Week 5 - M1: Packfile compression support #31

• Owner: @mahis1067 · minigit/objects.py, tests/test_objects.py
• Docs: docs/packfiles.md
• Depends on: Week 4 - M1: Object-store hardening + packfile groundwork #22
• Branch from updated main: feat/packfile-compression

Steps

1. Review Week 4's packfile proposal.
2. Read selected objects through read_object; deduplicate hashes.
3. Write a compressed pack + hash-to-offset index under .minigit/objects/.
4. Store each object's type + exact bytes; keep original hashes.
5. Write temporary files first; publish only a complete pack + index.
6. Keep loose originals. New writes can stay loose; pruning/delta compression later.
7. Extend read_object: loose + packed lookup, same return type.
8. Validate type, lengths, offsets, compression, and hash.
   • Missing -> ObjectNotFoundError
   • Corrupt -> ObjectCorruptError
9. Keep has_object and duplicate writes working for packed-only objects.
10. Document format, lookup order, packing method, and one usage example.
11. scripts/quality-check.sh -> linked PR -> teammate review.

Tests

• Fresh store: packed blobs, trees, commits -> original bytes
• Packed-only reads; mixed loose/packed reads
• Empty/binary/Unicode content; repeated hashes
• Truncated pack, bad index, broken zlib, wrong hash
• Interrupted write -> old objects still readable
• Existing tree/history tests still pass

Done when

• Pack round-trip works; hashes unchanged
• Packed-only lookup + corruption checks work
• Format + usage documented
• Quality check green; PR open; review requested