improve performance and reduce allocations of most UUID methods

This commit improves the performance and reduces the number of allocs of most UUID methods and adds a new UUID.Parse() method for parsing string encoded UUIDs. Parsing string encoded UUIDs is now 2x faster and no longer allocates. The NullUUID MarshalJSON() and UnmarshalJSON() methods have also been improved and no longer call out to json.Unmarshal. The UUID.Format method has been improved for common cases. Benchmark results: ``` goos: linux goarch: amd64 pkg: github.com/gofrs/uuid cpu: Intel(R) Core(TM) i9-9900K CPU @ 3.60GHz name old time/op new time/op delta UnmarshalText/canonical-16 47.7ns ± 0% 22.3ns ± 0% -53.26% (p=0.000 n=8+10) UnmarshalText/urn-16 47.8ns ± 1% 22.3ns ± 0% -53.32% (p=0.000 n=10+10) UnmarshalText/braced-16 47.8ns ± 1% 22.3ns ± 0% -53.34% (p=0.000 n=9+9) ParseV4-16 86.5ns ±10% 22.6ns ± 0% -73.85% (p=0.000 n=9+8) NullMarshalJSON/Valid-16 308ns ±21% 72ns ±12% -76.71% (p=0.000 n=10+10) NullMarshalJSON/Invalid-16 41.8ns ± 2% 1.4ns ± 0% -96.59% (p=0.000 n=10+8) Format/s-16 151ns ± 3% 143ns ± 7% -5.18% (p=0.003 n=10+10) Format/S-16 305ns ± 2% 161ns ± 2% -47.38% (p=0.000 n=10+9) Format/q-16 217ns ±12% 144ns ± 6% -33.52% (p=0.000 n=10+10) Format/x-16 170ns ±13% 123ns ± 1% -27.95% (p=0.000 n=10+10) Format/X-16 324ns ±11% 148ns ± 1% -54.15% (p=0.000 n=10+10) Format/v-16 156ns ± 4% 142ns ± 5% -8.68% (p=0.000 n=10+10) Format/+v-16 155ns ± 3% 142ns ± 6% -8.67% (p=0.000 n=10+10) Format/#v-16 894ns ± 1% 847ns ± 1% -5.22% (p=0.000 n=10+9) String-16 70.1ns ±36% 70.4ns ±29% ~ (p=0.971 n=10+10) FromBytes-16 1.81ns ± 0% 1.81ns ± 0% -0.15% (p=0.010 n=8+9) FromString/canonical-16 94.3ns ±20% 23.3ns ± 1% -75.25% (p=0.000 n=10+10) FromString/urn-16 93.7ns ±11% 23.8ns ± 1% -74.66% (p=0.000 n=10+8) FromString/braced-16 87.1ns ± 5% 23.5ns ± 1% -73.03% (p=0.000 n=9+10) MarshalBinary-16 0.20ns ± 3% 0.20ns ± 1% ~ (p=0.922 n=10+9) MarshalText-16 115ns ±25% 22ns ± 1% -80.66% (p=0.000 n=10+10) name old alloc/op new alloc/op delta UnmarshalText/canonical-16 0.00B 0.00B ~ (all equal) UnmarshalText/urn-16 0.00B 0.00B ~ (all equal) UnmarshalText/braced-16 0.00B 0.00B ~ (all equal) ParseV4-16 48.0B ± 0% 0.0B -100.00% (p=0.000 n=10+10) NullMarshalJSON/Valid-16 160B ± 0% 48B ± 0% -70.00% (p=0.000 n=10+10) NullMarshalJSON/Invalid-16 8.00B ± 0% 0.00B -100.00% (p=0.000 n=10+10) Format/s-16 48.0B ± 0% 48.0B ± 0% ~ (all equal) Format/S-16 96.0B ± 0% 48.0B ± 0% -50.00% (p=0.000 n=10+10) Format/q-16 96.0B ± 0% 48.0B ± 0% -50.00% (p=0.000 n=10+10) Format/x-16 64.0B ± 0% 32.0B ± 0% -50.00% (p=0.000 n=10+10) Format/X-16 112B ± 0% 32B ± 0% -71.43% (p=0.000 n=10+10) Format/v-16 48.0B ± 0% 48.0B ± 0% ~ (all equal) Format/+v-16 48.0B ± 0% 48.0B ± 0% ~ (all equal) Format/#v-16 128B ± 0% 16B ± 0% -87.50% (p=0.000 n=10+10) String-16 48.0B ± 0% 48.0B ± 0% ~ (all equal) FromBytes-16 0.00B 0.00B ~ (all equal) FromString/canonical-16 48.0B ± 0% 0.0B -100.00% (p=0.000 n=10+10) FromString/urn-16 48.0B ± 0% 0.0B -100.00% (p=0.000 n=10+10) FromString/braced-16 48.0B ± 0% 0.0B -100.00% (p=0.000 n=10+10) MarshalBinary-16 0.00B 0.00B ~ (all equal) MarshalText-16 96.0B ± 0% 0.0B -100.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta UnmarshalText/canonical-16 0.00 0.00 ~ (all equal) UnmarshalText/urn-16 0.00 0.00 ~ (all equal) UnmarshalText/braced-16 0.00 0.00 ~ (all equal) ParseV4-16 1.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) NullMarshalJSON/Valid-16 4.00 ± 0% 1.00 ± 0% -75.00% (p=0.000 n=10+10) NullMarshalJSON/Invalid-16 1.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) Format/s-16 1.00 ± 0% 1.00 ± 0% ~ (all equal) Format/S-16 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) Format/q-16 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) Format/x-16 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) Format/X-16 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) Format/v-16 1.00 ± 0% 1.00 ± 0% ~ (all equal) Format/+v-16 1.00 ± 0% 1.00 ± 0% ~ (all equal) Format/#v-16 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) String-16 1.00 ± 0% 1.00 ± 0% ~ (all equal) FromBytes-16 0.00 0.00 ~ (all equal) FromString/canonical-16 1.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) FromString/urn-16 1.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) FromString/braced-16 1.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) MarshalBinary-16 0.00 0.00 ~ (all equal) MarshalText-16 2.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) ```
gofrs · Dec 5, 2021 · 5f181ae · 5f181ae
1 parent 0c84a43
commit 5f181ae
Show file tree

Hide file tree

Showing 6 changed files with 365 additions and 160 deletions.
diff --git a/codec.go b/codec.go
@@ -22,8 +22,7 @@
 package uuid
 
 import (
-	"bytes"
-	"encoding/hex"
+	"errors"
 	"fmt"
 )
 
@@ -45,11 +44,77 @@ func FromBytesOrNil(input []byte) UUID {
 	return uuid
 }
 
+var errInvalidFormat = errors.New("uuid: invalid UUID format")
+
+func fromHexChar(c byte) byte {
+	switch {
+	case '0' <= c && c <= '9':
+		return c - '0'
+	case 'a' <= c && c <= 'f':
+		return c - 'a' + 10
+	case 'A' <= c && c <= 'F':
+		return c - 'A' + 10
+	}
+	return 255
+}
+
+// Parse parses the UUID stored in the string text. Parsing and supported
+// formats are the same as UnmarshalText.
+func (u *UUID) Parse(s string) error {
+	switch len(s) {
+	case 32: // hash
+	case 36: // canonical
+	case 34, 38:
+		if s[0] != '{' || s[len(s)-1] != '}' {
+			return fmt.Errorf("uuid: incorrect UUID format in string %q", s)
+		}
+		s = s[1 : len(s)-1]
+	case 41, 45:
+		if s[:9] != "urn:uuid:" {
+			return fmt.Errorf("uuid: incorrect UUID format in string %q", s[:9])
+		}
+		s = s[9:]
+	default:
+		return fmt.Errorf("uuid: incorrect UUID length %d in string %q", len(s), s)
+	}
+	// canonical
+	if len(s) == 36 {
+		if s[8] != '-' || s[13] != '-' || s[18] != '-' || s[23] != '-' {
+			return fmt.Errorf("uuid: incorrect UUID format in string %q", s)
+		}
+		for i, x := range [16]byte{
+			0, 2, 4, 6,
+			9, 11,
+			14, 16,
+			19, 21,
+			24, 26, 28, 30, 32, 34,
+		} {
+			v1 := fromHexChar(s[x])
+			v2 := fromHexChar(s[x+1])
+			if v1|v2 == 255 {
+				return errInvalidFormat
+			}
+			u[i] = (v1 << 4) | v2
+		}
+		return nil
+	}
+	// hash like
+	for i := 0; i < 32; i += 2 {
+		v1 := fromHexChar(s[i])
+		v2 := fromHexChar(s[i+1])
+		if v1|v2 == 255 {
+			return errInvalidFormat
+		}
+		u[i/2] = (v1 << 4) | v2
+	}
+	return nil
+}
+
 // FromString returns a UUID parsed from the input string.
 // Input is expected in a form accepted by UnmarshalText.
-func FromString(input string) (UUID, error) {
-	u := UUID{}
-	err := u.UnmarshalText([]byte(input))
+func FromString(text string) (UUID, error) {
+	var u UUID
+	err := u.Parse(text)
 	return u, err
 }
 
@@ -66,7 +131,9 @@ func FromStringOrNil(input string) UUID {
 // MarshalText implements the encoding.TextMarshaler interface.
 // The encoding is the same as returned by the String() method.
 func (u UUID) MarshalText() ([]byte, error) {
-	return []byte(u.String()), nil
+	var buf [36]byte
+	encodeHex(buf[:], u)
+	return buf[:], nil
 }
 
 // UnmarshalText implements the encoding.TextUnmarshaler interface.
@@ -103,96 +170,52 @@ func (u UUID) MarshalText() ([]byte, error) {
 //   braced := '{' plain '}' | '{' hashlike  '}'
 //   urn := URN ':' UUID-NID ':' plain
 //
-func (u *UUID) UnmarshalText(text []byte) error {
-	switch len(text) {
-	case 32:
-		return u.decodeHashLike(text)
+func (u *UUID) UnmarshalText(b []byte) error {
+	switch len(b) {
+	case 32: // hash
+	case 36: // canonical
 	case 34, 38:
-		return u.decodeBraced(text)
-	case 36:
-		return u.decodeCanonical(text)
+		if b[0] != '{' || b[len(b)-1] != '}' {
+			return fmt.Errorf("uuid: incorrect UUID format in string %q", b)
+		}
+		b = b[1 : len(b)-1]
 	case 41, 45:
-		return u.decodeURN(text)
+		if string(b[:9]) != "urn:uuid:" {
+			return fmt.Errorf("uuid: incorrect UUID format in string %q", b[:9])
+		}
+		b = b[9:]
 	default:
-		return fmt.Errorf("uuid: incorrect UUID length %d in string %q", len(text), text)
-	}
-}
-
-// decodeCanonical decodes UUID strings that are formatted as defined in RFC-4122 (section 3):
-// "6ba7b810-9dad-11d1-80b4-00c04fd430c8".
-func (u *UUID) decodeCanonical(t []byte) error {
-	if t[8] != '-' || t[13] != '-' || t[18] != '-' || t[23] != '-' {
-		return fmt.Errorf("uuid: incorrect UUID format in string %q", t)
+		return fmt.Errorf("uuid: incorrect UUID length %d in string %q", len(b), b)
 	}
-
-	src := t
-	dst := u[:]
-
-	for i, byteGroup := range byteGroups {
-		if i > 0 {
-			src = src[1:] // skip dash
+	if len(b) == 36 {
+		if b[8] != '-' || b[13] != '-' || b[18] != '-' || b[23] != '-' {
+			return fmt.Errorf("uuid: incorrect UUID format in string %q", b)
 		}
-		_, err := hex.Decode(dst[:byteGroup/2], src[:byteGroup])
-		if err != nil {
-			return err
+		for i, x := range [16]byte{
+			0, 2, 4, 6,
+			9, 11,
+			14, 16,
+			19, 21,
+			24, 26, 28, 30, 32, 34,
+		} {
+			v1 := fromHexChar(b[x])
+			v2 := fromHexChar(b[x+1])
+			if v1|v2 == 255 {
+				return errInvalidFormat
+			}
+			u[i] = (v1 << 4) | v2
 		}
-		src = src[byteGroup:]
-		dst = dst[byteGroup/2:]
-	}
-
-	return nil
-}
-
-// decodeHashLike decodes UUID strings that are using the following format:
-//  "6ba7b8109dad11d180b400c04fd430c8".
-func (u *UUID) decodeHashLike(t []byte) error {
-	src := t[:]
-	dst := u[:]
-
-	_, err := hex.Decode(dst, src)
-	return err
-}
-
-// decodeBraced decodes UUID strings that are using the following formats:
-//  "{6ba7b810-9dad-11d1-80b4-00c04fd430c8}"
-//  "{6ba7b8109dad11d180b400c04fd430c8}".
-func (u *UUID) decodeBraced(t []byte) error {
-	l := len(t)
-
-	if t[0] != '{' || t[l-1] != '}' {
-		return fmt.Errorf("uuid: incorrect UUID format in string %q", t)
+		return nil
 	}
-
-	return u.decodePlain(t[1 : l-1])
-}
-
-// decodeURN decodes UUID strings that are using the following formats:
-//  "urn:uuid:6ba7b810-9dad-11d1-80b4-00c04fd430c8"
-//  "urn:uuid:6ba7b8109dad11d180b400c04fd430c8".
-func (u *UUID) decodeURN(t []byte) error {
-	total := len(t)
-
-	urnUUIDPrefix := t[:9]
-
-	if !bytes.Equal(urnUUIDPrefix, urnPrefix) {
-		return fmt.Errorf("uuid: incorrect UUID format in string %q", t)
-	}
-
-	return u.decodePlain(t[9:total])
-}
-
-// decodePlain decodes UUID strings that are using the following formats:
-//  "6ba7b810-9dad-11d1-80b4-00c04fd430c8" or in hash-like format
-//  "6ba7b8109dad11d180b400c04fd430c8".
-func (u *UUID) decodePlain(t []byte) error {
-	switch len(t) {
-	case 32:
-		return u.decodeHashLike(t)
-	case 36:
-		return u.decodeCanonical(t)
-	default:
-		return fmt.Errorf("uuid: incorrect UUID length %d in string %q", len(t), t)
+	for i := 0; i < 32; i += 2 {
+		v1 := fromHexChar(b[i])
+		v2 := fromHexChar(b[i+1])
+		if v1|v2 == 255 {
+			return errInvalidFormat
+		}
+		u[i/2] = (v1 << 4) | v2
 	}
+	return nil
 }
 
 // MarshalBinary implements the encoding.BinaryMarshaler interface.