Fix emoji detection in certain cases (go-gitea#12320)

* Fix emoji detection certain cases

Previous tests weren't complicated enough so there were some situations where emojis were't detected properly. Find the earliest occurance in addition to checking for the longest combination.

Fixes go-gitea#12312

* ok spell bot

Co-authored-by: Lauris BH <[email protected]>

Author: 2 people

modules/emoji/emoji.go

@@ -130,18 +130,35 @@ func ReplaceAliases(s string) string {
 // FindEmojiSubmatchIndex returns index pair of longest emoji in a string
 func FindEmojiSubmatchIndex(s string) []int {
 	loadMap()
+	found := make(map[int]int)
+	keys := make([]int, 0)
 
 	//see if there are any emoji in string before looking for position of specific ones
 	//no performance difference when there is a match but 10x faster when there are not
 	if s == ReplaceCodes(s) {
 		return nil
 	}
 
+	// get index of first emoji occurrence while also checking for longest combination
 	for j := range GemojiData {
 		i := strings.Index(s, GemojiData[j].Emoji)
 		if i != -1 {
-			return []int{i, i + len(GemojiData[j].Emoji)}
+			if _, ok := found[i]; !ok {
+				if len(keys) == 0 || i < keys[0] {
+					found[i] = j
+					keys = []int{i}
+				}
+				if i == 0 {
+					break
+				}
+			}
 		}
 	}
+
+	if len(keys) > 0 {
+		index := keys[0]
+		return []int{index, index + len(GemojiData[found[index]].Emoji)}
+	}
+
 	return nil
 }

modules/markup/html_test.go

@@ -266,6 +266,10 @@ func TestRender_emoji(t *testing.T) {
 	test(
 		"Some text with 😄😄 2 emoji next to each other",
 		`<p>Some text with <span class="emoji" aria-label="grinning face with smiling eyes">😄</span><span class="emoji" aria-label="grinning face with smiling eyes">😄</span> 2 emoji next to each other</p>`)
+	test(
+		"😎🤪🔐🤑❓",
+		`<p><span class="emoji" aria-label="smiling face with sunglasses">😎</span><span class="emoji" aria-label="zany face">🤪</span><span class="emoji" aria-label="locked with key">🔐</span><span class="emoji" aria-label="money-mouth face">🤑</span><span class="emoji" aria-label="question mark">❓</span></p>`)
+
 	// should match nothing
 	test(
 		"2001:0db8:85a3:0000:0000:8a2e:0370:7334",