improve hashtag detection (#43)

src/parser.pegjs

@@ -66,7 +66,7 @@ fullParser
 	= nodes:(&. n:(block / inline) { return n; })* { return mergeText(nodes); }
 
 plainParser
-	= nodes:(&. n:(emojiCode / unicodeEmoji / text) { return n; })* { return mergeText(nodes); }
+	= nodes:(&. n:(emojiCode / unicodeEmoji / plainText) { return n; })* { return mergeText(nodes); }
 
 inlineParser
 	= nodes:(&. n:inline { return n; })* { return mergeText(nodes); }
@@ -165,7 +165,7 @@ inline
 	/ url
 	/ link
 	/ fn
-	/ text
+	/ inlineText
 
 // inline: emoji code
 
@@ -289,7 +289,7 @@ mentionHostPart
 // inline: hashtag
 
 hashtag
-	= "#" content:hashtagContent
+	= "#" !("\uFE0F"? "\u20E3") content:hashtagContent
 {
 	return HASHTAG(content);
 }
@@ -382,7 +382,13 @@ fnArg
 
 // inline: text
 
-text
+inlineText
+	= !(LF / _) . &hashtag . { return text(); } // hashtag ignore
+	/ . /* text node */
+
+// inline: text (for plainParser)
+
+plainText
 	= . /* text node */
 
 //

test/parser.ts

@@ -4,7 +4,29 @@ import {
 	TEXT, CENTER, FN, UNI_EMOJI, MENTION, EMOJI_CODE, HASHTAG, N_URL, BOLD, SMALL, ITALIC, STRIKE, QUOTE, MATH_BLOCK, SEARCH, CODE_BLOCK, LINK
 } from '../built/index';
 
-describe('parser', () => {
+describe('PlainParser', () => {
+	describe('text', () => {
+		it('basic', () => {
+			const input = 'abc';
+			const output = [TEXT('abc')];
+			assert.deepStrictEqual(mfm.parsePlain(input), output);
+		});
+
+		it('ignore hashtag', () => {
+			const input = 'abc#abc';
+			const output = [TEXT('abc#abc')];
+			assert.deepStrictEqual(mfm.parsePlain(input), output);
+		});
+
+		it('keycap number sign', () => {
+			const input = 'abc#️⃣abc';
+			const output = [TEXT('abc'), UNI_EMOJI('#️⃣'), TEXT('abc')];
+			assert.deepStrictEqual(mfm.parsePlain(input), output);
+		});
+	});
+});
+
+describe('FullParser', () => {
 	describe('text', () => {
 		it('普通のテキストを入力すると1つのテキストノードが返される', () => {
 			const input = 'abc';
@@ -222,6 +244,12 @@ describe('parser', () => {
 			const output = [TEXT('今起きた'), UNI_EMOJI('😇')];
 			assert.deepStrictEqual(mfm.parse(input), output);
 		});
+
+		it('keycap number sign', () => {
+			const input = 'abc#️⃣123';
+			const output = [TEXT('abc'), UNI_EMOJI('#️⃣'), TEXT('123')];
+			assert.deepStrictEqual(mfm.parse(input), output);
+		});
 	});
 
 	describe('big', () => {
@@ -397,9 +425,28 @@ describe('parser', () => {
 	// mention
 
 	describe('hashtag', () => {
-		it('and unicode emoji', () => {
-			const input = '#️⃣abc123#abc';
-			const output = [UNI_EMOJI('#️⃣'), TEXT('abc123'), HASHTAG('abc')];
+		it('basic', () => {
+			const input = 'before #abc after';
+			const output = [TEXT('before '), HASHTAG('abc'), TEXT(' after')];
+			assert.deepStrictEqual(mfm.parse(input), output);
+		});
+
+		it('with keycap number sign', () => {
+			const input = '#️⃣abc123 #abc';
+			const output = [UNI_EMOJI('#️⃣'), TEXT('abc123 '), HASHTAG('abc')];
+			assert.deepStrictEqual(mfm.parse(input), output);
+		});
+
+		it('with keycap number sign 2', () => {
+			const input = `abc
+#️⃣abc`;
+			const output = [TEXT('abc\n'), UNI_EMOJI('#️⃣'), TEXT('abc')];
+			assert.deepStrictEqual(mfm.parse(input), output);
+		});
+
+		it('ignore a hashtag if the before char is neither a space nor an LF', () => {
+			const input = 'abc#abc';
+			const output = [TEXT('abc#abc')];
 			assert.deepStrictEqual(mfm.parse(input), output);
 		});
 	});