Show repository sizes in list

`list -a` had sizes because archives are files; repositories are
directories, and a long listing reports the inode size for those -- 4096
for every single one. Taking that number would have filled the column
with the same meaningless value, so the real disk usage is asked of `du`
instead, appended to the same remote command so it still costs one round
trip.

The column is dropped entirely when no usable sizes come back, rather
than showing a column of zeroes, so a server without a working `du`
degrades to the previous output. The summary line carries the total.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

README.md

@@ -134,14 +134,20 @@ worked-on project sits closest to the prompt:
 
 ```
 < src > list
-  Betaflight3.0.0  Sep 28  2016
-  website          Mar  3  2024
-  notes            Jan  3 14:32
-3 repositories
+  Betaflight3.0.0  Sep 28  2016     181M
+  website          Mar  3  2024     2.1M
+  notes            Jan  3 14:32     876K
+3 repositories · 184M
 ```
 
-`list -a` lists the archives instead, with their sizes; a pattern filters by
-name (`list note`).
+The size is the repository's real disk usage on the server, asked of `du` in
+the same round trip as the listing — a long listing reports the inode size for
+a directory, which is the same number for every repository and says nothing. If
+the server produces no usable sizes the column is left out rather than filled
+with zeroes.
+
+`list -a` lists the archives instead, whose sizes come from the listing itself;
+a pattern filters by name (`list note`).
 
 ### Overview
 

commands.go

@@ -24,6 +24,45 @@ var (
 	wsRe      = regexp.MustCompile(`\s+`)
 )
 
+// listMarker separates the two sections of the combined listing command, so
+// `list` gets both the long listing and the disk usage in one round trip.
+const listMarker = "---mgsh---"
+
+// duRe matches one `du -sk` line: kilobytes, then the path.
+var duRe = regexp.MustCompile(`^(\d+)\s+(.*)$`)
+
+// splitAtMarker divides the remote output into the part before and after the
+// marker line. Everything is in the first section when the marker is absent —
+// which is what happens when only a plain listing was asked for.
+func splitAtMarker(lines []string, marker string) (before, after []string) {
+	for i, ln := range lines {
+		if strings.TrimSpace(ln) == marker {
+			return lines[:i], lines[i+1:]
+		}
+	}
+	return lines, nil
+}
+
+// parseDuSizes turns `du -sk` output into a name -> bytes map. A long listing
+// reports the inode size for a directory — the same number for every bare
+// repository — so this is the only way to say how large one actually is.
+// A symlinked repository reports the size of the link, not of its target.
+func parseDuSizes(lines []string) map[string]int64 {
+	out := map[string]int64{}
+	for _, ln := range lines {
+		m := duRe.FindStringSubmatch(strings.TrimRight(ln, "\r"))
+		if m == nil {
+			continue
+		}
+		kb, err := strconv.ParseInt(m[1], 10, 64)
+		if err != nil {
+			continue
+		}
+		out[strings.TrimPrefix(strings.TrimSpace(m[2]), "./")] = kb * 1024
+	}
+	return out
+}
+
 // lsEntry is one parsed entry of the server's listing.
 type lsEntry struct {
 	name string // with the ".git" / ".git.tar.gz" suffix removed
@@ -203,19 +242,34 @@ func runCommandDepth(line string, depth int) bool {
 			one, many = "archive", "archives"
 		}
 		pat := strings.ToLower(word(words, 1))
-		lines, err := sshOut("/bin/ls -ltr " + shq(path))
+		remote := "/bin/ls -ltr " + shq(path)
+		if !opt["a"] {
+			// archives are files and carry a real size; repositories are
+			// directories, whose listed size is the inode's, so ask du in the
+			// same round trip
+			remote += "; echo " + shq(listMarker) + "; du -sk *.git 2>/dev/null"
+		}
+		lines, err := sshOut(remote)
 		if err != nil {
 			errorln("could not list " + many + " on the git server")
 			break
 		}
+		lsLines, duLines := splitAtMarker(lines, listMarker)
+		sizes := parseDuSizes(duLines)
+
 		var entries []lsEntry
-		for _, ln := range lines {
+		var total int64
+		for _, ln := range lsLines {
 			e, ok := parseLsEntry(ln, suffix)
 			// the pattern filters the name, not the whole listing line — an
 			// accidental match on the date or the owner helps nobody
 			if !ok || (pat != "" && !strings.Contains(strings.ToLower(e.name), pat)) {
 				continue
 			}
+			if !opt["a"] {
+				e.size = sizes[e.name+suffix] // 0 when du said nothing
+			}
+			total += e.size
 			entries = append(entries, e)
 		}
 		if len(entries) == 0 {
@@ -226,12 +280,18 @@ func runCommandDepth(line string, depth int) bool {
 			fmt.Println(col(cGray, what))
 			break
 		}
-		fmt.Print(formatRepoList(entries, opt["a"]))
+		// no size column when the server gave no usable sizes, rather than a
+		// column of zeroes
+		fmt.Print(formatRepoList(entries, total > 0))
 		label := many
 		if len(entries) == 1 {
 			label = one
 		}
-		fmt.Println(col(cGray, fmt.Sprintf("%d %s", len(entries), label)))
+		summary := fmt.Sprintf("%d %s", len(entries), label)
+		if total > 0 {
+			summary += " · " + humanSize(total)
+		}
+		fmt.Println(col(cGray, summary))
 
 	case "show": // show a repository's log directly on the server
 		prj := PRJ

mgsh_test.go

@@ -943,3 +943,45 @@ func TestExpandAlias(t *testing.T) {
 		}
 	}
 }
+
+func TestSplitAtMarker(t *testing.T) {
+	lines := []string{"a", "b", "---mgsh---", "c", "d"}
+	before, after := splitAtMarker(lines, "---mgsh---")
+	if strings.Join(before, ",") != "a,b" || strings.Join(after, ",") != "c,d" {
+		t.Errorf("split = %v / %v", before, after)
+	}
+	// no marker: everything is the first section, so a server that produced no
+	// du output simply yields no sizes
+	before, after = splitAtMarker([]string{"a", "b"}, "---mgsh---")
+	if strings.Join(before, ",") != "a,b" || after != nil {
+		t.Errorf("split without marker = %v / %v", before, after)
+	}
+}
+
+func TestParseDuSizes(t *testing.T) {
+	lines := []string{
+		"185432\tBetaflight3.0.0.git",
+		"2144\twebsite.git",
+		"876   spaced-with-blanks.git", // some du implementations use spaces
+		"1024\t./with-dot-slash.git",
+		"1500\tmy project.git", // a name with a space survives
+		"garbage",
+		"",
+	}
+	got := parseDuSizes(lines)
+	want := map[string]int64{
+		"Betaflight3.0.0.git":    185432 * 1024,
+		"website.git":            2144 * 1024,
+		"spaced-with-blanks.git": 876 * 1024,
+		"with-dot-slash.git":     1024 * 1024,
+		"my project.git":         1500 * 1024,
+	}
+	if len(got) != len(want) {
+		t.Fatalf("parseDuSizes = %v, want %d entries", got, len(want))
+	}
+	for k, v := range want {
+		if got[k] != v {
+			t.Errorf("parseDuSizes[%q] = %d, want %d", k, got[k], v)
+		}
+	}
+}

version.txt

@@ -1 +1 @@
-4.0.29
+4.0.31