feat: AD1 (AccessData logical image) lazy mount + doc lifecycle#3
Merged
Conversation
Add the FsType::Ad1 variant + Display + FromStr so the tests compile, and two failing detection tests asserting the AccessData AD1 magic bytes map to FsType::Ad1. The magic check itself follows in GREEN. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add the magic-byte checks to detect_filesystem so AccessData AD1 images (and their ADCRYPT encrypted variant) resolve to FsType::Ad1. Both detection tests pass. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add the ad1 feature (default) + ad1-core dep, and dev-deps (ad1-core testfix + tempfile). New fs_ad1 module with the Ad1ForensicFs skeleton (stubbed) and four tests driven by ad1-core's testfix fixture builder: list the tree, read every file byte-identical to testfix's independently hashed ground truth, a mid-file range read across a zlib-chunk boundary, and fs_info. All four fail (unimplemented); implementation follows. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Ad1ForensicFs::open enumerates the AD1 logical tree into an ArchiveTree (file entries carry their reader index as payload); reads go through read_range, which resolves the entry and loops ad1::read_at (handling short reads) to inflate only the overlapping zlib chunks — no full extraction. ForensicFs methods delegate to the tree; encrypted (ADCRYPT) images surface as NotSupported via map_err. All four testfix-oracle tests pass; clippy -D warnings and fmt clean. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add the FsType::Ad1 arm to the mount dispatch. Unlike seekable-stream formats, AD1 opens by path (to find sibling segments) and exposes a logical tree, so it constructs Ad1ForensicFs directly rather than funnelling through build_filesystem. It inherits the DiskOverlay layout (ro/ rw/ deleted/) like the archive backends. ADCRYPT images fail open with a clear message and are refused. Binary builds; clippy/fmt clean. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Commit a 1150-byte AD1 fixture (built by ad1-core's spec-faithful testfix writer, verified to read back through Ad1ForensicFs), copied by gen-fixtures.sh and asserted by the `ad1` manifest row (root/hello.txt = "hello from ad1"). Provenance + md5 documented in tests/data/README.md. This mounts AD1 end-to-end through FUSE (Linux) and Dokan (Windows) alongside the other 13 formats. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Reverse-write ADR 0001 capturing why AD1 mounts as a logical tree built directly from the path (bypassing build_filesystem), reads lazily via ad1-core, refuses ADCRYPT, and uses the DiskOverlay layout. The planning handoff (HANDOFF-ad1.md) is retired now that the work is built and its rationale lives here — per the plan/doc lifecycle standard, the working tree holds the conclusion, git history holds the archaeology. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…ild output Apply the .gitignore standard: cover local agent state (.claude/) and cargo-fuzz's regenerated output (fuzz/target/, fuzz/corpus/, fuzz/artifacts/) plus a couple more editor cruft globs. Untrack the 3706 fuzz/target + fuzz/corpus files that had been committed by mistake — the files stay on disk and git history retains them, so nothing is lost; the working tree just stops carrying regenerated build output. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
|
Review the following changes in direct dependencies. Learn more about Socket for GitHub.
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Mount AccessData AD1 logical images (
4n6mount evidence.ad1 /mnt) lazily via the pure-Rustad1-corecrate — no full extraction of a potentially multi-GiB image. Implemented with strict TDD.AD1 support
ADSEGMENTEDFILE/ADCRYPTmagic →FsType::Ad1.fs_ad1(RED→GREEN):Ad1ForensicFsbuilds theForensicFstree directly from the path (AD1 is a logical tree opened by path, so it bypassesbuild_filesystem), reading bytes lazily viaad1::read_at(inflates only overlapping zlib chunks). Encrypted (ADCRYPT) images are refused loudly.ad1-core'stestfixindependent oracle (spec-faithful writer, flate2 + RustCrypto) — byte-identical reads incl. a mid-file range across a zlib-chunk boundary. Plus a committed 1150-byte fixture added to the mount-smoke matrix (14th format, FUSE + Dokan).ad1is a default feature.Doc lifecycle (per CLAUDE.md standard)
docs/decisions/) capturing the why; retired the implementedHANDOFF-ad1.md..gitignore: added.claude/+ cargo-fuzz output (fuzz/target/,fuzz/corpus/,fuzz/artifacts/); untracked 3706 fuzz build-output/corpus files committed by mistake (files kept on disk; history retains them).Test
Local: 202 lib + 10 integration tests green; clippy
-D warnings+ fmt clean;cargo denygreen withad1-corein the graph.🤖 Generated with Claude Code