Add fuzzing

[Eeems]

Summary by CodeRabbit

• New Features

  • Added an automated fuzzing harness and runnable fuzz target with deterministic seed handling.

• Chores

  • Broadened CI Python matrix, added a dedicated fuzz job, and made fuzzing a prerequisite for publish/release.
  • Added Makefile support for running fuzzing with timeout control and expanded phony targets.
  • Consolidated packaging config, introduced optional dependency groups, and removed dynamic dependency sourcing.
  • Updated ignore rules to exclude fuzz artifacts and seed corpus; trimmed requirements file.

• Refactor

  • Widespread type-annotation and import cleanups across the codebase.

[coderabbitai]

📝 Walkthrough

Walkthrough

Adds an Atheris fuzzing harness and CI fuzz job, moves dependency declarations into pyproject.toml with optional groups, extends Makefile with a fuzz target and timeout, updates .gitignore for fuzz artifacts, and applies widespread import regrouping and type-annotation tightening across the ext4 package.

Changes

Cohort / File(s)	Summary
Fuzz harness & CI fuzz.py, .github/workflows/build.yaml, .gitignore	Adds fuzz.py (Atheris harness with TestOneInput and custom mutator), a CI fuzz job and Python version matrices, updates publish/release jobs to depend on fuzz, and ignores fuzz artifacts (crash-*, timeout-*, corpus/seed/).
Makefile & script updates Makefile, test.sh	Makefile: remove requirements.txt as build input, switch bootstrap to pyproject.toml, add FUZZ_TIMEOUT default and fuzz target, install editable extras per-target, expand .PHONY; test.sh replaced inline pip installs with invoking make venv activation targets.
Packaging & tooling pyproject.toml, requirements.txt	Moves deps into pyproject.toml (explicit dependencies, optional groups dev/test/fuzz), removes dynamic file-sourced deps, adds tool.ruff/tool.pyright configs; removes cachetools/crcmod entries from requirements.txt.
Public exports ext4/__init__.py	Reorganized imports and declares an explicit __all__ enumerating exported symbols (block/device/IO types, directory/entry types, extents/htree/xattr types, enums/constants, Superblock, Volume, etc.).
Compatibility shim ext4/_compat.py	Conditional override import: use typing_extensions.override for Python <3.12 and typing.override for ≥3.12; removes custom fallback and adds Pyright directives.
Typing & imports (library) ext4/*.py ext4/block.py, ext4/blockdescriptor.py, ext4/directory.py, ext4/enum.py, ext4/extent.py, ext4/htree.py, ext4/inode.py, ext4/struct.py, ext4/superblock.py, ext4/volume.py, ext4/xattr.py	Widespread import regrouping and tightening of type annotations across many modules (added -> None, explicit return types on properties/methods, narrower forward refs). Notable functional tweaks: Ext4Struct short-read OSError now includes errno.EIO; BlockIO.seek uses os.strerror for the error string; Inode.__new__ now uses get_file_type + UnknownInode instead of immediate exception.
Tests test.py	Adds explicit return annotations, linters suppressions, small runtime assertions ensuring volume.superblock/bad_blocks/boot_loader/journal not None before further checks.

Sequence Diagram(s)

sequenceDiagram
  participant A as "Atheris runner"
  participant F as "fuzz.py"
  participant M as "mkfs.ext4 (subprocess)"
  participant V as "ext4.Volume"
  participant I as "Temporary image file"

  rect rgba(200,200,255,0.5)
  A->>F: Start fuzzing (Setup + TestOneInput)
  end

  rect rgba(200,255,200,0.5)
  F->>M: Invoke mkfs.ext4 with generated tree and features
  M-->>I: Produce filesystem image file
  end

  rect rgba(255,200,200,0.5)
  F->>V: Open image via ext4.Volume and exercise parsing (superblock, groups, inodes, htree, xattrs)
  V-->>A: Exceptions/returns exercised by library
  end

  rect rgba(220,220,220,0.5)
  F->>I: Cleanup temporary image
  end

Loading

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~45 minutes

Possibly related PRs

• Add more type hints #11: Overlapping bulk type-hint and signature refactors across ext4 modules (Inode, Volume, Struct, htree, extent).
• Fix #6 #7: Related changes to ext4/struct.py read semantics and I/O handling.

Poem

🐰 I tunneled bytes and stitched a tiny stream,

I seeded corpus paths where fuzzer-sprites gleam,
I nibbled imports and lined types neat and bright,
Atheris hums softly through the quiet night,
Little paws applaud — parse on, ext4 delight.

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 1.83% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'Add fuzzing' is directly related to the main change: introducing fuzzing support via fuzz.py, workflow updates, dependency management, and type annotation improvements.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}