Cleanup

Author: korthaj

README.md

@@ -4,7 +4,7 @@ Golang Bloom filter implementation
 
 ![Neutral density filter](ND-filter.jpg)
 
-*Neutral density filter, image by [Robert Emperley][re], [CC BY-SA 2.0][ccbysa].*
+*Image by [Robert Emperley][re], [CC BY-SA 2.0][ccbysa].*
 
 ### Installation
 

example_test.go

@@ -26,7 +26,7 @@ func Example_basics() {
 	// Output: https://rascal.com seems to be shady.
 }
 
-// Count the number of false positives.
+// Estimate the number of false positives.
 func Example_falsePositives() {
 	// Create a Bloom filter with room for n elements
 	// at a false-positives rate less than 1/p.
@@ -51,8 +51,8 @@ func Example_falsePositives() {
 }
 
 // Compute the union of two filters.
-func ExampleFilter_Or() {
-	// Create two Bloom filter with room for 1000 elements
+func ExampleFilter_Union() {
+	// Create two Bloom filters, each with room for 1000 elements
 	// at a false-positives rate less than 1/100.
 	n, p := 1000, 100
 	f1 := bloom.New(n, p)

filter.go

@@ -2,8 +2,10 @@
 //
 // Bloom filters
 //
-// A Bloom filter is a space-efficient probabilistic data structure
-// used to test set membership. A member test returns either
+// A Bloom filter is a fast and space-efficient probabilistic data structure
+// used to test set membership.
+//
+// A membership test returns either
 // ”likely member” or ”definitely not a member”. Only false positives
 // can occur: an element that has been added to the filter
 // will be identified as ”likely member”.
@@ -31,8 +33,8 @@
 //
 // This implementation is not intended for cryptographic use.
 // Each membership test makes a single call to a 128-bit MurmurHash3 function.
-// This saves on hashing without increasing the false-positives
-// probability as shown by Kirsch and Mitzenmacher.
+// This improves speed without increasing the false-positives rate
+// as shown by Kirsch and Mitzenmacher.
 //
 package bloom
 
@@ -49,23 +51,24 @@ const (
 type Filter struct {
 	data    []uint64 // Bit array, the length is a power of 2.
 	lookups int      // Lookups per query
-	count   int64    // Estimated number of unique elements
+	count   int64    // Estimate number of elements
 }
 
+// MurmurHash3 function.
 var murmur = new(digest)
 
 // New creates an empty Bloom filter with room for n elements
 // at a false-positives rate less than 1/p.
 func New(n int, p int) *Filter {
-	f := &Filter{}
 	minWords := int(0.0325 * math.Log(float64(p)) * float64(n))
 	words := 1
 	for words < minWords {
 		words *= 2
 	}
-	f.data = make([]uint64, words)
-	f.lookups = int(1.4*math.Log(float64(p)) + 1)
-	return f
+	return &Filter{
+		data:    make([]uint64, words),
+		lookups: int(1.4*math.Log(float64(p)) + 1),
+	}
 }
 
 // AddByte adds b to the filter and tells if b was already a likely member.
@@ -95,7 +98,7 @@ func (f *Filter) Add(s string) bool {
 	return f.AddByte(b)
 }
 
-// TestByte tells if b is a likely member of this filter.
+// TestByte tells if b is a likely member of the filter.
 func (f *Filter) TestByte(b []byte) bool {
 	h1, h2 := murmur.hash(b)
 	trunc := uint64(len(f.data))<<shift - 1
@@ -110,14 +113,14 @@ func (f *Filter) TestByte(b []byte) bool {
 	return true
 }
 
-// Test tells if s is a likely member of this filter.
+// Test tells if s is a likely member of the filter.
 func (f *Filter) Test(s string) bool {
 	b := make([]byte, len(s))
 	copy(b, s)
 	return f.TestByte(b)
 }
 
-// Count returns an estimate of the number of elements in this filter.
+// Count returns an estimate of the number of elements in the filter.
 func (f *Filter) Count() int64 {
 	return f.count
 }

filter_test.go

@@ -48,7 +48,7 @@ func TestFilter(t *testing.T) {
 
 			member = filter.Add(s3)
 			if member {
-				t.Errorf("Add(s1) = %v; want false\n", member)
+				t.Errorf("Add(s3) = %v; want false\n", member)
 			}
 			count = filter.Count()
 			if count != 2 {
@@ -72,15 +72,15 @@ func TestUnion(t *testing.T) {
 			or := f1.Union(f2)
 			member := or.Test(s1)
 			if !member {
-				t.Errorf("or.Test(s1) = %v; want true\n", member)
+				t.Errorf("f1.Union(f2).Test(s1) = %v; want true\n", member)
 			}
 			member = or.Test(s2)
 			if !member {
-				t.Errorf("or.Test(s2) = %v; want true\n", member)
+				t.Errorf("f1.Union(f2).Test(s2) = %v; want true\n", member)
 			}
 			member = or.Test(s3)
 			if !member {
-				t.Errorf("or.Test(s3) = %v; want true\n", member)
+				t.Errorf("f1.Union(f2).Test(s3) = %v; want true\n", member)
 			}
 		}
 	}
@@ -93,7 +93,7 @@ func BenchmarkAdd(b *testing.B) {
 	filter := New(1<<30, 200)
 	b.StartTimer()
 	for i := 0; i < b.N; i++ {
-		filter.Add(fox)
+		_ = filter.Add(fox)
 	}
 }
 
@@ -103,7 +103,7 @@ func BenchmarkAddByte(b *testing.B) {
 	b.StartTimer()
 	bytes := []byte(fox)
 	for i := 0; i < b.N; i++ {
-		filter.AddByte(bytes)
+		_ = filter.AddByte(bytes)
 	}
 }
 
@@ -112,7 +112,7 @@ func BenchmarkTest(b *testing.B) {
 	filter := New(1<<30, 200)
 	b.StartTimer()
 	for i := 0; i < b.N; i++ {
-		filter.Test(fox)
+		_ = filter.Test(fox)
 	}
 }
 
@@ -122,7 +122,7 @@ func BenchmarkTestByte(b *testing.B) {
 	b.StartTimer()
 	bytes := []byte(fox)
 	for i := 0; i < b.N; i++ {
-		filter.TestByte(bytes)
+		_ = filter.TestByte(bytes)
 	}
 }
 

hash_test.go

@@ -7,9 +7,8 @@ import (
 func TestHash(t *testing.T) {
 	d := new(digest)
 	var data = []struct {
-		h1 uint64
-		h2 uint64
-		s  string
+		h1, h2 uint64
+		s      string
 	}{
 		{0x0000000000000000, 0x0000000000000000, ""},
 		{0xcbd8a7b341bd9b02, 0x5b1e906a48ae1d19, "hello"},
@@ -23,7 +22,7 @@ func TestHash(t *testing.T) {
 			t.Errorf("hash(%q).h1 = %d; want %d\n", x.s, h1, x.h1)
 		}
 		if h2 != x.h2 {
-			t.Errorf("hash(%q).h1 = %d; want %d\n", x.s, h2, x.h2)
+			t.Errorf("hash(%q).h2 = %d; want %d\n", x.s, h2, x.h2)
 		}
 	}
 }