Исправление парсинга emoji и составных emoji (skin tones)

This commit is contained in:
RoyceDa
2026-04-08 22:58:49 +02:00
parent adfc6add6f
commit 130ad9c35a

View File

@@ -42,6 +42,20 @@ export function TextParser(props: TextParserProps) {
const theme = useMantineTheme();
let entityCount = 0;
const UNICODE_EMOJI_SEQUENCE_REGEX =
/(?:\p{Regional_Indicator}{2}|[0-9#*]\uFE0F?\u20E3|\p{Extended_Pictographic}(?:\uFE0F|\uFE0E)?(?:\p{Emoji_Modifier})?(?:\u200D\p{Extended_Pictographic}(?:\uFE0F|\uFE0E)?(?:\p{Emoji_Modifier})?)*)/u;
const UNICODE_EMOJI_SEQUENCE_REGEX_GLOBAL = new RegExp(
UNICODE_EMOJI_SEQUENCE_REGEX.source,
"gu"
);
const toUnified = (value: string): string =>
Array.from(value)
.map((ch) => ch.codePointAt(0)?.toString(16))
.filter(Boolean)
.join("-");
const formatRules : FormatRule[] = [
{
pattern: [
@@ -120,17 +134,20 @@ export function TextParser(props: TextParserProps) {
}
},
{
// unicode emojis
pattern: [/\p{Emoji_Presentation}/u],
// unicode emojis (including composite sequences)
pattern: [UNICODE_EMOJI_SEQUENCE_REGEX],
render: (match: string) => {
let textWithoutEmojis = props.text.replace(/\p{Emoji_Presentation}/gu, '');
const textWithoutEmojis = props.text.replace(UNICODE_EMOJI_SEQUENCE_REGEX_GLOBAL, "");
const unified = toUnified(match);
if (textWithoutEmojis.length <= (props.oversizeIfTextSmallerThan ?? 0)) {
return <Emoji size={40} unified={match.codePointAt(0)?.toString(16) || ''}></Emoji>;
return <Emoji size={40} unified={unified}></Emoji>;
}
return <Emoji unified={match.codePointAt(0)?.toString(16) || ''}></Emoji>;
return <Emoji unified={unified}></Emoji>;
},
flush: (match: string) => {
return <Emoji unified={match.codePointAt(0)?.toString(16) || ''}></Emoji>;
return <Emoji unified={toUnified(match)}></Emoji>;
}
},
{