replace sha1 with sha2-256

.chloggen/introduce_sha2Hasher.yaml

@@ -0,0 +1,16 @@
+# One of 'breaking', 'deprecation', 'new_component', 'enhancement', 'bug_fix'
+change_type: enhancement
+
+# The name of the component, or a single word describing the area of concern, (e.g. filelogreceiver)
+component: attributesprocessor, resourceprocessor
+
+# A brief description of the change.  Surround your text with quotes ("") if it needs to start with a backtick (`).
+note: Add feature gate to support using SHA2-256 instead of SHA-1 for hashing
+
+# One or more tracking issues related to the change
+issues: [4759, 5576]
+
+# (Optional) One or more lines of additional information to render under the primary note.
+# These lines will be padded with 2 spaces and then inserted directly into the document.
+# Use pipe (|) for multiline entries.
+subtext: enable switching to use SHA-2 256 with the feature gate `coreinternal.attraction.hash.sha256`

internal/coreinternal/attraction/attraction.go

@@ -21,10 +21,17 @@ import (
 	"strings"
 
 	"go.opentelemetry.io/collector/client"
+	"go.opentelemetry.io/collector/featuregate"
 	"go.opentelemetry.io/collector/pdata/pcommon"
 	"go.uber.org/zap"
 )
 
+var enableSha256Gate = featuregate.GlobalRegistry().MustRegister(
+	"coreinternal.attraction.hash.sha256",
+	featuregate.StageAlpha,
+	featuregate.WithRegisterDescription("When enabled, switches hashing algorithm from SHA-1 to SHA-2 256"),
+)
+
 // Settings specifies the processor settings.
 type Settings struct {
 	// Actions specifies the list of attributes to act on.
@@ -85,7 +92,9 @@ type ActionKeyValue struct {
 	// DELETE  - Deletes the attribute. If the key doesn't exist,
 	//           no action is performed.
 	// HASH    - Calculates the SHA-1 hash of an existing value and overwrites the
-	//           value with it's SHA-1 hash result.
+	//           value with its SHA-1 hash result. If the feature gate
+	//           `coreinternal.attraction.hash.sha256` is enabled, it uses SHA2-256
+	//           instead.
 	// EXTRACT - Extracts values using a regular expression rule from the input
 	//           'key' to target keys specified in the 'rule'. If a target key
 	//           already exists, it will be overridden.
@@ -132,8 +141,8 @@ const (
 	// Supports pattern which is matched against attribute key.
 	DELETE Action = "delete"
 
-	// HASH calculates the SHA-1 hash of an existing value and overwrites the
-	// value with it's SHA-1 hash result.
+	// HASH calculates the SHA-256 hash of an existing value and overwrites the
+	// value with it's SHA-256 hash result.
 	// Supports pattern which is matched against attribute key.
 	HASH Action = "hash"
 
@@ -405,7 +414,11 @@ func getSourceAttributeValue(ctx context.Context, action attributeAction, attrs
 
 func hashAttribute(key string, attrs pcommon.Map) {
 	if value, exists := attrs.Get(key); exists {
-		sha1Hasher(value)
+		if enableSha256Gate.IsEnabled() {
+			sha2Hasher(value)
+		} else {
+			sha1Hasher(value)
+		}
 	}
 }
 

internal/coreinternal/attraction/attraction_test.go

@@ -16,7 +16,8 @@ package attraction
 
 import (
 	"context"
-	"crypto/sha1" // #nosec
+	"crypto/sha1"
+	"crypto/sha256"
 	"encoding/binary"
 	"errors"
 	"fmt"
@@ -630,7 +631,7 @@ func TestAttributes_HashValue(t *testing.T) {
 				"updateme": "foo",
 			},
 			expectedAttributes: map[string]interface{}{
-				"updateme": sha1Hash([]byte("foo")),
+				"updateme": hash([]byte("foo")),
 			},
 		},
 		// Ensure int data types are hashed correctly
@@ -640,7 +641,7 @@ func TestAttributes_HashValue(t *testing.T) {
 				"updateme": intVal,
 			},
 			expectedAttributes: map[string]interface{}{
-				"updateme": sha1Hash(intBytes),
+				"updateme": hash(intBytes),
 			},
 		},
 		// Ensure double data types are hashed correctly
@@ -650,7 +651,7 @@ func TestAttributes_HashValue(t *testing.T) {
 				"updateme": doubleVal,
 			},
 			expectedAttributes: map[string]interface{}{
-				"updateme": sha1Hash(doubleBytes),
+				"updateme": hash(doubleBytes),
 			},
 		},
 		// Ensure bool data types are hashed correctly
@@ -660,7 +661,7 @@ func TestAttributes_HashValue(t *testing.T) {
 				"updateme": true,
 			},
 			expectedAttributes: map[string]interface{}{
-				"updateme": sha1Hash([]byte{1}),
+				"updateme": hash([]byte{1}),
 			},
 		},
 		// Ensure bool data types are hashed correctly
@@ -670,7 +671,7 @@ func TestAttributes_HashValue(t *testing.T) {
 				"updateme": false,
 			},
 			expectedAttributes: map[string]interface{}{
-				"updateme": sha1Hash([]byte{0}),
+				"updateme": hash([]byte{0}),
 			},
 		},
 		// Ensure regex pattern is being used
@@ -681,7 +682,7 @@ func TestAttributes_HashValue(t *testing.T) {
 				"donotupdatemebyregexp": false,
 			},
 			expectedAttributes: map[string]interface{}{
-				"updatemebyregexp":      sha1Hash([]byte{0}),
+				"updatemebyregexp":      hash([]byte{0}),
 				"donotupdatemebyregexp": false,
 			},
 		},
@@ -939,13 +940,26 @@ func TestValidConfiguration(t *testing.T) {
 
 }
 
+func hash(b []byte) string {
+	if enableSha256Gate.IsEnabled() {
+		return sha2Hash(b)
+	} else {
+		return sha1Hash(b)
+	}
+}
+
 func sha1Hash(b []byte) string {
-	// #nosec
 	h := sha1.New()
 	h.Write(b)
 	return fmt.Sprintf("%x", h.Sum(nil))
 }
 
+func sha2Hash(b []byte) string {
+	h := sha256.New()
+	h.Write(b)
+	return fmt.Sprintf("%x", h.Sum(nil))
+}
+
 type mockInfoAuth map[string]interface{}
 
 func (a mockInfoAuth) GetAttribute(name string) interface{} {

internal/coreinternal/attraction/hasher.go

@@ -15,7 +15,8 @@
 package attraction // import "github.com/open-telemetry/opentelemetry-collector-contrib/internal/coreinternal/attraction"
 
 import (
-	"crypto/sha1" // #nosec
+	"crypto/sha1"
+	"crypto/sha256"
 	"encoding/binary"
 	"encoding/hex"
 	"math"
@@ -33,6 +34,7 @@ var (
 	byteFalse = [1]byte{0}
 )
 
+// Deprecated: [v0.75.0] use sha2Hasher instead.
 // sha1Hasher hashes an AttributeValue using SHA1 and returns a
 // hashed version of the attribute. In practice, this would mostly be used
 // for string attributes but we support all types for completeness/correctness
@@ -69,3 +71,39 @@ func sha1Hasher(attr pcommon.Value) {
 
 	attr.SetStr(hashed)
 }
+
+// sha2Hasher hashes an AttributeValue using SHA2-256 and returns a
+// hashed version of the attribute. In practice, this would mostly be used
+// for string attributes but we support all types for completeness/correctness
+// and eliminate any surprises.
+func sha2Hasher(attr pcommon.Value) {
+	var val []byte
+	switch attr.Type() {
+	case pcommon.ValueTypeStr:
+		val = []byte(attr.Str())
+	case pcommon.ValueTypeBool:
+		if attr.Bool() {
+			val = byteTrue[:]
+		} else {
+			val = byteFalse[:]
+		}
+	case pcommon.ValueTypeInt:
+		val = make([]byte, int64ByteSize)
+		binary.LittleEndian.PutUint64(val, uint64(attr.Int()))
+	case pcommon.ValueTypeDouble:
+		val = make([]byte, float64ByteSize)
+		binary.LittleEndian.PutUint64(val, math.Float64bits(attr.Double()))
+	}
+
+	var hashed string
+	if len(val) > 0 {
+		h := sha256.New()
+		_, _ = h.Write(val)
+		val = h.Sum(nil)
+		hashedBytes := make([]byte, hex.EncodedLen(len(val)))
+		hex.Encode(hashedBytes, val)
+		hashed = string(hashedBytes)
+	}
+
+	attr.SetStr(hashed)
+}

internal/coreinternal/go.mod

@@ -5,6 +5,7 @@ go 1.19
 require (
 	github.com/stretchr/testify v1.8.2
 	go.opentelemetry.io/collector v0.74.0
+	go.opentelemetry.io/collector/featuregate v0.74.0
 	go.opentelemetry.io/collector/pdata v1.0.0-rc8
 	go.opentelemetry.io/collector/semconv v0.74.0
 	go.uber.org/zap v1.24.0