downloadFile opened the temp file with os.Create, which opens and
empties a file already at that name. If that file was a hard link to a
file outside the destination directory, fetch overwrote the outside
file. It now removes whatever is at the temp name, which removes only
that name, and creates the temp file with O_EXCL, so the create fails if
anything is still or again there. A leftover temp file from an
interrupted run is still replaced. The new test puts a hard link at the
temp name and checks that fetch succeeds and the outside file is
unchanged. The command name is now the constant cmdFetch, like the other
command names, because lint requires it once a third test uses it.
Model: opus-5-5
script/test now runs go test -timeout 30s -race -cover ./..., quiet on success and rerunning verbose on failure. -race exposed a data race on the process-global apex/log logger: Init reconfigured it on every CLI run while other goroutines logged through it. internal/log now mutates the global under the write lock and reads it under the read lock; WithError is dropped because its Entry logged outside that lock, and run() reports a failed command's error via log.Errorf instead (still shown under -q). The corruption fuzz test is scaled to 1500 files / 100 iterations to fit the budget under -race; it pins the same claims at reduced breadth. Independent review ran the Docker lint gate uncached.
Model: opus-4-8 (implementation, review)
model: claude-fable-5
Adopts golangci-lint v2.12.2 and the canonical .golangci.yml (default: all), and fixes all resulting findings across the tree.
Two intended behavior changes: absent MFFilePath.Mtime is handled explicitly in freshen, list and export rather than dereferenced (main panicked); gpg positional key IDs now follow an explicit -- end-of-options marker.
All twelve reworded user-visible error messages restored to byte-identical parity with main and pinned by tests.
Generates a ~1MB manifest (20000 files with random names), then:
- Verifies truncated manifest causes check to fail
- Runs 500 iterations of random single-byte corruption
- Each iteration verifies check detects the corruption
- Atomic writes for mfer gen: writes to temp file, renames on success,
cleans up temp on error/interrupt. Prevents empty manifests on Ctrl-C.
- Humanized byte sizes using dustin/go-humanize (e.g., "10 MiB" not "10485760")
- Progress lines clear when done (using ANSI escape \r\033[K])
- Debug logging when files are added to manifest (mfer gen -vv)
- Move -v/-q flags from global to per-command for better UX
- Add tests for atomic write behavior with failing filesystem mock
- Version command now uses mfer.Version constant instead of empty build flags
- Running just 'mfer' shows banner + help
- Unknown commands still return error with exit code 1
- pathIsHidden(".") was returning true, causing freshen to skip entire
directory tree when dotfiles excluded
- Banner now prints directly to stdout to avoid log prefix artifacts
Changed the default behavior to exclude dotfiles (files/dirs starting with .)
which is the more common use case. Added --include-dotfiles flag for when
hidden files need to be included in the manifest.
Major changes:
- Refactor CLI to accept injected I/O streams and filesystem (afero.Fs)
for testing without touching the real filesystem
- Add RunOptions struct and RunWithOptions() for configurable CLI execution
- Add internal/scanner package with two-phase manifest generation:
- Phase 1 (Enumeration): walk directories, collect metadata
- Phase 2 (Scan): read contents, compute hashes, write manifest
- Add internal/checker package for manifest verification with progress
reporting and channel-based result streaming
- Add mfer/builder.go for incremental manifest construction
- Add --no-extra-files flag to check command to detect files not in manifest
- Add timing summaries showing file count, size, elapsed time, and throughput
- Add comprehensive tests using afero.MemMapFs (no real filesystem access)
- Add contrib/usage.sh integration test script
- Fix banner ASCII art alignment (consistent spacing)
- Fix verbosity levels so summaries display at default log level
- Update internal/log to support configurable output writers
clawbot
sneak