[*] chore: Update official reference

Author: Andrew-M-C

internal/official/emoji-sequences.go

@@ -1,5 +1,5 @@
 # emoji-sequences.txt
-# Date: 2025-04-23, 17:41:13 GMT
+# Date: 2025-07-25, 17:54:32 GMT
 # © 2025 Unicode®, Inc.
 # Unicode and the Unicode Logo are registered trademarks of Unicode, Inc. in the U.S. and other countries.
 # For terms of use and license, see https://www.unicode.org/terms_of_use.html
@@ -325,7 +325,6 @@
 1FAD7..1FAD9  ; Basic_Emoji                  ; pouring liquid..jar                                            # E14.0  [3] (🫗..🫙)
 1FADA..1FADB  ; Basic_Emoji                  ; ginger root..pea pod                                           # E15.0  [2] (🫚..🫛)
 1FADC         ; Basic_Emoji                  ; root vegetable                                                 # E16.0  [1] (🫜)
-1FADD         ; Basic_Emoji                  ; apple core                                                     # E17.0  [1] (🫝)
 1FADF         ; Basic_Emoji                  ; splatter                                                       # E16.0  [1] (🫟)
 1FAE0..1FAE7  ; Basic_Emoji                  ; melting face..bubbles                                          # E14.0  [8] (🫠..🫧)
 1FAE8         ; Basic_Emoji                  ; shaking face                                                   # E15.0  [1] (🫨)
@@ -542,7 +541,7 @@
 1F6F0 FE0F    ; Basic_Emoji                  ; satellite                                                      # E0.7   [1] (🛰️)
 1F6F3 FE0F    ; Basic_Emoji                  ; passenger ship                                                 # E0.7   [1] (🛳️)
 
-# Total elements: 1401
+# Total elements: 1400
 
 # ================================================
 

internal/official/emoji-sequences.txt

@@ -6,50 +6,22 @@ import (
 	"io"
 	"net/http"
 	"os"
-	"regexp"
-	"strconv"
-	"strings"
 )
 
 func downloadEmoji() {
 	// list and find latest emoji versions
-	s, err := httpGet(emojiOfficialURL)
-	var lines []string
-	check(err)
-
-	if find := strings.SplitN(s, "Parent Directory", 2); len(find) == 1 {
-		printf("error, cannot find emoji sub directories")
-		os.Exit(-1)
-	} else {
-		lines = strings.Split(find[1], "\n")
-	}
-
-	latest := ""
-	rep := regexp.MustCompile(`(\d+\.\d+)\/.+20\d\d-[01]\d-\d\d`)
-	for _, line := range lines {
-		match := rep.FindStringSubmatch(line)
-		if len(match) < 2 {
-			continue
-		}
-
-		v := match[1]
-		printf("found version: %s", v)
-
-		if vercmp(latest, v) < 0 {
-			latest = v
-		}
-	}
+	latest := emojiLatestVersion
 
 	// check old emoji-data file
-	data, err := httpGet(emojiOfficialURL + emojiOldDataFileVer + "/" + emojiOfficialDataFile)
+	data, err := httpGet(emojiDataFileURL())
 	check(err)
 
 	printf("latest version: %v", latest)
 
 	// get emoji and zwj files
-	basic, err := httpGet(emojiOfficialURL + latest + "/" + emojiOfficialSeqFile)
+	basic, err := httpGet(emojiSeqFileURL(latest))
 	check(err)
-	zwj, err := httpGet(emojiOfficialURL + latest + "/" + emojiOfficialZwjSeqFile)
+	zwj, err := httpGet(emojiZwjSeqFileURL(latest))
 	check(err)
 
 	err = os.WriteFile(emojiDataFile, []byte(data), 0644)
@@ -59,45 +31,12 @@ func downloadEmoji() {
 	err = os.WriteFile(emojiZwjSeqFile, []byte(zwj), 0644)
 	check(err)
 
-	referenceURL = emojiOfficialURL + latest + "/"
-}
-
-func vercmp(left, right string) int {
-	if left == "" {
-		if right == "" {
-			return 0
-		}
-		return -1
-	}
-	if right == "" {
-		return 1
-	}
-
-	toCode := func(s string) uint32 {
-		nums := strings.Split(s, ".")
-		if len(nums) == 1 {
-			n, _ := strconv.ParseUint(s, 10, 32)
-			return uint32(n) << 16
-		}
-
-		hi, _ := strconv.ParseUint(nums[0], 10, 32)
-		lo, _ := strconv.ParseUint(nums[1], 10, 32)
-		return (uint32(hi) << 16) + uint32(lo)
-	}
-
-	leftCode := toCode(left)
-	rightCode := toCode(right)
-
-	if leftCode < rightCode {
-		return -1
-	}
-	if leftCode > rightCode {
-		return 1
-	}
-	return 0
+	referenceURL = dataFile(latest, "")
 }
 
 func httpGet(u string) (res string, err error) {
+	printf("request: %s", u)
+
 	resp, err := http.Get(u)
 	if err != nil {
 		return "", fmt.Errorf("http.Get(%s) error: %w", u, err)

internal/tool/const/download.go

@@ -16,27 +16,42 @@ import (
 // https://www.jianshu.com/p/9682f8ce1260
 
 const (
-	emojiOfficialURL = "https://www.unicode.org/Public/emoji/"
-
 	emojiOldDataFileVer = "12.1"
 
-	emojiOfficialDataFile   = "emoji-data.txt"
-	emojiOfficialSeqFile    = "emoji-sequences.txt"
-	emojiOfficialZwjSeqFile = "emoji-zwj-sequences.txt"
-
 	emojiDataFile   = "../../../internal/official/emoji-data.txt"
 	emojiSeqFile    = "../../../internal/official/emoji-sequences.txt"
 	emojiZwjSeqFile = "../../../internal/official/emoji-zwj-sequences.txt"
 
 	emojiDataGoFile = "../../../internal/official/emoji-sequences.go"
 
+	emojiLatestVersion = "17.0.0"
+
 	// typeEmojiBasic            = " Basic_Emoji "
 	// typeEmojiKeycapSequence   = " Emoji_Keycap_Sequence "
 	// typeEmojiFlagSequence     = " RGI_Emoji_Flag_Sequence "
 	// typeEmojiTagSequence      = " RGI_Emoji_Tag_Sequence "
 	// typeEmojiModifierSequence = " RGI_Emoji_Modifier_Sequence "
 )
 
+func dataFile(version, filename string) string {
+	return fmt.Sprintf(
+		"https://www.unicode.org/Public/%s/emoji/%s",
+		version, filename,
+	)
+}
+
+func emojiDataFileURL() string {
+	return fmt.Sprintf("https://www.unicode.org/Public/emoji/%s/emoji-data.txt", emojiOldDataFileVer)
+}
+
+func emojiSeqFileURL(version string) string {
+	return dataFile(version, "emoji-sequences.txt")
+}
+
+func emojiZwjSeqFileURL(version string) string {
+	return dataFile(version, "emoji-zwj-sequences.txt")
+}
+
 var (
 	referenceURL = ""
 	longestRunes [][]rune