Add tests for RegExp.escape

Author: leobalter

features.txt

@@ -68,6 +68,10 @@ Array.fromAsync
 # https://github.com./tc39/proposal-json-parse-with-source
 json-parse-with-source
 
+# RegExp.escape
+# https://github.com./tc39/proposal-regex-escaping
+RegExp.escape
+
 # Regular expression modifiers
 # https://github.com./tc39/proposal-regexp-modifiers
 regexp-modifiers

test/built-ins/RegExp/escape/cross-realm.js

@@ -0,0 +1,17 @@
+// Copyright 2024 Leo Balter. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+esid: sec-regexp.escape
+description: escape called with a RegExp object from another realm
+features: [RegExp.escape, cross-realm]
+---*/
+
+const str = "oi+hello";
+const other = $262.createRealm().global;
+
+assert.sameValue(typeof other.RegExp.escape, "function", "other.RegExp.escape is a function");
+
+const res = other.RegExp.escape.call(RegExp, str);
+
+assert.sameValue(res, RegExp.escape(str), "cross-realm escape works correctly");

test/built-ins/RegExp/escape/escaped-control-characters.js

@@ -0,0 +1,25 @@
+// Copyright (C) 2024 Leo Balter. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+esid: sec-encodeforregexescape
+description: Encodes control characters with their ControlEscape sequences
+info: |
+  EncodeForRegExpEscape ( c )
+
+  2. If c is the code point listed in some cell of the “Code Point” column of Table 64, then
+    a. Return the string-concatenation of 0x005C (REVERSE SOLIDUS) and the string in the “ControlEscape” column of the row whose “Code Point” column contains c.
+
+  ControlEscape, Numeric Value, Code Point, Unicode Name, Symbol
+  t 9 U+0009 CHARACTER TABULATION <HT>
+  n 10 U+000A LINE FEED (LF) <LF>
+  v 11 U+000B LINE TABULATION <VT>
+  f 12 U+000C FORM FEED (FF) <FF>
+  r 13 U+000D CARRIAGE RETURN (CR) <CR>
+features: [RegExp.escape]
+---*/
+
+const controlCharacters = '\t\n\v\f\r';
+const expectedEscapedCharacters = '\\t\\n\\v\\f\\r';
+
+assert.sameValue(RegExp.escape(controlCharacters), expectedEscapedCharacters, 'Control characters are correctly escaped');

test/built-ins/RegExp/escape/escaped-lineterminator.js

@@ -0,0 +1,48 @@
+// Copyright (C) 2024 Leo Balter. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+esid: sec-regexp.escape
+description: Escaped lineterminator characters (simple assertions)
+info: |
+  EncodeForRegExpEscape ( c )
+
+  ...
+  3. Let otherPunctuators be the string-concatenation of ",-=<>#&!%:;@~'`" and the code unit 0x0022 (QUOTATION MARK).
+  4. Let toEscape be StringToCodePoints(otherPunctuators).
+  5. If toEscape ..., c is matched by WhiteSpace or LineTerminator, ..., then
+    a. If c ≤ 0xFF, then
+      i. Let hex be Number::toString(𝔽(c), 16).
+      ii. Return the string-concatenation of the code unit 0x005C (REVERSE SOLIDUS), "x", and StringPad(hex, 2, "0", START).
+    b. Let escaped be the empty String.
+    c. Let codeUnits be UTF16EncodeCodePoint(c).
+    d. For each code unit cu of codeUnits, do
+      i. Set escaped to the string-concatenation of escaped and UnicodeEscape(cu).
+    e. Return escaped.
+  6. Return UTF16EncodeCodePoint(c).
+
+  LineTerminator ::
+    <LF>
+    <CR>
+    <LS>
+    <PS>
+
+  Exceptions:
+
+    2. If c is the code point listed in some cell of the “Code Point” column of Table 64, then
+    a. Return the string-concatenation of 0x005C (REVERSE SOLIDUS) and the string in the “ControlEscape” column of the row whose “Code Point” column contains c.
+
+  ControlEscape, Numeric Value, Code Point, Unicode Name, Symbol
+  t 9 U+0009 CHARACTER TABULATION <HT>
+  n 10 U+000A LINE FEED (LF) <LF>
+  v 11 U+000B LINE TABULATION <VT>
+  f 12 U+000C FORM FEED (FF) <FF>
+  r 13 U+000D CARRIAGE RETURN (CR) <CR>
+features: [RegExp.escape]
+---*/
+
+assert.sameValue(RegExp.escape('\u2028'), '\\u2028', 'line terminator \\u2028 is escaped correctly to \\u2028');
+assert.sameValue(RegExp.escape('\u2029'), '\\u2029', 'line terminator \\u2029 is escaped correctly to \\u2029'); 
+
+assert.sameValue(RegExp.escape('\u2028\u2029'), '\\u2028\\u2029', 'line terminators are escaped correctly');
+assert.sameValue(RegExp.escape('\u2028a\u2029a'), '\\u2028a\\u2029a', 'mixed line terminators are escaped correctly');

test/built-ins/RegExp/escape/escaped-otherpunctuators.js

@@ -0,0 +1,58 @@
+// Copyright (C) 2024 Leo Balter. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+esid: sec-regexp.escape
+description: Escaped other punctuators characters
+info: |
+  EncodeForRegExpEscape ( c )
+
+  ...
+  3. Let otherPunctuators be the string-concatenation of ",-=<>#&!%:;@~'`" and the code unit 0x0022 (QUOTATION MARK).
+  4. Let toEscape be StringToCodePoints(otherPunctuators).
+  5. If toEscape contains c, (...), then
+    a. If c ≤ 0xFF, then
+      i. Let hex be Number::toString(𝔽(c), 16).
+      ii. Return the string-concatenation of the code unit 0x005C (REVERSE SOLIDUS), "x", and StringPad(hex, 2, "0", START).
+    b. Let escaped be the empty String.
+    c. Let codeUnits be UTF16EncodeCodePoint(c).
+    d. For each code unit cu of codeUnits, do
+      i. Set escaped to the string-concatenation of escaped and UnicodeEscape(cu).
+    e. Return escaped.
+  6. Return UTF16EncodeCodePoint(c).
+
+  codePoints
+  0x002c ,
+  0x002d -
+  0x003d =
+  0x003c <
+  0x003e >
+  0x0023 #
+  0x0026 &
+  0x0021 !
+  0x0025 %
+  0x003a :
+  0x003b ;
+  0x0040 @
+  0x007e ~
+  0x0027 '
+  0x0060 `
+  0x0022 "
+features: [RegExp.escape]
+---*/
+
+const otherPunctuators = ",-=<>#&!%:;@~'`\"";
+
+// Return the string-concatenation of the code unit 0x005C (REVERSE SOLIDUS), "x", and StringPad(hex, 2, "0", START).
+for (const c of otherPunctuators) {
+  const expected = `\\x${c.codePointAt(0).toString(16)}`;
+  assert.sameValue(RegExp.escape(c), expected, `${c} is escaped correctly`);
+}
+
+const otherPunctuatorsExpected = "\\x2c\\x2d\\x3d\\x3c\\x3e\\x23\\x26\\x21\\x25\\x3a\\x3b\\x40\\x7e\\x27\\x60\\x22";
+
+assert.sameValue(
+  RegExp.escape(otherPunctuators),
+  otherPunctuatorsExpected,
+  'all other punctuators are escaped correctly'
+);

test/built-ins/RegExp/escape/escaped-solidus-character-mixed.js

@@ -0,0 +1,17 @@
+// Copyright (C) 2024 Leo Balter. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+esid: sec-regexp.escape
+description: Escaped U+002F (SOLIDUS) characters (mixed assertions)
+info: |
+  EncodeForRegExpEscape ( c )
+
+  1. If c is matched by SyntaxCharacter or c is U+002F (SOLIDUS), then
+    a. Return the string-concatenation of 0x005C (REVERSE SOLIDUS) and UTF16EncodeCodePoint(c).
+features: [RegExp.escape]
+---*/
+
+assert.sameValue(RegExp.escape('.a/b'), '\\.a\\/b', 'mixed string with solidus character is escaped correctly');
+assert.sameValue(RegExp.escape('/./'), '\\/\\.\\/', 'solidus character is escaped correctly - regexp similar');
+assert.sameValue(RegExp.escape('./a\\/*b+c?d^e$f|g{2}h[i]j\\k'), '\\.\\/a\\\\\\/\\*b\\+c\\?d\\^e\\$f\\|g\\{2\\}h\\[i\\]j\\\\k', 'complex string with multiple special characters is escaped correctly');

test/built-ins/RegExp/escape/escaped-solidus-character-simple.js

@@ -0,0 +1,18 @@
+// Copyright (C) 2024 Leo Balter. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+esid: sec-regexp.escape
+description: Escaped U+002F (SOLIDUS) character (simple assertions)
+info: |
+  EncodeForRegExpEscape ( c )
+
+  1. If c is matched by SyntaxCharacter or c is U+002F (SOLIDUS), then
+    a. Return the string-concatenation of 0x005C (REVERSE SOLIDUS) and UTF16EncodeCodePoint(c).
+features: [RegExp.escape]
+---*/
+
+assert.sameValue(RegExp.escape('/'), '\\/', 'solidus character is escaped correctly');
+assert.sameValue(RegExp.escape('//'), '\\/\\/', 'solidus character is escaped correctly - multiple occurrences 1');
+assert.sameValue(RegExp.escape('///'), '\\/\\/\\/', 'solidus character is escaped correctly - multiple occurrences 2');
+assert.sameValue(RegExp.escape('////'), '\\/\\/\\/\\/', 'solidus character is escaped correctly - multiple occurrences 3');