Add scan --exclude PATH (closes #163)

The walk neither enters nor records an --exclude path or anything
below it, and prints "scan: skipped PATH (--exclude)" for each one it
reaches. The flag repeats. The walk knows each --exclude path by
device ID and inode number, as it knows the database files, so it also
leaves one out when it reaches it by another path, such as a hard
link. The records below an excluded path are kept, as for a default
exclude. An empty value is a usage error. An --exclude path whose real
path scan cannot find, or an operand at or below one, is a fatal error
before the scan starts. walkRules names its two lists excludePaths and
defaultExcludes. printDatabasePath, split out of runScan, keeps runScan
under the lint length limit.

Model: opus-5-5

README.md

@@ -848,8 +848,7 @@ The summary line on stderr gives:
 - the records read;
 - the duplicate-tree groups;
 - the dupe trees, which is the number of rows;
-- the reclaimable bytes, the sum of `size` over all rows, leaving out a row
+- the reclaimable bytes, the sum of `size` over all rows.
-  whose `dupe` lies inside another row's `dupe`.
 
 ```
 trees: 14 records read, 1 duplicate tree groups, 1 dupe trees, 3.0 KiB reclaimable

TODO.md

@@ -29,10 +29,6 @@
 Newest first. An entry ends with its date and its issue or branch, where those
 are known.
 
-- The reclaimable total of `trees` is the sum of `size` over all rows, leaving
-  out a row whose `dupe` lies inside another row's `dupe`. Before, the bytes of
-  such a row were counted twice (2026-11-11,
-  https://git.eeqj.de/sneak/sfdupes/issues/168)
 - New `scan` flag `--exclude PATH`, which may be repeated. The walk neither
   enters nor records an excluded path, and keeps the records below it. An
   operand at or below one is a fatal error (2026-11-11,

internal/sfdupes/trees.go

@@ -101,7 +101,7 @@ func runTrees(ctx context.Context, stdout io.Writer) error {
 		label: treesPhase("write"), unit: "groups", total: int64(len(dupes)),
 		stalledOn: "out",
 	})
-	dupeTrees, err := writeTreeGroups(stdout, dupes, prog)
+	dupeTrees, reclaimable, err := writeTreeGroups(stdout, dupes, prog)
 
 	prog.finish()
 
@@ -112,7 +112,7 @@ func runTrees(ctx context.Context, stdout io.Writer) error {
 	fmt.Fprintf(os.Stderr,
 		"trees: %d records read, %d duplicate tree groups, %d dupe trees, "+
 			"%s reclaimable\n",
-		records, len(dupes), dupeTrees, humanBytes(reclaimableBytes(dupes)))
+		records, len(dupes), dupeTrees, humanBytes(reclaimable))
 
 	return nil
 }
@@ -146,19 +146,22 @@ func readTree(ctx context.Context, db *sql.DB,
 // a row for every member of each group but its first, counting each
 // group on prog once its rows are written. The disk_size column is the
 // dupe's own. stdout may be the terminal prog is drawn on, so the lines
-// go through prog's writer. It returns the number of rows (dupe trees).
+// go through prog's writer. It returns the number of rows (dupe trees)
+// and the bytes they could reclaim.
 func writeTreeGroups(stdout io.Writer, dupes [][]*treeNode,
 	prog *progress,
-) (int, error) {
+) (int, int64, error) {
 	out := prog.writer(stdout)
 
 	_, err := fmt.Fprintln(out, "first\tdupe\tfiles\tsize\tdisk_size")
 	if err != nil {
-		return 0, fmt.Errorf("write stdout: %w", err)
+		return 0, 0, fmt.Errorf("write stdout: %w", err)
 	}
 
 	dupeTrees := 0
 
+	var reclaimable int64
+
 	for _, g := range dupes {
 		first := g[0]
 		for _, n := range g[1:] {
@@ -166,10 +169,11 @@ func writeTreeGroups(stdout io.Writer, dupes [][]*treeNode,
 				escapePath(first.path), escapePath(n.path),
 				first.fileCount, first.totalSize, n.diskSize)
 			if err != nil {
-				return 0, fmt.Errorf("write stdout: %w", err)
+				return 0, 0, fmt.Errorf("write stdout: %w", err)
 			}
 
 			dupeTrees++
+			reclaimable += first.totalSize
 		}
 
 		prog.increment()
@@ -177,44 +181,10 @@ func writeTreeGroups(stdout io.Writer, dupes [][]*treeNode,
 
 	err = out.Flush()
 	if err != nil {
-		return 0, fmt.Errorf("write stdout: %w", err)
+		return 0, 0, fmt.Errorf("write stdout: %w", err)
 	}
 
-	return dupeTrees, nil
+	return dupeTrees, reclaimable, nil
-}
-
-// reclaimableBytes returns the sum of the size column over the rows
-// that writeTreeGroups writes for dupes, leaving out a row whose dupe
-// lies inside another row's dupe.
-func reclaimableBytes(dupes [][]*treeNode) int64 {
-	isDupe := make(map[*treeNode]bool)
-
-	for _, g := range dupes {
-		for _, n := range g[1:] {
-			isDupe[n] = true
-		}
-	}
-
-	var total int64
-
-	for n := range isDupe {
-		if !insideDupe(n, isDupe) {
-			total += n.totalSize
-		}
-	}
-
-	return total
-}
-
-// insideDupe reports whether a directory above n is in isDupe.
-func insideDupe(n *treeNode, isDupe map[*treeNode]bool) bool {
-	for p := n.parent; p != nil; p = p.parent {
-		if isDupe[p] {
-			return true
-		}
-	}
-
-	return false
 }
 
 // treeBuilder reconstructs the directory hierarchy from records added

internal/sfdupes/trees_test.go

@@ -313,7 +313,7 @@ func TestWriteTreeGroupsOnTerminal(t *testing.T) {
 		label: treesPhase("write"), unit: "groups", total: int64(len(dupes)),
 	}, terminalWidth, time.Hour, time.Now)
 
-	_, err := writeTreeGroups(os.Stderr, dupes, p)
+	_, _, err := writeTreeGroups(os.Stderr, dupes, p)
 
 	p.finish()
 
@@ -587,64 +587,3 @@ func TestCollectTreeGroupsKeepsGroupLeftWithTwoCopies(t *testing.T) {
 		t.Errorf("groups = %v, want %v", got, want)
 	}
 }
-
-// TestReclaimableBytesCountsNestedDupeOnce checks that a dupe tree
-// inside another dupe tree adds nothing to the reclaimable bytes. /B/x
-// lies in /B, so the total adds the sizes of /B and /C/x, 1500 + 1000.
-func TestReclaimableBytesCountsNestedDupeOnce(t *testing.T) {
-	t.Parallel()
-
-	fh, ft, fc := testHash("fh"), testHash("ft"), testHash("fc")
-	gh, gt, gc := testHash("gh"), testHash("gt"), testHash("gc")
-	recs := []scanRec{
-		{size: 1000, head: fh, tail: ft, content: fc, path: "/A/x/f"},
-		{size: 500, head: gh, tail: gt, content: gc, path: "/A/y/g"},
-		{size: 1000, head: fh, tail: ft, content: fc, path: "/B/x/f"},
-		{size: 500, head: gh, tail: gt, content: gc, path: "/B/y/g"},
-		{size: 1000, head: fh, tail: ft, content: fc, path: "/C/x/f"},
-	}
-
-	super, dirs := treeOf(t, recs)
-	dupes := collectTreeGroups(dirs, super, everyPathExists, nil)
-
-	wantGroups := [][]string{{"/A", "/B"}, {"/A/x", "/B/x", "/C/x"}}
-	if got := groupPaths(dupes); !slices.EqualFunc(got, wantGroups,
-		slices.Equal) {
-		t.Fatalf("groups = %v, want %v", got, wantGroups)
-	}
-
-	if got := reclaimableBytes(dupes); got != 2500 {
-		t.Errorf("reclaimableBytes = %d, want 2500", got)
-	}
-}
-
-// TestReclaimableBytesCountsDupeInsideGrandparentDupeOnce checks that a
-// dupe tree adds nothing when a dupe tree above its parent holds it.
-// /B/m/x lies in /B, and /B/m is no dupe, since the group of /A and /B
-// implies that of /A/m and /B/m. The total adds the sizes of /B and
-// /C/x, 1300 + 1000.
-func TestReclaimableBytesCountsDupeInsideGrandparentDupeOnce(t *testing.T) {
-	t.Parallel()
-
-	f, w := testHash("f"), testHash("w")
-	recs := []scanRec{
-		{size: 1000, head: f, tail: f, content: f, path: "/A/m/x/f"},
-		{size: 300, head: w, tail: w, content: w, path: "/A/m/w"},
-		{size: 1000, head: f, tail: f, content: f, path: "/B/m/x/f"},
-		{size: 300, head: w, tail: w, content: w, path: "/B/m/w"},
-		{size: 1000, head: f, tail: f, content: f, path: "/C/x/f"},
-	}
-
-	super, dirs := treeOf(t, recs)
-	dupes := collectTreeGroups(dirs, super, everyPathExists, nil)
-
-	wantGroups := [][]string{{"/A", "/B"}, {"/A/m/x", "/B/m/x", "/C/x"}}
-	if got := groupPaths(dupes); !slices.EqualFunc(got, wantGroups,
-		slices.Equal) {
-		t.Fatalf("groups = %v, want %v", got, wantGroups)
-	}
-
-	if got := reclaimableBytes(dupes); got != 2300 {
-		t.Errorf("reclaimableBytes = %d, want 2300", got)
-	}
-}