first part

bucket.go

@@ -129,16 +129,23 @@ func (b *Bucket) Bucket(name []byte) *Bucket {
 	}
 
 	// Otherwise create a bucket and cache it.
-	var child = newBucket(b.tx)
-	child.bucket = &bucket{}
-	*child.bucket = *(*bucket)(unsafe.Pointer(&v[0]))
-	b.buckets[string(name)] = &child
+	var child = b.openBucket(v)
+	b.buckets[string(name)] = child
 
 	// Save a reference to the inline page if the bucket is inline.
 	if child.root == 0 {
 		child.page = (*page)(unsafe.Pointer(&v[bucketHeaderSize]))
 	}
 
+	return child
+}
+
+// Helper method that re-interprets a sub-bucket value
+// from a parent into a Bucket
+func (b *Bucket) openBucket(value []byte) *Bucket {
+	var child = newBucket(b.tx)
+	child.bucket = &bucket{}
+	*child.bucket = *(*bucket)(unsafe.Pointer(&value[0]))
 	return &child
 }
 
@@ -354,7 +361,7 @@ func (b *Bucket) ForEach(fn func(k, v []byte) error) error {
 
 // Stat returns stats on a bucket.
 func (b *Bucket) Stats() BucketStats {
-	var s BucketStats
+	var s, subStats BucketStats
 	pageSize := b.tx.db.pageSize
 	b.forEachPage(func(p *page, depth int) {
 		if (p.flags & leafPageFlag) != 0 {
@@ -365,6 +372,13 @@ func (b *Bucket) Stats() BucketStats {
 			used += int(lastElement.pos + lastElement.ksize + lastElement.vsize)
 			s.LeafInuse += used
 			s.LeafOverflowN += int(p.overflow)
+
+			// Recurse into sub-buckets
+			for _, e := range p.leafPageElements() {
+				if e.flags&bucketLeafFlag != 0 {
+					subStats.Add(b.openBucket(e.value()).Stats())
+				}
+			}
 		} else if (p.flags & branchPageFlag) != 0 {
 			s.BranchPageN++
 			lastElement := p.branchPageElement(p.count - 1)
@@ -380,6 +394,10 @@ func (b *Bucket) Stats() BucketStats {
 	})
 	s.BranchAlloc = (s.BranchPageN + s.BranchOverflowN) * pageSize
 	s.LeafAlloc = (s.LeafPageN + s.LeafOverflowN) * pageSize
+
+	// add the max depth of sub-buckets to get total nested depth
+	s.Depth += subStats.Depth
+	s.Add(subStats)
 	return s
 }
 
@@ -643,6 +661,21 @@ type BucketStats struct {
 	LeafInuse   int // bytes actually used for leaf data
 }
 
+func (s *BucketStats) Add(other BucketStats) {
+	s.BranchPageN += other.BranchPageN
+	s.BranchOverflowN += other.BranchOverflowN
+	s.LeafPageN += other.LeafPageN
+	s.LeafOverflowN += other.LeafOverflowN
+	s.KeyN += s.KeyN
+	if s.Depth < other.Depth {
+		s.Depth = other.Depth
+	}
+	s.BranchAlloc += other.BranchAlloc
+	s.BranchInuse += other.BranchInuse
+	s.LeafAlloc += other.LeafAlloc
+	s.LeafInuse += other.LeafInuse
+}
+
 // cloneBytes returns a copy of a given slice.
 func cloneBytes(v []byte) []byte {
 	var clone = make([]byte, len(v))

cmd/bolt/main.go

@@ -91,6 +91,14 @@ func NewApp() *cli.App {
 				Check(path)
 			},
 		},
+		{
+			Name:  "stats",
+			Usage: "Retrieve statistics for a bucket (aggregated recursively)",
+			Action: func(c *cli.Context) {
+				path, name := c.Args().Get(0), c.Args().Get(1)
+				Stats(path, name)
+			},
+		},
 		{
 			Name:  "bench",
 			Usage: "Performs a synthetic benchmark",

cmd/bolt/stats.go

@@ -0,0 +1,54 @@
+package main
+
+import (
+	"os"
+
+	"github.com/boltdb/bolt"
+)
+
+// Keys retrieves a list of keys for a given bucket.
+func Stats(path, name string) {
+	if _, err := os.Stat(path); os.IsNotExist(err) {
+		fatal(err)
+		return
+	}
+
+	db, err := bolt.Open(path, 0600)
+	if err != nil {
+		fatal(err)
+		return
+	}
+	defer db.Close()
+
+	err = db.View(func(tx *bolt.Tx) error {
+		// Find bucket.
+		b := tx.Bucket([]byte(name))
+		if b == nil {
+			fatalf("bucket not found: %s", name)
+			return nil
+		}
+
+		// Iterate over each key.
+		s := b.Stats()
+		println("Page count statistics")
+		printf("\tNumber of logical branch pages: %d\n", s.BranchPageN)
+		printf("\tNumber of physical branch overflow pages: %d\n", s.BranchOverflowN)
+		printf("\tNumber of logical leaf pages: %d\n", s.LeafPageN)
+		printf("\tNumber of physical leaf overflow pages: %d\n", s.LeafOverflowN)
+
+		println("Tree statistics")
+		printf("\tNumber of keys/value pairs: %d\n", s.KeyN)
+		printf("\tNumber of levels in B+tree: %d\n", s.Depth)
+
+		println("Page size utilization")
+		printf("\tBytes allocated for physical branch pages: %d\n", s.BranchAlloc)
+		printf("\tBytes actually used for branch data: %d\n", s.BranchInuse)
+		printf("\tBytes allocated for physical leaf pages: %d\n", s.LeafAlloc)
+		printf("\tBytes actually used for leaf data: %d\n", s.LeafInuse)
+		return nil
+	})
+	if err != nil {
+		fatal(err)
+		return
+	}
+}

page.go

@@ -63,7 +63,7 @@ func (p *page) leafPageElement(index uint16) *leafPageElement {
 
 // leafPageElements retrieves a list of leaf nodes.
 func (p *page) leafPageElements() []leafPageElement {
-	return ((*[maxNodesPerPage]leafPageElement)(unsafe.Pointer(&p.ptr)))[:]
+	return ((*[maxNodesPerPage]leafPageElement)(unsafe.Pointer(&p.ptr)))[:p.count]
 }
 
 // branchPageElement retrieves the branch node by index
@@ -73,7 +73,7 @@ func (p *page) branchPageElement(index uint16) *branchPageElement {
 
 // branchPageElements retrieves a list of branch nodes.
 func (p *page) branchPageElements() []branchPageElement {
-	return ((*[maxNodesPerPage]branchPageElement)(unsafe.Pointer(&p.ptr)))[:]
+	return ((*[maxNodesPerPage]branchPageElement)(unsafe.Pointer(&p.ptr)))[:p.count]
 }
 
 // dump writes n bytes of the page to STDERR as hex output.