Merge pull request #84 from bits-and-blooms/dlemire/issue68

Improve support of big endian systems

README.md

@@ -108,7 +108,7 @@ make qa
 
 A Bloom filter has two parameters: _m_, the number of bits used in storage, and _k_, the number of hashing functions on elements of the set. (The actual hashing functions are important, too, but this is not a parameter for this implementation). A Bloom filter is backed by a [BitSet](https://github.com/bits-and-blooms/bitset); a key is represented in the filter by setting the bits at each value of the  hashing functions (modulo _m_). Set membership is done by _testing_ whether the bits at each value of the hashing functions (again, modulo _m_) are set. If so, the item is in the set. If the item is actually in the set, a Bloom filter will never fail (the true positive rate is 1.0); but it is susceptible to false positives. The art is to choose _k_ and _m_ correctly.
 
-In this implementation, the hashing functions used is [murmurhash](https://github.com/spaolacci/murmur3), a non-cryptographic hashing function.
+In this implementation, the hashing functions used is [murmurhash](github.com/twmb/murmur3), a non-cryptographic hashing function.
 
 
 Given the particular hashing scheme, it's best to be empirical about this. Note

go.mod

@@ -4,5 +4,5 @@ go 1.14
 
 require (
 	github.com/bits-and-blooms/bitset v1.2.0
-	github.com/spaolacci/murmur3 v1.1.0
+	github.com/twmb/murmur3 v1.1.6
 )

go.sum

@@ -1,4 +1,4 @@
 github.com/bits-and-blooms/bitset v1.2.0 h1:Kn4yilvwNtMACtf1eYDlG8H77R07mZSPbMjLyS07ChA=
 github.com/bits-and-blooms/bitset v1.2.0/go.mod h1:gIdJ4wp64HaoK2YrL1Q5/N7Y16edYb8uY+O0FJTyyDA=
-github.com/spaolacci/murmur3 v1.1.0 h1:7c1g84S4BPRrfL5Xrdp6fOJ206sU9y293DDHaoy0bLI=
-github.com/spaolacci/murmur3 v1.1.0/go.mod h1:JwIasOWyU6f++ZhiEuf87xNszmSA2myDM2Kzu9HwQUA=
+github.com/twmb/murmur3 v1.1.6 h1:mqrRot1BRxm+Yct+vavLMou2/iJt0tNVTTC0QoIjaZg=
+github.com/twmb/murmur3 v1.1.6/go.mod h1:Qq/R7NUyOfr65zD+6Q5IHKsJLwP7exErjN6lyyq3OSQ=

murmur.go

@@ -267,9 +267,6 @@ func (d *digest128) sum256(data []byte) (hash1, hash2, hash3, hash4 uint64) {
 	// we do not want to append to an actual array!!!
 	if tail_length+1 == block_size {
 		// We are left with no tail!!!
-		// Note that murmur3 is sensitive to endianess and so are we.
-		// We assume a little endian system. Go effectively never run
-		// on big endian systems so this is fine.
 		word1 := *(*uint64)(unsafe.Pointer(&tail[0]))
 		word2 := uint64(*(*uint32)(unsafe.Pointer(&tail[8])))
 		word2 = word2 | (uint64(tail[12]) << 32) | (uint64(tail[13]) << 40) | (uint64(tail[14]) << 48)

murmur_test.go

@@ -4,7 +4,7 @@ import (
 	"math/rand"
 	"testing"
 
-	"github.com/spaolacci/murmur3"
+	"github.com/twmb/murmur3"
 )
 
 // We want to preserve backward compatibility