From f5359d4656772bca0c3b10eda259cbe8ca6654d0 Mon Sep 17 00:00:00 2001
From: dakkar <dakkar@thenautilus.net>
Date: Sat, 9 Mar 2024 09:39:48 +0000
Subject: [PATCH] normalise emoji text
also, tests
---
src/internal/parser.ts | 2 +-
test/parser.ts | 24 ++++++++++++++++++++++++
2 files changed, 25 insertions(+), 1 deletion(-)
diff --git a/src/internal/parser.ts b/src/internal/parser.ts
index ef3264d..a31109b 100644
--- a/src/internal/parser.ts
+++ b/src/internal/parser.ts
@@ -629,7 +629,7 @@ export const language = P.createLanguage({
P.regexp(/[\p{Letter}\p{Number}\p{Mark}_+-]+/iu),
mark,
P.alt([P.lineEnd, side]),
- ], 2).map(name => M.EMOJI_CODE(name as string));
+ ], 2).map(name => M.EMOJI_CODE((name as string).normalize('NFC')));
},
link: r => {
diff --git a/test/parser.ts b/test/parser.ts
index d3000df..1bad6c8 100644
--- a/test/parser.ts
+++ b/test/parser.ts
@@ -338,6 +338,30 @@ hoge`;
const output = [EMOJI_CODE('abc')];
assert.deepStrictEqual(mfm.parse(input), output);
});
+
+ test('non-ASCII', () => {
+ const input = ':taneÅima_Äojas:, :मारà¥à¤œà¤¾à¤°à¤ƒ:, :é¹…:, :taneÅima_malsanas:, :à¹à¸¡à¸§:, and :लकà¥à¤·à¤£à¤¾:';
+ const output = [
+ EMOJI_CODE('taneÅima_Äojas'),
+ TEXT(', '),
+ EMOJI_CODE('मारà¥à¤œà¤¾à¤°à¤ƒ'),
+ TEXT(', '),
+ EMOJI_CODE('é¹…'),
+ TEXT(', '),
+ EMOJI_CODE('taneÅima_malsanas'),
+ TEXT(', '),
+ EMOJI_CODE('à¹à¸¡à¸§'),
+ TEXT(', and '),
+ EMOJI_CODE('लकà¥à¤·à¤£à¤¾'),
+ ];
+ assert.deepStrictEqual(mfm.parse(input), output);
+ });
+
+ test('non-ASCII normalization', () => {
+ const input = ":fo\u{0308}o:";
+ const output = [EMOJI_CODE("f\u{00F6}o")];
+ assert.deepStrictEqual(mfm.parse(input), output);
+ });
});
describe('unicode emoji', () => {
--
GitLab