Files
mfer/mfer/checker_test.go
T
clawbot 2a270b40c5
check / check (push) Failing after 4s
Record file mode in the manifest, 0000 unless asked (closes #161)
MFFilePath gains mode (field 304): a file's permission bits, 0777 at
most, or 0000, meaning none recorded. gen and freshen record real modes
only with --include-permissions (ScannerOptions.IncludePermissions); the
builder keeps only mode.Perm(), so setuid, setgid and sticky are never
written. list -l and export show the mode in octal. check reports
MODE_MISMATCH for a recorded mode other than 0000 the file lacks. fetch
refuses a manifest with a mode above 0777 before requesting any file,
sets only the permission bits of each recorded mode on the files it
writes, and downloads again a present file whose mode differs. The decoding-cost bound counts a file
entry at 176 bytes, up from 160.

Model: opus-5-5
2026-10-06 11:43:18 +02:00

843 lines
22 KiB
Go

//nolint:testpackage // white-box tests exercise unexported internals
package mfer
import (
"bytes"
"context"
"fmt"
"os"
"path/filepath"
"testing"
"time"
"github.com/spf13/afero"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
const (
testFile1 = "file1.txt"
testFile2 = "file2.txt"
testExistsFile = "exists.txt"
testManifestPath = "/manifest.mf"
testDataDir = "/data"
)
func TestStatusString(t *testing.T) {
t.Parallel()
tests := []struct {
status Status
expected string
}{
{StatusOK, "OK"},
{StatusMissing, "MISSING"},
{StatusSizeMismatch, "SIZE_MISMATCH"},
{StatusHashMismatch, "HASH_MISMATCH"},
{StatusModeMismatch, "MODE_MISMATCH"},
{StatusExtra, "EXTRA"},
{StatusError, "ERROR"},
{Status(99), "UNKNOWN"},
}
for _, tt := range tests {
t.Run(tt.expected, func(t *testing.T) {
t.Parallel()
assert.Equal(t, tt.expected, tt.status.String())
})
}
}
// createTestManifest creates a manifest file in the filesystem with the given files.
func createTestManifest(
t *testing.T, fs afero.Fs, manifestPath string, files map[string][]byte,
) {
t.Helper()
builder := NewBuilder()
for path, content := range files {
reader := bytes.NewReader(content)
_, err := builder.AddFile(
RelFilePath(path), FileSize(len(content)), ModTime(time.Now()), 0, reader, nil,
)
require.NoError(t, err)
}
var buf bytes.Buffer
require.NoError(t, builder.Build(context.Background(), &buf))
require.NoError(t, afero.WriteFile(fs, manifestPath, buf.Bytes(), 0o644))
}
// createFilesOnDisk creates the given files on the filesystem under
// testDataDir.
func createFilesOnDisk(t *testing.T, fs afero.Fs, files map[string][]byte) {
t.Helper()
for path, content := range files {
fullPath := testDataDir + "/" + path
require.NoError(t, fs.MkdirAll(testDataDir, 0o755))
require.NoError(t, afero.WriteFile(fs, fullPath, content, 0o644))
}
}
func TestNewChecker(t *testing.T) {
t.Parallel()
t.Run("valid manifest", func(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
files := map[string][]byte{
testFile1: []byte("hello"),
testFile2: []byte("world"),
}
createTestManifest(t, fs, testManifestPath, files)
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: "/",
Fs: fs,
})
require.NoError(t, err)
assert.NotNil(t, chk)
assert.Equal(t, FileCount(2), chk.FileCount())
})
t.Run("missing manifest", func(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
_, err := NewChecker(&CheckerOptions{
ManifestPath: "/nonexistent.mf",
BasePath: "/",
Fs: fs,
})
assert.Error(t, err)
})
t.Run("invalid manifest", func(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
require.NoError(t, afero.WriteFile(fs, "/bad.mf", []byte("not a manifest"), 0o644))
_, err := NewChecker(&CheckerOptions{
ManifestPath: "/bad.mf",
BasePath: "/",
Fs: fs,
})
assert.Error(t, err)
})
}
func TestNewCheckerRequiredPaths(t *testing.T) {
t.Parallel()
for _, tc := range []struct {
name string
opts *CheckerOptions
want string
is error
}{
{
name: "nil options",
opts: nil,
want: "manifest path cannot be empty",
is: errManifestPathEmpty,
},
{
name: "empty manifest path",
opts: &CheckerOptions{BasePath: testDataDir},
want: "manifest path cannot be empty",
is: errManifestPathEmpty,
},
{
name: "empty base path",
opts: &CheckerOptions{ManifestPath: testManifestPath},
want: "base path cannot be empty",
is: errBasePathEmpty,
},
} {
t.Run(tc.name, func(t *testing.T) {
t.Parallel()
chk, err := NewChecker(tc.opts)
require.ErrorIs(t, err, tc.is)
require.EqualError(t, err, tc.want)
assert.Nil(t, chk)
})
}
}
func TestNewCheckerNilFsUsesOsFs(t *testing.T) {
t.Parallel()
dir := t.TempDir()
manifestPath := filepath.Join(dir, "index.mf")
content := []byte("hello")
createTestManifest(t, afero.NewOsFs(), manifestPath, map[string][]byte{
testFile1: content,
})
require.NoError(t, os.WriteFile(filepath.Join(dir, testFile1), content, 0o600))
chk, err := NewChecker(&CheckerOptions{ManifestPath: manifestPath, BasePath: dir})
require.NoError(t, err)
results := make(chan Result, 1)
require.NoError(t, chk.Check(context.Background(), results, nil))
assert.Equal(t, StatusOK, (<-results).Status)
}
func TestCheckerFileCountAndTotalBytes(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
files := map[string][]byte{
"small.txt": []byte("hi"),
"medium.txt": []byte("hello world"),
"large.txt": bytes.Repeat([]byte("x"), 1000),
}
createTestManifest(t, fs, testManifestPath, files)
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: "/",
Fs: fs,
})
require.NoError(t, err)
assert.Equal(t, FileCount(3), chk.FileCount())
assert.Equal(t, FileSize(2+11+1000), chk.TotalBytes())
}
func TestCheckAllFilesOK(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
files := map[string][]byte{
testFile1: []byte("content one"),
testFile2: []byte("content two"),
}
createTestManifest(t, fs, testManifestPath, files)
createFilesOnDisk(t, fs, files)
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
err = chk.Check(context.Background(), results, nil)
require.NoError(t, err)
var resultList []Result
for r := range results {
resultList = append(resultList, r)
}
assert.Len(t, resultList, 2)
for _, r := range resultList {
assert.Equal(t, StatusOK, r.Status, "file %s should be OK", r.Path)
}
}
func TestCheckMissingFile(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
files := map[string][]byte{
testExistsFile: []byte("I exist"),
"missing.txt": []byte("I don't exist on disk"),
}
createTestManifest(t, fs, testManifestPath, files)
// Only create one file
createFilesOnDisk(t, fs, map[string][]byte{
testExistsFile: []byte("I exist"),
})
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
err = chk.Check(context.Background(), results, nil)
require.NoError(t, err)
var okCount, missingCount int
for r := range results {
switch r.Status {
case StatusOK:
okCount++
case StatusMissing:
missingCount++
assert.Equal(t, RelFilePath("missing.txt"), r.Path)
case StatusSizeMismatch, StatusHashMismatch, StatusModeMismatch,
StatusExtra, StatusError:
// Not expected in this test; counted assertions below will fail.
}
}
assert.Equal(t, 1, okCount)
assert.Equal(t, 1, missingCount)
}
func TestCheckSizeMismatch(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
files := map[string][]byte{
testFileName: []byte("original content"),
}
createTestManifest(t, fs, testManifestPath, files)
// Create file with different size
createFilesOnDisk(t, fs, map[string][]byte{
testFileName: []byte("short"),
})
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
err = chk.Check(context.Background(), results, nil)
require.NoError(t, err)
r := <-results
assert.Equal(t, StatusSizeMismatch, r.Status)
assert.Equal(t, RelFilePath(testFileName), r.Path)
}
func TestCheckHashMismatch(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
originalContent := []byte("original content")
files := map[string][]byte{
testFileName: originalContent,
}
createTestManifest(t, fs, testManifestPath, files)
// Create file with same size but different content
differentContent := []byte("different contnt") // same length (16 bytes) but different
require.Len(t, differentContent, len(originalContent), "test requires same length")
createFilesOnDisk(t, fs, map[string][]byte{
testFileName: differentContent,
})
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
err = chk.Check(context.Background(), results, nil)
require.NoError(t, err)
r := <-results
assert.Equal(t, StatusHashMismatch, r.Status)
assert.Equal(t, RelFilePath(testFileName), r.Path)
}
// A recorded mode other than 0000 that differs from the file's permission
// bits fails the check; a recorded 0000 is never checked.
func TestCheckMode(t *testing.T) {
t.Parallel()
for _, tc := range []struct {
name string
recorded os.FileMode
onDisk os.FileMode
want Status
}{
{"recorded mode matches", 0o640, 0o640, StatusOK},
{"recorded mode differs", 0o640, 0o600, StatusModeMismatch},
{"0000 is not checked", 0, 0o600, StatusOK},
} {
t.Run(tc.name, func(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
content := []byte("content")
b := NewBuilder()
_, err := b.AddFile(testFileName, FileSize(len(content)), ModTime{},
tc.recorded, bytes.NewReader(content), nil)
require.NoError(t, err)
var buf bytes.Buffer
require.NoError(t, b.Build(context.Background(), &buf))
require.NoError(t, afero.WriteFile(fs, testManifestPath, buf.Bytes(), 0o644))
require.NoError(t, fs.MkdirAll(testDataDir, 0o755))
require.NoError(t, afero.WriteFile(fs,
testDataDir+"/"+testFileName, content, tc.onDisk))
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 1)
require.NoError(t, chk.Check(context.Background(), results, nil))
assert.Equal(t, tc.want, (<-results).Status)
})
}
}
func TestCheckWithProgress(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
files := map[string][]byte{
testFile1: bytes.Repeat([]byte("a"), 100),
testFile2: bytes.Repeat([]byte("b"), 200),
}
createTestManifest(t, fs, testManifestPath, files)
createFilesOnDisk(t, fs, files)
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
progress := make(chan CheckStatus, 10)
err = chk.Check(context.Background(), results, progress)
require.NoError(t, err)
// results is fully buffered and closed; no draining needed
// Check progress was sent
var progressUpdates []CheckStatus
for p := range progress {
progressUpdates = append(progressUpdates, p)
}
assert.NotEmpty(t, progressUpdates)
// Final progress should show all files checked
final := progressUpdates[len(progressUpdates)-1]
assert.Equal(t, FileCount(2), final.TotalFiles)
assert.Equal(t, FileCount(2), final.CheckedFiles)
assert.Equal(t, FileSize(300), final.TotalBytes)
assert.Equal(t, FileSize(300), final.CheckedBytes)
assert.Equal(t, FileCount(0), final.Failures)
}
func TestCheckContextCancellation(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
// Create many files to ensure we have time to cancel
files := make(map[string][]byte)
for i := range 100 {
files[string(rune('a'+i%26))+".txt"] = bytes.Repeat([]byte("x"), 1000)
}
createTestManifest(t, fs, testManifestPath, files)
createFilesOnDisk(t, fs, files)
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
ctx, cancel := context.WithCancel(context.Background())
cancel() // Cancel immediately
results := make(chan Result, 200)
err = chk.Check(ctx, results, nil)
assert.ErrorIs(t, err, context.Canceled)
}
func TestFindExtraFiles(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
// Manifest only contains file1
manifestFiles := map[string][]byte{
testFile1: []byte("in manifest"),
}
createTestManifest(t, fs, testManifestPath, manifestFiles)
// Disk has file1 and file2
createFilesOnDisk(t, fs, map[string][]byte{
testFile1: []byte("in manifest"),
testFile2: []byte("extra file"),
})
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
err = chk.FindExtraFiles(context.Background(), results)
require.NoError(t, err)
var extras []Result
for r := range results {
extras = append(extras, r)
}
assert.Len(t, extras, 1)
assert.Equal(t, RelFilePath(testFile2), extras[0].Path)
assert.Equal(t, StatusExtra, extras[0].Status)
assert.Equal(t, "not in manifest", extras[0].Message)
}
// TestFindExtraFilesReportsHiddenFilesButNotManifest keeps the manifest
// inside the checked tree: hidden files and directories are reported, the
// manifest is not. The manifest is recognized by file identity, which needs
// the real filesystem.
func TestFindExtraFilesReportsHiddenFilesButNotManifest(t *testing.T) {
t.Parallel()
dir := t.TempDir()
manifestPath := filepath.Join(dir, "index.mf")
fs := afero.NewOsFs()
createTestManifest(t, fs, manifestPath, map[string][]byte{
testFile1: []byte("in manifest"),
})
unlisted := []RelFilePath{"extra.txt", ".hidden", ".git/config"}
for _, p := range append([]RelFilePath{testFile1}, unlisted...) {
path := filepath.Join(dir, string(p))
require.NoError(t, fs.MkdirAll(filepath.Dir(path), 0o750))
require.NoError(t, afero.WriteFile(fs, path, []byte("x"), 0o600))
}
chk, err := NewChecker(&CheckerOptions{
ManifestPath: manifestPath,
BasePath: dir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
require.NoError(t, chk.FindExtraFiles(context.Background(), results))
var extras []RelFilePath
for r := range results {
extras = append(extras, r.Path)
}
assert.ElementsMatch(t, unlisted, extras)
}
// TestFindExtraFilesSkipsManifestReachedThroughSymlink checks a tree whose
// index.mf is a symlink to the manifest kept outside the tree: the symlink is
// not reported.
func TestFindExtraFilesSkipsManifestReachedThroughSymlink(t *testing.T) {
t.Parallel()
dir := t.TempDir()
tree := filepath.Join(dir, "tree")
manifestPath := filepath.Join(dir, "real.mf")
linkPath := filepath.Join(tree, "index.mf")
fs := afero.NewOsFs()
createTestManifest(t, fs, manifestPath, map[string][]byte{testFile1: []byte("x")})
require.NoError(t, fs.MkdirAll(tree, 0o750))
require.NoError(t,
afero.WriteFile(fs, filepath.Join(tree, testFile1), []byte("x"), 0o600))
require.NoError(t, os.Symlink(manifestPath, linkPath))
chk, err := NewChecker(&CheckerOptions{
ManifestPath: linkPath,
BasePath: tree,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
require.NoError(t, chk.FindExtraFiles(context.Background(), results))
var extras []RelFilePath
for r := range results {
extras = append(extras, r.Path)
}
assert.Empty(t, extras)
}
// TestFindExtraFilesSearchesBaseNamedThroughSymlink names the checked tree
// through a symlink to it: the files in the tree are searched, and the
// symlink itself is not reported.
func TestFindExtraFilesSearchesBaseNamedThroughSymlink(t *testing.T) {
t.Parallel()
dir := t.TempDir()
tree := filepath.Join(dir, "tree")
link := filepath.Join(dir, "link")
manifestPath := filepath.Join(dir, "index.mf")
fs := afero.NewOsFs()
createTestManifest(t, fs, manifestPath, map[string][]byte{testFile1: []byte("x")})
require.NoError(t, fs.MkdirAll(tree, 0o750))
for _, name := range []string{testFile1, testFile2} {
require.NoError(t,
afero.WriteFile(fs, filepath.Join(tree, name), []byte("x"), 0o600))
}
require.NoError(t, os.Symlink(tree, link))
chk, err := NewChecker(&CheckerOptions{
ManifestPath: manifestPath,
BasePath: link,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
require.NoError(t, chk.FindExtraFiles(context.Background(), results))
var extras []RelFilePath
for r := range results {
extras = append(extras, r.Path)
}
assert.Equal(t, []RelFilePath{testFile2}, extras)
}
func TestFindExtraFilesContextCancellation(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
files := map[string][]byte{testFileName: []byte("data")}
createTestManifest(t, fs, testManifestPath, files)
createFilesOnDisk(t, fs, files)
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
ctx, cancel := context.WithCancel(context.Background())
cancel() // Cancel immediately
results := make(chan Result, 10)
err = chk.FindExtraFiles(ctx, results)
assert.ErrorIs(t, err, context.Canceled)
}
func TestCheckNilChannels(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
files := map[string][]byte{testFileName: []byte("data")}
createTestManifest(t, fs, testManifestPath, files)
createFilesOnDisk(t, fs, files)
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
// Should not panic with nil channels
err = chk.Check(context.Background(), nil, nil)
assert.NoError(t, err)
}
func TestFindExtraFilesNilChannel(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
files := map[string][]byte{testFileName: []byte("data")}
createTestManifest(t, fs, testManifestPath, files)
createFilesOnDisk(t, fs, files)
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
// Should not panic with nil channel
err = chk.FindExtraFiles(context.Background(), nil)
assert.NoError(t, err)
}
func TestCheckSubdirectories(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
files := map[string][]byte{
"dir1/file1.txt": []byte("content1"),
"dir1/dir2/file2.txt": []byte("content2"),
"dir1/dir2/dir3/deep.txt": []byte("deep content"),
}
createTestManifest(t, fs, testManifestPath, files)
// Create files with full directory structure
for path, content := range files {
fullPath := "/data/" + path
require.NoError(t, fs.MkdirAll("/data/dir1/dir2/dir3", 0o755))
require.NoError(t, afero.WriteFile(fs, fullPath, content, 0o644))
}
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
err = chk.Check(context.Background(), results, nil)
require.NoError(t, err)
var okCount int
for r := range results {
assert.Equal(t, StatusOK, r.Status, "file %s should be OK", r.Path)
okCount++
}
assert.Equal(t, 3, okCount)
}
func TestCheckMissingFileDetectedWithoutFallback(t *testing.T) {
t.Parallel()
// Regression test: errors.Is(err, errors.New("...")) never matches because
// errors.New creates a new value each time. The fix uses os.ErrNotExist instead.
fs := afero.NewMemMapFs()
files := map[string][]byte{
testExistsFile: []byte("here"),
"missing.txt": []byte("not on disk"),
}
createTestManifest(t, fs, testManifestPath, files)
// Only create one file on disk
createFilesOnDisk(t, fs, map[string][]byte{
testExistsFile: []byte("here"),
})
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 10)
err = chk.Check(context.Background(), results, nil)
require.NoError(t, err)
statusCounts := map[Status]int{}
for r := range results {
statusCounts[r.Status]++
if r.Status == StatusMissing {
assert.Equal(t, RelFilePath("missing.txt"), r.Path)
}
}
assert.Equal(t, 1, statusCounts[StatusOK], "one file should be OK")
assert.Equal(t, 1, statusCounts[StatusMissing], "one file should be MISSING")
assert.Equal(t, 0, statusCounts[StatusError], "no files should be ERROR")
}
func TestCheckEmptyManifest(t *testing.T) {
t.Parallel()
fs := afero.NewMemMapFs()
// Create manifest with no files
createTestManifest(t, fs, testManifestPath, map[string][]byte{})
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
assert.Equal(t, FileCount(0), chk.FileCount())
assert.Equal(t, FileSize(0), chk.TotalBytes())
results := make(chan Result, 10)
err = chk.Check(context.Background(), results, nil)
require.NoError(t, err)
var count int
for range results {
count++
}
assert.Equal(t, 0, count)
}
func TestCheckProgressRateLimited(t *testing.T) {
t.Parallel()
// Create many small files - progress should be rate-limited, not one per file.
// With rate-limiting to once per second, we should get far fewer progress
// updates than files (plus one final update).
fs := afero.NewMemMapFs()
files := make(map[string][]byte, 100)
for i := range 100 {
name := fmt.Sprintf("file%03d.txt", i)
files[name] = []byte("content")
}
createTestManifest(t, fs, testManifestPath, files)
createFilesOnDisk(t, fs, files)
chk, err := NewChecker(&CheckerOptions{
ManifestPath: testManifestPath,
BasePath: testDataDir,
Fs: fs,
})
require.NoError(t, err)
results := make(chan Result, 200)
progress := make(chan CheckStatus, 200)
err = chk.Check(context.Background(), results, progress)
require.NoError(t, err)
// results is fully buffered and closed; no draining needed
// Count progress updates
var progressCount int
for range progress {
progressCount++
}
// Should be far fewer than 100 (rate-limited to once per second)
// At minimum we get the final update
assert.GreaterOrEqual(t, progressCount, 1,
"should get at least the final progress update")
assert.Less(t, progressCount, 100,
"progress should be rate-limited, not one per file")
}