From 4af19e8a4c5c1363872ca4c37f181a19e67da4c0 Mon Sep 17 00:00:00 2001
From: Luke Shumaker <lukeshu@beefcake.parabola.nu>
Date: Sun, 3 Jun 2018 13:20:19 -0400
Subject: cow-dedupe: Shorter status line; better statistics

---
 go/src/cow-dedupe/dedupe.go | 37 +++++++++++++++++--------------------
 1 file changed, 17 insertions(+), 20 deletions(-)

diff --git a/go/src/cow-dedupe/dedupe.go b/go/src/cow-dedupe/dedupe.go
index 5717e30..3fc2254 100644
--- a/go/src/cow-dedupe/dedupe.go
+++ b/go/src/cow-dedupe/dedupe.go
@@ -176,19 +176,17 @@ func dedupe(srcFile string, dupFiles []string) error {
 	return err
 }
 
-// [ 0s ] size-set[n/d]->c :: sha256-set[t+n/d]->c :: (summed=n deduped=n) :: verb[m/n]
+// [ 0s ] set[n/(d+e)]->c :: (summed=n deduped=n) :: verb[m/n]
 type fancyStatus struct {
-	sizeN, sizeD, sizeC    int
-	sumT, sumN, sumD, sumC int
-	summed, errs           int
-	deduped                struct{ srcs, dstMaps, dstFiles int }
-	verb                   string
+	set          struct{ n, d, p, m, size int }
+	summed, errs int
+	deduped      struct{ srcs, dstMaps, dstFiles int }
+	verb         string
 }
 
 func (s fancyStatus) String() string {
-	return fmt.Sprintf("size-set[%d/%d]->%d :: sha256-set[%d+%d/%d]->%d :: (summed=%d deduped={%dmaps:%dfiles->%d} errs=%d) :: %s",
-		s.sizeN, s.sizeD, s.sizeC,
-		s.sumT, s.sumN, s.sumD, s.sumC,
+	return fmt.Sprintf("set[%d/(%d+%d-%d)]->%d :: summed=%d deduped={%dmaps:%dfiles->%d} errs=%d :: %s",
+		s.set.n, s.set.d, s.set.p, s.set.m, s.set.size,
 		s.summed,
 		s.deduped.dstMaps+s.deduped.srcs, s.deduped.dstFiles+s.deduped.srcs, s.deduped.srcs,
 		s.errs,
@@ -236,9 +234,9 @@ func main() {
 
 	sl = myStatusLine()
 	var status fancyStatus
-	status.sizeD = len(size2filenames)
+	status.set.d = len(size2filenames)
 	for _, filenames := range size2filenames {
-		status.sizeC = len(filenames)
+		status.set.size = len(filenames)
 		// Now do strict hashing, instead of the incredibly
 		// sloppy (but fast) size-bucketing.
 		status.verb = "sha256sum[%v/%v]"
@@ -248,9 +246,13 @@ func main() {
 		sl.Put(status.String())
 		pruneSingles(checksum2filenames)
 		// And loop over the smaller, precise buckets
-		status.sumD = len(checksum2filenames)
+		if len(checksum2filenames) == 0 {
+			status.set.m++
+		} else {
+			status.set.p += len(checksum2filenames) - 1
+		}
 		for _, filenames := range checksum2filenames {
-			status.sumC = len(filenames)
+			status.set.size = len(filenames)
 			status.verb = "prep"
 			sl.Put(status.String())
 			var fiemaps []string
@@ -285,7 +287,7 @@ func main() {
 			if len(name) > 21 {
 				name = name[:20] + "…"
 			}
-			status.verb = fmt.Sprintf("dedupe %q (and %d more)", name, len(dupFiles))
+			status.verb = fmt.Sprintf("dedupe %q<-[%d]file{…}", name, len(dupFiles))
 			sl.Put(status.String())
 			err := dedupe(srcFile, dupFiles)
 			if err != nil {
@@ -295,13 +297,8 @@ func main() {
 				status.deduped.dstMaps += len(fiemaps) - 1
 				status.deduped.dstFiles += len(dupFiles)
 			}
-			status.sumN++
+			status.set.n++
 		}
-		status.sumT += len(checksum2filenames)
-		status.sumN = 0
-		status.sumD = 0
-		status.sumC = 0
-		status.sizeN++
 	}
 	status.verb = "done"
 	sl.Put(status.String())
-- 
cgit v1.2.3