JS: Improved number pattern (#3149)

PrismJS · Oct 14, 2021 · 5a24cbf · 5a24cbf
1 parent 3c61c8f
commit 5a24cbf
Show file tree

Hide file tree

Showing 4 changed files with 127 additions and 4 deletions.
diff --git a/components/prism-javascript.js b/components/prism-javascript.js
@@ -18,7 +18,34 @@ Prism.languages.javascript = Prism.languages.extend('clike', {
 	],
 	// Allow for all non-ASCII characters (See http://stackoverflow.com/a/2008444)
 	'function': /#?(?!\s)[_$a-zA-Z\xA0-\uFFFF](?:(?!\s)[$\w\xA0-\uFFFF])*(?=\s*(?:\.\s*(?:apply|bind|call)\s*)?\()/,
-	'number': /\b(?:(?:0[xX](?:[\dA-Fa-f](?:_[\dA-Fa-f])?)+|0[bB](?:[01](?:_[01])?)+|0[oO](?:[0-7](?:_[0-7])?)+)n?|(?:\d(?:_\d)?)+n|NaN|Infinity)\b|(?:\b(?:\d(?:_\d)?)+\.?(?:\d(?:_\d)?)*|\B\.(?:\d(?:_\d)?)+)(?:[Ee][+-]?(?:\d(?:_\d)?)+)?/,
+	'number': {
+		pattern: RegExp(
+			/(^|[^\w$])/.source +
+			'(?:' +
+			(
+				// constant
+				/NaN|Infinity/.source +
+				'|' +
+				// binary integer
+				/0[bB][01]+(?:_[01]+)*n?/.source +
+				'|' +
+				// octal integer
+				/0[oO][0-7]+(?:_[0-7]+)*n?/.source +
+				'|' +
+				// hexadecimal integer
+				/0[xX][\dA-Fa-f]+(?:_[\dA-Fa-f]+)*n?/.source +
+				'|' +
+				// decimal bigint
+				/\d+(?:_\d+)*n/.source +
+				'|' +
+				// decimal number (integer or float) but no bigint
+				/(?:\d+(?:_\d+)*(?:\.(?:\d+(?:_\d+)*)?)?|\.\d+(?:_\d+)*)(?:[Ee][+-]?\d+(?:_\d+)*)?/.source
+			) +
+			')' +
+			/(?![\w$])/.source
+		),
+		lookbehind: true
+	},
 	'operator': /--|\+\+|\*\*=?|=>|&&=?|\|\|=?|[!=]==|<<=?|>>>?=?|[-+*/%&|^!=<>]=?|\.{3}|\?\?=?|\?\.?|[~:]/
 });
 

diff --git a/components/prism-javascript.min.js b/components/prism-javascript.min.js
diff --git a/prism.js b/prism.js
@@ -1567,7 +1567,34 @@ Prism.languages.javascript = Prism.languages.extend('clike', {
 	],
 	// Allow for all non-ASCII characters (See http://stackoverflow.com/a/2008444)
 	'function': /#?(?!\s)[_$a-zA-Z\xA0-\uFFFF](?:(?!\s)[$\w\xA0-\uFFFF])*(?=\s*(?:\.\s*(?:apply|bind|call)\s*)?\()/,
-	'number': /\b(?:(?:0[xX](?:[\dA-Fa-f](?:_[\dA-Fa-f])?)+|0[bB](?:[01](?:_[01])?)+|0[oO](?:[0-7](?:_[0-7])?)+)n?|(?:\d(?:_\d)?)+n|NaN|Infinity)\b|(?:\b(?:\d(?:_\d)?)+\.?(?:\d(?:_\d)?)*|\B\.(?:\d(?:_\d)?)+)(?:[Ee][+-]?(?:\d(?:_\d)?)+)?/,
+	'number': {
+		pattern: RegExp(
+			/(^|[^\w$])/.source +
+			'(?:' +
+			(
+				// constant
+				/NaN|Infinity/.source +
+				'|' +
+				// binary integer
+				/0[bB][01]+(?:_[01]+)*n?/.source +
+				'|' +
+				// octal integer
+				/0[oO][0-7]+(?:_[0-7]+)*n?/.source +
+				'|' +
+				// hexadecimal integer
+				/0[xX][\dA-Fa-f]+(?:_[\dA-Fa-f]+)*n?/.source +
+				'|' +
+				// decimal bigint
+				/\d+(?:_\d+)*n/.source +
+				'|' +
+				// decimal number (integer or float) but no bigint
+				/(?:\d+(?:_\d+)*(?:\.(?:\d+(?:_\d+)*)?)?|\.\d+(?:_\d+)*)(?:[Ee][+-]?\d+(?:_\d+)*)?/.source
+			) +
+			')' +
+			/(?![\w$])/.source
+		),
+		lookbehind: true
+	},
 	'operator': /--|\+\+|\*\*=?|=>|&&=?|\|\|=?|[!=]==|<<=?|>>>?=?|[-+*/%&|^!=<>]=?|\.{3}|\?\?=?|\?\.?|[~:]/
 });
 

diff --git a/tests/languages/javascript/number_feature.test b/tests/languages/javascript/number_feature.test
@@ -23,6 +23,39 @@ Infinity
 0.000_001
 1e10_000
 
+0xf_f_f_f_ffn
+0xf_f_f_f_f
+0b1_1_1_1
+0b0_0_1_0_1_0_1n
+0o1_1_3_2
+0o1_1_3_2n
+2.2_4_54e64_33
+2.2_4_54e+64_33
+2.2_4_54e-64_33
+3_4_5
+3_4_5n
+3_4_5.5_5_6
+
+// not numbers
+$1234;
+$0xFF;
+$0b0101;
+$1234n;
+$0xFFn;
+$0b0101n;
+_1234;
+_0xFF;
+_0b0101;
+_1234n;
+_0xFFn;
+_0b0101n;
+abc1234;
+abc0xFF;
+abc0b0101;
+abc1234n;
+abc0xFFn;
+abc0b0101n;
+
 ----------------------------------------------------
 
 [
@@ -35,18 +68,54 @@ Infinity
 	["number", "0o571"],
 	["number", "0xbabe"],
 	["number", "0xBABE"],
+
 	["number", "NaN"],
 	["number", "Infinity"],
+
 	["number", "123n"],
 	["number", "0x123n"],
+
 	["number", "1_000_000_000_000"],
 	["number", "1_000_000.220_720"],
 	["number", "0b0101_0110_0011_1000"],
 	["number", "0o12_34_56"],
 	["number", "0x40_76_38_6A_73"],
 	["number", "4_642_473_943_484_686_707n"],
 	["number", "0.000_001"],
-	["number", "1e10_000"]
+	["number", "1e10_000"],
+
+	["number", "0xf_f_f_f_ffn"],
+	["number", "0xf_f_f_f_f"],
+	["number", "0b1_1_1_1"],
+	["number", "0b0_0_1_0_1_0_1n"],
+	["number", "0o1_1_3_2"],
+	["number", "0o1_1_3_2n"],
+	["number", "2.2_4_54e64_33"],
+	["number", "2.2_4_54e+64_33"],
+	["number", "2.2_4_54e-64_33"],
+	["number", "3_4_5"],
+	["number", "3_4_5n"],
+	["number", "3_4_5.5_5_6"],
+
+	["comment", "// not numbers"],
+	"\r\n$1234", ["punctuation", ";"],
+	"\r\n$0xFF", ["punctuation", ";"],
+	"\r\n$0b0101", ["punctuation", ";"],
+	"\r\n$1234n", ["punctuation", ";"],
+	"\r\n$0xFFn", ["punctuation", ";"],
+	"\r\n$0b0101n", ["punctuation", ";"],
+	"\r\n_1234", ["punctuation", ";"],
+	"\r\n_0xFF", ["punctuation", ";"],
+	"\r\n_0b0101", ["punctuation", ";"],
+	"\r\n_1234n", ["punctuation", ";"],
+	"\r\n_0xFFn", ["punctuation", ";"],
+	"\r\n_0b0101n", ["punctuation", ";"],
+	"\r\nabc1234", ["punctuation", ";"],
+	"\r\nabc0xFF", ["punctuation", ";"],
+	"\r\nabc0b0101", ["punctuation", ";"],
+	"\r\nabc1234n", ["punctuation", ";"],
+	"\r\nabc0xFFn", ["punctuation", ";"],
+	"\r\nabc0b0101n", ["punctuation", ";"]
 ]
 
 ----------------------------------------------------