Here's an example of salting and hashing a given password in python.
import scrypt
import os
# Length of salt
PW_SALT_BYTES = 32
# Length of scrypt hash of passwords
PW_HASH_BYTES = 64
# test password
password = "hello"
salt = os.urandom(PW_SALT_BYTES).encode('hex')
# hash(password, salt, N=1 << 14, r=8, p=1, buflen=64)
hashed_password = scrypt.hash(str(password), salt.decode('hex'), buflen=PW_HASH_BYTES).encode('hex')
print(hashed_password)
Which would give us a hashed and salted string in return:-
4d1da45b401961fccb10e094ecd70ec79510f05483ca293d300bbd0024e35866ca39fe09fbc15f83a359431021a1ed9644f7d2b871b357e37a186300877edb18
How would I implement this in golang?
What is a Salt? A salt is a random character string that is added to the beginning or the end of a password. This salt is unique to each user, and is stored in the database along with the username and salted-hashed password. An example username-password database using the SHA256 hashing function with a salt.
Bcrypt was not designed for encrypting large amounts of data. It is best implemented for passwords, however SHA-256 is better for large amounts of data because it is less costly and faster.
According to OWASP Guidelines, a salt is a value generated by a cryptographically secure function that is added to the input of hash functions to create unique hashes for every input, regardless of the input not being unique.
The Salted SHA512 Password Storage Scheme provides a mechanism for encoding user passwords using a salted form of the 512-bit SHA-2 message digest algorithm.
Rather than using scrypt
, a great library for securely hashing passwords with random salts in Golang is golang.org/x/crypto/bcrypt, as mentioned in the following answer:
Bcrypt password hashing in Golang (compatible with Node.js)?
A couple benefits of using bcrypt
instead of scrypt
:
Here's an example of using bcrypt taken from the above answer:
package main
import (
"golang.org/x/crypto/bcrypt"
"fmt"
)
func main() {
password := []byte("MyDarkSecret")
// Hashing the password with the default cost of 10
hashedPassword, err := bcrypt.GenerateFromPassword(password, bcrypt.DefaultCost)
if err != nil {
panic(err)
}
fmt.Println(string(hashedPassword))
// Comparing the password with the hash
err = bcrypt.CompareHashAndPassword(hashedPassword, password)
fmt.Println(err) // nil means it is a match
}
Go doesn't have scrypt in the standard library but there is an "official" implementation in the go.crypto repo.
import (
"crypto/rand"
"fmt"
"io"
"log"
"code.google.com/p/go.crypto/scrypt"
)
const (
PW_SALT_BYTES = 32
PW_HASH_BYTES = 64
password = "hello"
)
func main() {
salt := make([]byte, PW_SALT_BYTES)
_, err := io.ReadFull(rand.Reader, salt)
if err != nil {
log.Fatal(err)
}
hash, err := scrypt.Key([]byte(password), salt, 1<<14, 8, 1, PW_HASH_BYTES)
if err != nil {
log.Fatal(err)
}
fmt.Printf("%x\n", hash)
}
It looks like now Go has scrypt in official library. Its subrepository x/crypto among many other crypto functions has an scrypt.
Here is an example of how you can use it:
package main
import (
"golang.org/x/crypto/scrypt"
"fmt"
)
func main(){
salt := []byte("asdfasdf")
dk, err := scrypt.Key([]byte("some password"), salt, 16384, 8, 1, 32)
fmt.Println(dk)
fmt.Println(err)
}
Here's a complete hashing utilities funcs I wrote based on RFC 2898 / PKCS #5 v2.0
.
Hash
can be used to hash passwords, straight forward something like Hash("hello")
while Verify
can be used to raw password against a hash, basically what it does is to hashes the raw string and compares it with the actual hash.
package common
import (
"crypto/rand"
"crypto/sha1"
"encoding/base64"
"errors"
"fmt"
"golang.org/x/crypto/pbkdf2"
"io"
"strconv"
"strings"
)
const (
SALT_BYTE_SIZE = 24
HASH_BYTE_SIZE = 24
PBKDF2_ITERATIONS = 1000
)
func Hash(password string) (string, error) {
salt := make([]byte, SALT_BYTE_SIZE)
if _, err := io.ReadFull(rand.Reader, salt); err != nil {
fmt.Print("Err generating random salt")
return "", errors.New("Err generating random salt")
}
//todo: enhance: randomize itrs as well
hbts := pbkdf2.Key([]byte(password), salt, PBKDF2_ITERATIONS, HASH_BYTE_SIZE, sha1.New)
//hbtstr := fmt.Sprintf("%x", hbts)
return fmt.Sprintf("%v:%v:%v",
PBKDF2_ITERATIONS,
base64.StdEncoding.EncodeToString(salt),
base64.StdEncoding.EncodeToString(hbts)), nil
}
func Verify(raw, hash string) (bool, error) {
hparts := strings.Split(hash, ":")
itr, err := strconv.Atoi(hparts[0])
if err != nil {
fmt.Printf("wrong hash %v", hash)
return false, errors.New("wrong hash, iteration is invalid")
}
salt, err := base64.StdEncoding.DecodeString(hparts[1])
if err != nil {
fmt.Print("wrong hash, salt error:", err)
return false, errors.New("wrong hash, salt error:" + err.Error())
}
hsh, err := base64.StdEncoding.DecodeString(hparts[2])
if err != nil {
fmt.Print("wrong hash, hash error:", err)
return false, errors.New("wrong hash, hash error:" + err.Error())
}
rhash := pbkdf2.Key([]byte(raw), salt, itr, len(hsh), sha1.New)
return equal(rhash, hsh), nil
}
//bytes comparisons
func equal(h1, h2 []byte) bool {
diff := uint32(len(h1)) ^ uint32(len(h2))
for i := 0; i < len(h1) && i < len(h2); i++ {
diff |= uint32(h1[i] ^ h2[i])
}
return diff == 0
}
Here's unit test that would help you figuring out how to call such funcs
package common
import (
"github.com/stretchr/testify/assert"
"testing"
)
func TestHash(t *testing.T) {
hash, err := Hash("hello")
assert.Nil(t, err)
assert.NotEmpty(t, hash)
}
func TestVerify(t *testing.T) {
hash, err := Hash("hello")
assert.Nil(t, err)
assert.NotEmpty(t, hash)
ok, err := Verify("hello", hash)
assert.Nil(t, err)
assert.True(t, ok)
}
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With