Merge pull request #5 from yourbasic/tip

Adds binary format

Author: korthaj

binary.go

@@ -0,0 +1,48 @@
+package bloom
+
+import (
+	"bytes"
+	"encoding/gob"
+)
+
+const binVer0 = 0
+
+// Data to be included in binary representation of Filter.
+type marshalFilter struct {
+	Version int
+	Data    []uint64
+	Lookups int
+	Count   int64
+}
+
+// MarshalBinary returns a binary representation of the filter.
+//
+// This method implements the encoding.BinaryMarshaler interface.
+// The packages encoding/gob, encoding/json, and encoding/xml
+// all check for this interface.
+func (f *Filter) MarshalBinary() ([]byte, error) {
+	mf := marshalFilter{
+		Version: binVer0,
+		Data:    f.data,
+		Lookups: f.lookups,
+		Count:   f.count,
+	}
+	var b bytes.Buffer
+	err := gob.NewEncoder(&b).Encode(mf)
+	return b.Bytes(), err
+}
+
+// UnmarshalBinary imports binary data created by MarshalBinary
+// into an empty filter. If the filter is not empty, all previous
+// entries are overwritten.
+func (f *Filter) UnmarshalBinary(data []byte) error {
+	var mf marshalFilter
+	err := gob.NewDecoder(bytes.NewBuffer(data)).Decode(&mf)
+	if err != nil {
+		return err
+	}
+	f.data = mf.Data
+	f.lookups = mf.Lookups
+	f.count = mf.Count
+	return nil
+}

example_test.go

@@ -1,8 +1,11 @@
 package bloom_test
 
 import (
+	"bytes"
+	"encoding/gob"
 	"fmt"
 	"github.com/yourbasic/bloom"
+	"log"
 	"math/rand"
 	"strconv"
 )
@@ -72,3 +75,30 @@ func ExampleFilter_Union() {
 	fmt.Println("f1 ∪ f2:", f1.Union(f2).Count())
 	// Output: f1 ∪ f2: 505
 }
+
+// Send a filter over a network using the encoding/gob package.
+func ExampleFilter_MarshalBinary_network() {
+	// Create a mock network and a new Filter.
+	var network bytes.Buffer
+	f1 := bloom.New(1000, 100)
+	f1.Add("Hello, filter!")
+
+	// Create an encoder and send the filter to the network.
+	enc := gob.NewEncoder(&network)
+	if err := enc.Encode(f1); err != nil {
+		log.Fatal("encode error:", err)
+	}
+
+	// Create a decoder and receive the filter from the network.
+	dec := gob.NewDecoder(&network)
+	var f2 bloom.Filter
+	if err := dec.Decode(&f2); err != nil {
+		log.Fatal("decode error:", err)
+	}
+
+	// Check that we got the same filter back.
+	if f2.Test("Hello, filter!") {
+		fmt.Println("Filter arrived safely.")
+	}
+	// Output: Filter arrived safely.
+}

filter.go

@@ -5,10 +5,9 @@
 // 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”.
+// 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 always be identified as ”likely member”.
 //
 // Elements can be added, but not removed. With more elements in the filter,
 // the probability of false positives increases.
@@ -97,11 +96,13 @@ func (f *Filter) add(h1, h2 uint64) bool {
 }
 
 // TestByte tells if b is a likely member of the filter.
+// If true, b is probably a member; if false, b is definitely not a member.
 func (f *Filter) TestByte(b []byte) bool {
 	return f.test(hash(b))
 }
 
 // Test tells if s is a likely member of the filter.
+// If true, s is probably a member; if false, s is definitely not a member.
 func (f *Filter) Test(s string) bool {
 	return f.test(hashString(s))
 }

filter_test.go

@@ -1,6 +1,9 @@
 package bloom
 
 import (
+	"bytes"
+	"encoding/gob"
+	"reflect"
 	"testing"
 )
 
@@ -140,6 +143,27 @@ func TestUnion(t *testing.T) {
 	}
 }
 
+func TestMarshal(t *testing.T) {
+	var network bytes.Buffer
+	f1 := New(10000, 100)
+	f1.Add("Hello, filter!")
+
+	enc := gob.NewEncoder(&network)
+	if err := enc.Encode(f1); err != nil {
+		t.Errorf("Encode->err = %v; want nil\n", err)
+	}
+
+	dec := gob.NewDecoder(&network)
+	var f2 *Filter
+	if err := dec.Decode(&f2); err != nil {
+		t.Errorf("Decode->err = %v; want nil\n", err)
+	}
+
+	if !reflect.DeepEqual(f1, f2) {
+		t.Errorf("Encode(Code(f)) = %v; want %v\n", f1, f2)
+	}
+}
+
 var fox string = "The quick brown fox jumps over the lazy dog."
 
 func BenchmarkAdd(b *testing.B) {