From 3500f9490c976af6dde00f40d449747f6d9b69df Mon Sep 17 00:00:00 2001
From: Michael Eischer <michael.eischer@fau.de>
Date: Fri, 6 Nov 2020 23:41:04 +0100
Subject: [PATCH] check: Simplify blob status tracking

UnusedBlobs now directly reads the list of existing blobs from the
repository index. This removes the need for the blobStatusExists flag,
which in turn allows converting the blobRefs map into a BlobSet.
---
 cmd/restic/cmd_check.go          |  2 +-
 internal/checker/checker.go      | 36 +++++++++++++-------------------
 internal/checker/checker_test.go |  2 +-
 internal/checker/testing.go      |  2 +-
 4 files changed, 18 insertions(+), 24 deletions(-)

diff --git a/cmd/restic/cmd_check.go b/cmd/restic/cmd_check.go
index 529b2d71d..84eaa5ac4 100644
--- a/cmd/restic/cmd_check.go
+++ b/cmd/restic/cmd_check.go
@@ -255,7 +255,7 @@ func runCheck(opts CheckOptions, gopts GlobalOptions, args []string) error {
 	}
 
 	if opts.CheckUnused {
-		for _, id := range chkr.UnusedBlobs() {
+		for _, id := range chkr.UnusedBlobs(gopts.ctx) {
 			Verbosef("unused blob %v\n", id)
 			errorsFound = true
 		}
diff --git a/internal/checker/checker.go b/internal/checker/checker.go
index 6466693c4..09a71f9de 100644
--- a/internal/checker/checker.go
+++ b/internal/checker/checker.go
@@ -25,8 +25,7 @@ type Checker struct {
 	packs    map[restic.ID]int64
 	blobRefs struct {
 		sync.Mutex
-		// see flags below
-		M map[restic.BlobHandle]blobStatus
+		M restic.BlobSet
 	}
 
 	masterIndex *repository.MasterIndex
@@ -34,13 +33,6 @@ type Checker struct {
 	repo restic.Repository
 }
 
-type blobStatus uint8
-
-const (
-	blobStatusExists blobStatus = 1 << iota
-	blobStatusReferenced
-)
-
 // New returns a new checker which runs on repo.
 func New(repo restic.Repository) *Checker {
 	c := &Checker{
@@ -49,7 +41,7 @@ func New(repo restic.Repository) *Checker {
 		repo:        repo,
 	}
 
-	c.blobRefs.M = make(map[restic.BlobHandle]blobStatus)
+	c.blobRefs.M = restic.NewBlobSet()
 
 	return c
 }
@@ -162,8 +154,6 @@ func (c *Checker) LoadIndex(ctx context.Context) (hints []error, errs []error) {
 			debug.Log("process blobs")
 			cnt := 0
 			for blob := range res.Index.Each(wgCtx) {
-				h := restic.BlobHandle{ID: blob.ID, Type: blob.Type}
-				c.blobRefs.M[h] = blobStatusExists
 				cnt++
 
 				if _, ok := packToIndex[blob.PackID]; !ok {
@@ -529,9 +519,9 @@ func (c *Checker) filterTrees(ctx context.Context, backlog restic.IDs, loaderCha
 			// even when a file references a tree blob
 			c.blobRefs.Lock()
 			h := restic.BlobHandle{ID: nextTreeID, Type: restic.TreeBlob}
-			status := c.blobRefs.M[h]
+			blobReferenced := c.blobRefs.M.Has(h)
 			c.blobRefs.Unlock()
-			if (status & blobStatusReferenced) != 0 {
+			if blobReferenced {
 				continue
 			}
 
@@ -552,7 +542,7 @@ func (c *Checker) filterTrees(ctx context.Context, backlog restic.IDs, loaderCha
 			loadCh = nil
 			c.blobRefs.Lock()
 			h := restic.BlobHandle{ID: nextTreeID, Type: restic.TreeBlob}
-			c.blobRefs.M[h] |= blobStatusReferenced
+			c.blobRefs.M.Insert(h)
 			c.blobRefs.Unlock()
 
 		case j, ok := <-inCh:
@@ -687,7 +677,7 @@ func (c *Checker) checkTree(id restic.ID, tree *restic.Tree) (errs []error) {
 	for _, blobID := range blobs {
 		c.blobRefs.Lock()
 		h := restic.BlobHandle{ID: blobID, Type: restic.DataBlob}
-		c.blobRefs.M[h] |= blobStatusReferenced
+		c.blobRefs.M.Insert(h)
 		debug.Log("blob %v is referenced", blobID)
 		c.blobRefs.Unlock()
 	}
@@ -696,15 +686,19 @@ func (c *Checker) checkTree(id restic.ID, tree *restic.Tree) (errs []error) {
 }
 
 // UnusedBlobs returns all blobs that have never been referenced.
-func (c *Checker) UnusedBlobs() (blobs restic.BlobHandles) {
+func (c *Checker) UnusedBlobs(ctx context.Context) (blobs restic.BlobHandles) {
 	c.blobRefs.Lock()
 	defer c.blobRefs.Unlock()
 
 	debug.Log("checking %d blobs", len(c.blobRefs.M))
-	for id, flags := range c.blobRefs.M {
-		if (flags & blobStatusReferenced) == 0 {
-			debug.Log("blob %v not referenced", id)
-			blobs = append(blobs, id)
+	ctx, cancel := context.WithCancel(ctx)
+	defer cancel()
+
+	for blob := range c.repo.Index().Each(ctx) {
+		h := restic.BlobHandle{ID: blob.ID, Type: blob.Type}
+		if !c.blobRefs.M.Has(h) {
+			debug.Log("blob %v not referenced", h)
+			blobs = append(blobs, h)
 		}
 	}
 
diff --git a/internal/checker/checker_test.go b/internal/checker/checker_test.go
index 8218a94fc..333e5573b 100644
--- a/internal/checker/checker_test.go
+++ b/internal/checker/checker_test.go
@@ -181,7 +181,7 @@ func TestUnreferencedBlobs(t *testing.T) {
 	test.OKs(t, checkPacks(chkr))
 	test.OKs(t, checkStruct(chkr))
 
-	blobs := chkr.UnusedBlobs()
+	blobs := chkr.UnusedBlobs(context.TODO())
 	sort.Sort(blobs)
 
 	test.Equals(t, unusedBlobsBySnapshot, blobs)
diff --git a/internal/checker/testing.go b/internal/checker/testing.go
index 8e49b562c..c6d9a08ef 100644
--- a/internal/checker/testing.go
+++ b/internal/checker/testing.go
@@ -37,7 +37,7 @@ func TestCheckRepo(t testing.TB, repo restic.Repository) {
 	}
 
 	// unused blobs
-	blobs := chkr.UnusedBlobs()
+	blobs := chkr.UnusedBlobs(context.TODO())
 	if len(blobs) > 0 {
 		t.Errorf("unused blobs found: %v", blobs)
 	}