Merge remote-tracking branch 'origin/main' into merge-check

ingestion/internal/brainfs/brainfs.go

@@ -104,6 +104,27 @@ func (b *Brain) ReadDirRegular(relDir string) ([]string, error) {
 	return names, nil
 }
 
+// ReadDirDirs lists the real subdirectories directly in a brain-relative
+// directory, by name. Every component of relDir must be a real directory
+// (ErrNotRegular otherwise); symlinks to directories are left out.
+func (b *Brain) ReadDirDirs(relDir string) ([]string, error) {
+	relDir = filepath.ToSlash(relDir)
+	if err := b.realDir(relDir); err != nil {
+		return nil, err
+	}
+	ents, err := fs.ReadDir(b.root.FS(), relDir)
+	if err != nil {
+		return nil, err
+	}
+	var names []string
+	for _, e := range ents {
+		if e.Type().IsDir() {
+			names = append(names, e.Name())
+		}
+	}
+	return names, nil
+}
+
 // ReadRegular reads a brain-relative path, refusing anything that is not a
 // regular file, both before the open and on the opened handle.
 func (b *Brain) ReadRegular(rel string) ([]byte, error) {

ingestion/internal/brainfs/brainfs_test.go

@@ -262,3 +262,21 @@ func TestReadRegularDoesNotReturnAnInBrainFileSwappedInBehindASymlink(t *testing
 	assert.Zero(t, leaks)
 	assert.Positive(t, ok, "positive control: the regular file is read")
 }
+
+func TestReadDirDirsListsOnlyRealSubdirsAndRefusesSymlinkedComponents(t *testing.T) {
+	dir := t.TempDir()
+	require.NoError(t, os.MkdirAll(filepath.Join(dir, "a", "real"), 0o755))
+	require.NoError(t, os.MkdirAll(filepath.Join(dir, "other"), 0o755))
+	require.NoError(t, os.Symlink(filepath.Join("..", "other"), filepath.Join(dir, "a", "linked")))
+	require.NoError(t, os.WriteFile(filepath.Join(dir, "a", "f.md"), []byte("f"), 0o644))
+	require.NoError(t, os.Symlink("a", filepath.Join(dir, "alias")))
+	b, err := Open(dir)
+	require.NoError(t, err)
+	defer func() { _ = b.Close() }()
+
+	names, err := b.ReadDirDirs("a")
+	require.NoError(t, err)
+	assert.Equal(t, []string{"real"}, names)
+	_, err = b.ReadDirDirs("alias")
+	assert.ErrorIs(t, err, ErrNotRegular)
+}

ingestion/internal/brainfs/brainfstest/brainfstest.go

@@ -29,7 +29,7 @@ const outsideContent = "---\ntitle: " + Sentinel + "\naliases:\n  - " + Sentinel
 	"\n\n" + Sentinel + " [[" + Sentinel + "]]\n"
 
 // Outside writes a file outside any brain dir containing Sentinel.
-func Outside(t *testing.T) string {
+func Outside(t testing.TB) string {
 	t.Helper()
 	p := filepath.Join(t.TempDir(), "secret.md")
 	if err := os.WriteFile(p, []byte(outsideContent), 0o644); err != nil {
@@ -44,9 +44,12 @@ var swapSeq atomic.Int64
 // a symlink to outside or a FIFO into path, until the test ends. FIFOs get a
 // writer held open until the swapper exits, so a reader that skips the
 // handle check parks forever instead of reading EOF. The swapper is stopped
-// and awaited before t.TempDir cleanup.
-func StartSwap(t *testing.T, path, inside, outside string, fifo bool) {
+// and awaited before t.TempDir cleanup. It returns the number of symlinks
+// or FIFOs actually renamed into path so far, so a caller can tell a race
+// that ran from one whose every swap failed (hyperguild#177).
+func StartSwap(t testing.TB, path, inside, outside string, fifo bool) *atomic.Int64 {
 	t.Helper()
+	var swaps atomic.Int64
 	var stop atomic.Bool
 	var wg sync.WaitGroup
 	wg.Add(1)
@@ -69,22 +72,26 @@ func StartSwap(t *testing.T, path, inside, outside string, fifo bool) {
 		}()
 		for i := 0; !stop.Load(); i++ {
 			_ = os.Remove(tmp)
+			var made error
 			switch {
 			case i%2 == 0:
 				_ = os.WriteFile(tmp, []byte(inside), 0o644)
 			case fifo:
-				if err := syscall.Mkfifo(tmp, 0o644); err == nil {
+				if made = syscall.Mkfifo(tmp, 0o644); made == nil {
 					if w, err := os.OpenFile(tmp, os.O_RDWR, 0); err == nil {
 						writers = append(writers, w)
 					}
 				}
 			default:
-				_ = os.Symlink(link, tmp)
+				made = os.Symlink(link, tmp)
+			}
+			if err := os.Rename(tmp, path); err == nil && made == nil && i%2 == 1 {
+				swaps.Add(1)
 			}
-			_ = os.Rename(tmp, path)
 		}
 	}()
 	t.Cleanup(func() { stop.Store(true); wg.Wait() })
+	return &swaps
 }
 
 // WriteSentinelAt writes a file carrying Sentinel at p (creating its
@@ -109,18 +116,41 @@ func AssertNoLeak(t *testing.T, path, inside string, d time.Duration, read func(
 	AssertNoLeakFrom(t, path, inside, Outside(t), d, read)
 }
 
+// InsideMarker must appear in the inside note's content and in read's
+// output whenever read saw that note: the floor below counts such reads.
+const InsideMarker = "inside"
+
 // AssertNoLeakFrom is AssertNoLeak with the symlink pointing at target, e.g.
 // a file elsewhere in the brain written with WriteSentinelAt.
-func AssertNoLeakFrom(t *testing.T, path, inside, target string, d time.Duration, read func() string) {
+//
+// Beyond "no Sentinel in any output" it sets two floors, because without
+// them the assertion passes when nothing was tested (hyperguild#177): at
+// least one symlink must have been swapped into path, and at least one read
+// must have returned InsideMarker (a reader that always errors returns
+// nothing, and so never leaks).
+func AssertNoLeakFrom(t testing.TB, path, inside, target string, d time.Duration, read func() string) {
 	t.Helper()
-	StartSwap(t, path, inside, target, false)
-	leaks, iters := 0, 0
+	if !strings.Contains(inside, InsideMarker) {
+		t.Fatalf("inside content must contain %q", InsideMarker)
+	}
+	swaps := StartSwap(t, path, inside, target, false)
+	leaks, iters, insideReads := 0, 0, 0
 	for deadline := time.Now().Add(d); time.Now().Before(deadline); iters++ {
-		leaks += strings.Count(read(), Sentinel)
+		out := read()
+		leaks += strings.Count(out, Sentinel)
+		if strings.Contains(out, InsideMarker) {
+			insideReads++
+		}
 	}
 	if leaks > 0 {
 		t.Fatalf("symlink target %s leaked %d times in %d iterations", target, leaks, iters)
 	}
+	if n := swaps.Load(); n == 0 {
+		t.Fatalf("no symlink was swapped into %s in %d iterations: the race did not run", path, iters)
+	}
+	if insideReads == 0 {
+		t.Fatalf("no read in %d iterations returned the inside note (%q): the reader read nothing", iters, InsideMarker)
+	}
 }
 
 // AssertNoHang loops read for d while a FIFO swapper (held writers) runs on

ingestion/internal/brainfs/brainfstest/brainfstest_test.go

@@ -0,0 +1,79 @@
+package brainfstest
+
+import (
+	"fmt"
+	"os"
+	"path/filepath"
+	"runtime"
+	"syscall"
+	"testing"
+	"time"
+)
+
+// recordingTB stands in for the test under AssertNoLeakFrom so a fatal
+// there can be observed instead of failing this test. Fatalf ends the
+// calling goroutine the way testing.T does.
+type recordingTB struct {
+	testing.TB
+	failed bool
+	msg    string
+}
+
+func (r *recordingTB) Fatalf(format string, args ...any) {
+	r.failed, r.msg = true, fmt.Sprintf(format, args...)
+	runtime.Goexit()
+}
+
+func runAssert(t *testing.T, path, target string, read func() string) *recordingTB {
+	t.Helper()
+	r := &recordingTB{TB: t}
+	done := make(chan struct{})
+	go func() {
+		defer close(done)
+		AssertNoLeakFrom(r, path, "inside", target, 200*time.Millisecond, read)
+	}()
+	<-done
+	return r
+}
+
+// hyperguild#177: with every swap failing, no leak can occur, so the leak
+// assertion passed without testing anything.
+func TestAssertNoLeakFromFailsWhenNoSymlinkSwapSucceeds(t *testing.T) {
+	path := filepath.Join(t.TempDir(), "missing-dir", "r.md")
+	r := runAssert(t, path, Outside(t), func() string { return "inside" })
+	if !r.failed {
+		t.Fatal("AssertNoLeakFrom passed with zero successful symlink swaps")
+	}
+}
+
+// hyperguild#177: a reader that always errors returns nothing, so it never
+// leaks either.
+func TestAssertNoLeakFromFailsWhenNoReadReturnsTheInsideNote(t *testing.T) {
+	path := filepath.Join(t.TempDir(), "r.md")
+	r := runAssert(t, path, Outside(t), func() string { return "" })
+	if !r.failed {
+		t.Fatal("AssertNoLeakFrom passed with zero reads returning the inside note")
+	}
+}
+
+// Positive control: a reader that refuses symlinks and reads regular files
+// passes, so the floors do not fail a working reader.
+func TestAssertNoLeakFromPassesAReaderThatRefusesSymlinks(t *testing.T) {
+	path := filepath.Join(t.TempDir(), "r.md")
+	read := func() string {
+		f, err := os.OpenFile(path, os.O_RDONLY|noFollow, 0)
+		if err != nil {
+			return ""
+		}
+		defer func() { _ = f.Close() }()
+		b := make([]byte, 64)
+		n, _ := f.Read(b)
+		return string(b[:n])
+	}
+	r := runAssert(t, path, Outside(t), read)
+	if r.failed {
+		t.Fatalf("a symlink-refusing reader failed: %s", r.msg)
+	}
+}
+
+const noFollow = syscall.O_NOFOLLOW