing

PRs welcome! Please:

1. Add or update tests
2. Run black . && isort . && pytest -q
3. If adding public API, include docstrings & update docs nav (mkdocs.yml)

See docs/contributing.md for details.

---

📄 License

MIT © 2025 Jinu Jang

---

🔖 Status

Beta. Interfaces may change before 0.1. Feedback appreciated.