feat: add UTF-8 support and base256emoji

This include fixes for UTF-8 as well as base256emoji encoding (an encoding which actually use UTF-8).

base256emoji.go

@@ -0,0 +1,94 @@
+package multibase
+
+import (
+	"strconv"
+	"strings"
+	"unicode/utf8"
+)
+
+var base256emojiTable = [256]rune{
+	// Curated list, this is just a list of things that *somwhat* are related to our comunity
+	'🚀', '🪐', '☄', '🛰', '🌌', // Space
+	'🌑', '🌒', '🌓', '🌔', '🌕', '🌖', '🌗', '🌘', // Moon
+	'🌍', '🌏', '🌎', // Our Home, for now (earth)
+	'☉', '☀', // Our Garden, for now (sol)
+	'💻', '🖥', '💾', '💿', // Computer
+	// The rest is completed from https://home.unicode.org/emoji/emoji-frequency/ at the time of creation (december 2021) (the data is from 2019), most used first until we reach 256.
+	// We exclude modifier based emojies (such as flags) as they are bigger than one single codepoint.
+	// Some other emojies were removed adhoc for various reasons.
+	'😂', '❤', '😍', '🤣', '😊', '🙏', '💕', '😭', '😘', '👍',
+	'😅', '👏', '😁', '🔥', '🥰', '💔', '💖', '💙', '😢', '🤔',
+	'😆', '🙄', '💪', '😉', '☺', '👌', '🤗', '💜', '😔', '😎',
+	'😇', '🌹', '🤦', '🎉', '💞', '✌', '✨', '🤷', '😱', '😌',
+	'🌸', '🙌', '😋', '💗', '💚', '😏', '💛', '🙂', '💓', '🤩',
+	'😄', '😀', '🖤', '😃', '💯', '🙈', '👇', '🎶', '😒', '🤭',
+	'❣', '😜', '💋', '👀', '😪', '😑', '💥', '🙋', '😞', '😩',
+	'😡', '🤪', '👊', '🥳', '😥', '🤤', '👉', '💃', '😳', '✋',
+	'😚', '😝', '😴', '🌟', '😬', '🙃', '🍀', '🌷', '😻', '😓',
+	'⭐', '✅', '🥺', '🌈', '😈', '🤘', '💦', '✔', '😣', '🏃',
+	'💐', '☹', '🎊', '💘', '😠', '☝', '😕', '🌺', '🎂', '🌻',
+	'😐', '🖕', '💝', '🙊', '😹', '🗣', '💫', '💀', '👑', '🎵',
+	'🤞', '😛', '🔴', '😤', '🌼', '😫', '⚽', '🤙', '☕', '🏆',
+	'🤫', '👈', '😮', '🙆', '🍻', '🍃', '🐶', '💁', '😲', '🌿',
+	'🧡', '🎁', '⚡', '🌞', '🎈', '❌', '✊', '👋', '😰', '🤨',
+	'😶', '🤝', '🚶', '💰', '🍓', '💢', '🤟', '🙁', '🚨', '💨',
+	'🤬', '✈', '🎀', '🍺', '🤓', '😙', '💟', '🌱', '😖', '👶',
+	'🥴', '▶', '➡', '❓', '💎', '💸', '⬇', '😨', '🌚', '🦋',
+	'😷', '🕺', '⚠', '🙅', '😟', '😵', '👎', '🤲', '🤠', '🤧',
+	'📌', '🔵', '💅', '🧐', '🐾', '🍒', '😗', '🤑', '🌊', '🤯',
+	'🐷', '☎', '💧', '😯', '💆', '👆', '🎤', '🙇', '🍑', '❄',
+	'🌴', '💣', '🐸', '💌', '📍', '🥀', '🤢', '👅', '💡', '💩',
+	'👐', '📸', '👻', '🤐', '🤮', '🎼', '🥵', '🚩', '🍎', '🍊',
+	'👼', '💍', '📣', '🥂',
+}
+
+var base256emojiReverseTable map[rune]byte
+
+func init() {
+	base256emojiReverseTable = make(map[rune]byte, len(base256emojiTable))
+	for i, v := range base256emojiTable {
+		base256emojiReverseTable[v] = byte(i)
+	}
+}
+
+func base256emojiEncode(in []byte) string {
+	var l int
+	for _, v := range in {
+		l += utf8.RuneLen(base256emojiTable[v])
+	}
+	var out strings.Builder
+	out.Grow(l)
+	for _, v := range in {
+		out.WriteRune(base256emojiTable[v])
+	}
+	return out.String()
+}
+
+type base256emojiCorruptInputError struct {
+	index int
+	char  rune
+}
+
+func (e base256emojiCorruptInputError) Error() string {
+	return "illegal base256emoji data at input byte " + strconv.FormatInt(int64(e.index), 10) + ", char: '" + string(e.char) + "'"
+}
+
+func (e base256emojiCorruptInputError) String() string {
+	return e.Error()
+}
+
+func base256emojiDecode(in string) ([]byte, error) {
+	out := make([]byte, utf8.RuneCountInString(in))
+	var stri int
+	for i := 0; len(in) > 0; i++ {
+		r, n := utf8.DecodeRuneInString(in)
+		in = in[n:]
+		var ok bool
+		out[i], ok = base256emojiReverseTable[r]
+		if !ok {
+			return nil, base256emojiCorruptInputError{stri, r}
+		}
+		stri += n
+	}
+	return out, nil
+}

base256emoji_test.go

@@ -0,0 +1,26 @@
+package multibase
+
+import "testing"
+
+func TestBase256EmojiAlphabet(t *testing.T) {
+	var c uint
+	for _, v := range base256emojiTable {
+		if v != rune(0) {
+			c++
+		}
+	}
+	if c != 256 {
+		t.Errorf("Base256Emoji count is wrong, expected 256, got %d.", c)
+	}
+}
+
+func TestBase256EmojiUniq(t *testing.T) {
+	m := make(map[rune]struct{}, len(base256emojiTable))
+	for i, v := range base256emojiTable {
+		_, ok := m[v]
+		if ok {
+			t.Errorf("Base256Emoji duplicate %s at index %d.", string(v), i)
+		}
+		m[v] = struct{}{}
+	}
+}

encoder.go

@@ -2,6 +2,7 @@ package multibase
 
 import (
 	"fmt"
+	"unicode/utf8"
 )
 
 // Encoder is a multibase encoding that is verified to be supported and
@@ -36,8 +37,9 @@ func EncoderByName(str string) (Encoder, error) {
 	var ok bool
 	if len(str) == 0 {
 		return Encoder{-1}, fmt.Errorf("empty multibase encoding")
-	} else if len(str) == 1 {
-		base = Encoding(str[0])
+	} else if utf8.RuneCountInString(str) == 1 {
+		r, _ := utf8.DecodeRuneInString(str)
+		base = Encoding(r)
 		_, ok = EncodingToStr[base]
 	} else {
 		base, ok = Encodings[str]

encoder_test.go

@@ -2,6 +2,7 @@ package multibase
 
 import (
 	"testing"
+	"unicode/utf8"
 )
 
 func TestInvalidCode(t *testing.T) {
@@ -43,9 +44,10 @@ func TestEncoder(t *testing.T) {
 		}
 		// Test that an encoder can be created from the single letter
 		// prefix
-		_, err = EncoderByName(str[0:1])
+		r, _ := utf8.DecodeRuneInString(str)
+		_, err = EncoderByName(string(r))
 		if err != nil {
-			t.Fatalf("EncoderByName(%s) failed: %v", str[0:1], err)
+			t.Fatalf("EncoderByName(%s) failed: %v", string(r), err)
 		}
 	}
 }

multibase.go

@@ -4,6 +4,7 @@ import (
 	"encoding/base64"
 	"encoding/hex"
 	"fmt"
+	"unicode/utf8"
 
 	b58 "github.com/mr-tron/base58/base58"
 	b32 "github.com/multiformats/go-base32"
@@ -38,31 +39,33 @@ const (
 	Base64url         = 'u'
 	Base64pad         = 'M'
 	Base64urlPad      = 'U'
+	Base256Emoji      = '🚀'
 )
 
 // EncodingToStr is a map of the supported encoding, unsupported encoding
 // specified in standard are left out
 var EncodingToStr = map[Encoding]string{
-	0x00: "identity",
-	'0':  "base2",
-	'f':  "base16",
-	'F':  "base16upper",
-	'b':  "base32",
-	'B':  "base32upper",
-	'c':  "base32pad",
-	'C':  "base32padupper",
-	'v':  "base32hex",
-	'V':  "base32hexupper",
-	't':  "base32hexpad",
-	'T':  "base32hexpadupper",
-	'k':  "base36",
-	'K':  "base36upper",
-	'z':  "base58btc",
-	'Z':  "base58flickr",
-	'm':  "base64",
-	'u':  "base64url",
-	'M':  "base64pad",
-	'U':  "base64urlpad",
+	0x00:         "identity",
+	'0':          "base2",
+	'f':          "base16",
+	'F':          "base16upper",
+	'b':          "base32",
+	'B':          "base32upper",
+	'c':          "base32pad",
+	'C':          "base32padupper",
+	'v':          "base32hex",
+	'V':          "base32hexupper",
+	't':          "base32hexpad",
+	'T':          "base32hexpadupper",
+	'k':          "base36",
+	'K':          "base36upper",
+	'z':          "base58btc",
+	'Z':          "base58flickr",
+	'm':          "base64",
+	'u':          "base64url",
+	'M':          "base64pad",
+	'U':          "base64urlpad",
+	Base256Emoji: "base256emoji",
 }
 
 var Encodings = map[string]Encoding{}
@@ -123,6 +126,8 @@ func Encode(base Encoding, data []byte) (string, error) {
 		return string(Base64url) + base64.RawURLEncoding.EncodeToString(data), nil
 	case Base64:
 		return string(Base64) + base64.RawStdEncoding.EncodeToString(data), nil
+	case Base256Emoji:
+		return string(Base256Emoji) + base256emojiEncode(data), nil
 	default:
 		return "", ErrUnsupportedEncoding
 	}
@@ -135,7 +140,8 @@ func Decode(data string) (Encoding, []byte, error) {
 		return 0, nil, fmt.Errorf("cannot decode multibase for zero length string")
 	}
 
-	enc := Encoding(data[0])
+	r, _ := utf8.DecodeRuneInString(data)
+	enc := Encoding(r)
 
 	switch enc {
 	case Identity:
@@ -179,6 +185,9 @@ func Decode(data string) (Encoding, []byte, error) {
 	case Base64url:
 		bytes, err := base64.RawURLEncoding.DecodeString(data[1:])
 		return Base64url, bytes, err
+	case Base256Emoji:
+		bytes, err := base256emojiDecode(data[4:])
+		return Base256Emoji, bytes, err
 	default:
 		return -1, nil, ErrUnsupportedEncoding
 	}

multibase_test.go

@@ -44,6 +44,7 @@ var encodedSamples = map[Encoding]string{
 	Base64url:         "uRGVjZW50cmFsaXplIGV2ZXJ5dGhpbmchISE",
 	Base64pad:         "MRGVjZW50cmFsaXplIGV2ZXJ5dGhpbmchISE=",
 	Base64urlPad:      "URGVjZW50cmFsaXplIGV2ZXJ5dGhpbmchISE=",
+	Base256Emoji:      "🚀💛✋💃✋😻😈🥺🤤🍀🌟💐✋😅✋💦✋🥺🏃😈😴🌟😻😝👏👏👏",
 }
 
 func testEncode(t *testing.T, encoding Encoding, bytes []byte, expected string) {