Add tests for RegExp Buffer Boundaries

features.txt

@@ -43,6 +43,10 @@ decorators
 # https://github.com/tc39/proposal-duplicate-named-capturing-groups
 regexp-duplicate-named-groups
 
+# RegExp Buffer Boundaries
+# https://github.com/tc39/proposal-regexp-buffer-boundaries
+regexp-buffer-boundaries
+
 # https://tc39.es/proposal-array-from-async/
 Array.fromAsync
 

test/annexB/built-ins/RegExp/buffer-boundaries/not-supported-outside-unicode-modes.js

@@ -0,0 +1,44 @@
+// Copyright 2026 Ron Buckton. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+author: Ron Buckton
+description: >
+  `\A` and `\z` (and `\Z`) are treated as IdentityEscape in Annex B outside any unicode mode
+info: |
+  Patterns
+
+  Assertion[UnicodeMode, UnicodeSetsMode, NamedCaptureGroups] :
+    ...
+    [+UnicodeMode] `\` `A`
+    [+UnicodeMode] `\` `z`
+    ...
+
+  IdentityEscape[UnicodeMode, NamedCaptureGroups] :
+    [+UnicodeMode] SyntaxCharacter
+    [+UnicodeMode] `/`
+    [~UnicodeMode] SourceCharacterIdentityEscape[?NamedCaptureGroups]
+
+  SourceCharacterIdentityEscape[NamedCaptureGroups] ::
+    [~NamedCaptureGroups] SourceCharacter but not `c`
+    [+NamedCaptureGroups] SourceCharacter but not one of `c` or `k`
+
+esid: sec-regular-expressions-patterns
+features: [regexp-buffer-boundaries]
+---*/
+
+// `\A` is |IdentityEscape| outside of any unicode mode
+assert(/^\A$/.test("A"), "Expected \\A to match literal 'A' outside of unicode mode.");
+assert(!/\Ax/.test("x"), "Expected \\A to not match start of buffer outside of unicode mode.");
+assert(new RegExp("^\\A$").test("A"), "Expected \\A to match literal 'A' outside of unicode mode.");
+assert(!new RegExp("\\Ax").test("x"), "Expected \\A to not match start of buffer outside of unicode mode.");
+
+// `\z` is |IdentityEscape| outside of any unicode mode
+assert(/^\z$/.test("z"), "Expected \\z to match literal 'z' outside of unicode mode.");
+assert(!/x\z/.test("x"), "Expected \\z to not match end of buffer outside of unicode mode.");
+assert(new RegExp("^\\z$").test("z"), "Expected \\z to match literal 'z' outside of unicode mode.");
+assert(!new RegExp("x\\z").test("x"), "Expected \\z to not match end of buffer outside of unicode mode.");
+
+// NOTE: ensures reserved `\Z` is |IdentityEscape| outside of any unicode mode
+assert(/^\Z$/.test("Z"), "Expected \\Z to match literal 'Z' outside of unicode mode.");
+assert(new RegExp("^\\Z$").test("Z"), "Expected \\Z to match literal 'Z' outside of unicode mode.");

test/built-ins/RegExp/buffer-boundaries/slash-lower-case-z-matches-end-of-buffer.js

@@ -0,0 +1,51 @@
+// Copyright 2026 Ron Buckton. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+author: Ron Buckton
+description: >
+  `\z` (lower-case z) matches end of buffer in any unicode mode.
+info: |
+  Runtime Semantics: CompileAssertion
+
+  Assertion :: \z
+    1. Return a new Matcher with parameters (x, c) that captures nothing and performs the following steps when called:
+      a. Assert: x is a MatchState.
+      b. Assert: c is a MatcherContinuation.
+      c. Let Input be x.[[Input]].
+      d. Let e be x.[[EndIndex]].
+      e. Let InputLength be the number of elements in Input.
+      f. If e = InputLength, return c(x).
+      g. Return failure.
+
+esid: sec-patterns
+features: [regexp-buffer-boundaries]
+---*/
+
+assert(/x\z/u.test("x"), "Expected \\z to match the end of the buffer outside of multiline mode");
+assert(/x\z/um.test("x"), "Expected \\z to match the end of the buffer inside of multiline mode");
+assert(!/x\z/u.test("xy"), "Expected \\z to not match when not at the end of the buffer outside of multiline mode");
+assert(!/x\z/um.test("xy"), "Expected \\z to not match when not at the end of the buffer inside of multiline mode");
+assert(!/x\z/u.test("x\ny"), "Expected \\z to not match when only at the end of a line outside of multiline mode");
+assert(!/x\z/um.test("x\ny"), "Expected \\z to not match when only at the end of a line inside of multiline mode");
+
+assert(/x\z/v.test("x"), "Expected \\z to match the end of the buffer outside of multiline mode");
+assert(/x\z/vm.test("x"), "Expected \\z to match the end of the buffer inside of multiline mode");
+assert(!/x\z/v.test("xy"), "Expected \\z to not match when not at the end of the buffer outside of multiline mode");
+assert(!/x\z/vm.test("xy"), "Expected \\z to not match when not at the end of the buffer inside of multiline mode");
+assert(!/x\z/v.test("x\ny"), "Expected \\z to not match when only at the end of a line outside of multiline mode");
+assert(!/x\z/vm.test("x\ny"), "Expected \\z to not match when only at the end of a line inside of multiline mode");
+
+assert(new RegExp("x\\z", "u").test("x"), "Expected \\z to match the end of the buffer outside of multiline mode");
+assert(new RegExp("x\\z", "um").test("x"), "Expected \\z to match the end of the buffer inside of multiline mode");
+assert(!new RegExp("x\\z", "u").test("xy"), "Expected \\z to not match when not at the end of the buffer outside of multiline mode");
+assert(!new RegExp("x\\z", "um").test("xy"), "Expected \\z to not match when not at the end of the buffer inside of multiline mode");
+assert(!new RegExp("x\\z", "u").test("x\ny"), "Expected \\z to not match when only at the end of a line outside of multiline mode");
+assert(!new RegExp("x\\z", "um").test("x\ny"), "Expected \\z to not match when only at the end of a line inside of multiline mode");
+
+assert(new RegExp("x\\z", "v").test("x"), "Expected \\z to match the end of the buffer outside of multiline mode");
+assert(new RegExp("x\\z", "vm").test("x"), "Expected \\z to match the end of the buffer inside of multiline mode");
+assert(!new RegExp("x\\z", "v").test("xy"), "Expected \\z to not match when not at the end of the buffer outside of multiline mode");
+assert(!new RegExp("x\\z", "vm").test("xy"), "Expected \\z to not match when not at the end of the buffer inside of multiline mode");
+assert(!new RegExp("x\\z", "v").test("x\ny"), "Expected \\z to not match when only at the end of a line outside of multiline mode");
+assert(!new RegExp("x\\z", "vm").test("x\ny"), "Expected \\z to not match when only at the end of a line inside of multiline mode");

test/built-ins/RegExp/buffer-boundaries/slash-upper-case-a-matches-start-of-buffer.js

@@ -0,0 +1,49 @@
+// Copyright 2026 Ron Buckton. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+author: Ron Buckton
+description: >
+  `\A` matches start of buffer in any unicode mode.
+info: |
+  Runtime Semantics: CompileAssertion
+
+  Assertion :: \A
+    1. Return a new Matcher with parameters (x, c) that captures nothing and performs the following steps when called:
+      a. Assert: x is a MatchState.
+      b. Assert: c is a MatcherContinuation.
+      c. Let e be x.[[EndIndex]].
+      d. If e = 0, return c(x).
+      e. Return ~failure~.
+
+esid: sec-patterns
+features: [regexp-buffer-boundaries]
+---*/
+
+assert(/\Ax/u.test("x"), "Expected \\A to match the start of the buffer outside of multiline mode");
+assert(/\Ax/um.test("x"), "Expected \\A to match the start of the buffer inside of multiline mode");
+assert(!/\Ax/u.test("yx"), "Expected \\A to not match when not at the start of the buffer outside of multiline mode");
+assert(!/\Ax/um.test("yx"), "Expected \\A to not match when not at the start of the buffer inside of multiline mode");
+assert(!/\Ax/u.test("y\nx"), "Expected \\A to not match when at the start of a new line outside of multiline mode");
+assert(!/\Ax/um.test("y\nx"), "Expected \\A to not match when at the start of a new line inside of multiline mode");
+
+assert(/\Ax/v.test("x"), "Expected \\A to match the start of the buffer outside of multiline mode");
+assert(/\Ax/vm.test("x"), "Expected \\A to match the start of the buffer inside of multiline mode");
+assert(!/\Ax/v.test("yx"), "Expected \\A to not match when not at the start of the buffer outside of multiline mode");
+assert(!/\Ax/vm.test("yx"), "Expected \\A to not match when not at the start of the buffer inside of multiline mode");
+assert(!/\Ax/v.test("y\nx"), "Expected \\A to not match when at the start of a new line outside of multiline mode");
+assert(!/\Ax/vm.test("y\nx"), "Expected \\A to not match when at the start of a new line inside of multiline mode");
+
+assert(new RegExp("\\Ax", "u").test("x"), "Expected \\A to match the start of the buffer outside of multiline mode");
+assert(new RegExp("\\Ax", "um").test("x"), "Expected \\A to match the start of the buffer inside of multiline mode");
+assert(!new RegExp("\\Ax", "u").test("yx"), "Expected \\A to not match when not at the start of the buffer outside of multiline mode");
+assert(!new RegExp("\\Ax", "um").test("yx"), "Expected \\A to not match when not at the start of the buffer inside of multiline mode");
+assert(!new RegExp("\\Ax", "u").test("y\nx"), "Expected \\A to not match when at the start of a new line outside of multiline mode");
+assert(!new RegExp("\\Ax", "um").test("y\nx"), "Expected \\A to not match when at the start of a new line inside of multiline mode");
+
+assert(new RegExp("\\Ax", "v").test("x"), "Expected \\A to match the start of the buffer outside of multiline mode");
+assert(new RegExp("\\Ax", "vm").test("x"), "Expected \\A to match the start of the buffer inside of multiline mode");
+assert(!new RegExp("\\Ax", "v").test("yx"), "Expected \\A to not match when not at the start of the buffer outside of multiline mode");
+assert(!new RegExp("\\Ax", "vm").test("yx"), "Expected \\A to not match when not at the start of the buffer inside of multiline mode");
+assert(!new RegExp("\\Ax", "v").test("y\nx"), "Expected \\A to not match when at the start of a new line outside of multiline mode");
+assert(!new RegExp("\\Ax", "vm").test("y\nx"), "Expected \\A to not match when at the start of a new line inside of multiline mode");

test/built-ins/RegExp/buffer-boundaries/syntax/u-mode.js

@@ -0,0 +1,48 @@
+// Copyright 2026 Ron Buckton. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+author: Ron Buckton
+description: >
+  `\A` and `\z` are parsed successfully in u-mode
+info: |
+  Patterns
+
+  Assertion[UnicodeMode, UnicodeSetsMode, NamedCaptureGroups] :
+    ...
+    [+UnicodeMode] `\` `A`
+    [+UnicodeMode] `\` `z`
+    ...
+
+esid: sec-patterns
+features: [regexp-buffer-boundaries]
+---*/
+
+/\A/u;
+/\A/um;
+/\A/umi;
+/\A/us;
+/\A/usi;
+/\A/usm;
+/\A/usmi;
+/\z/u;
+/\z/um;
+/\z/umi;
+/\z/us;
+/\z/usi;
+/\z/usm;
+/\z/usmi;
+new RegExp("\\A", "u");
+new RegExp("\\A", "um");
+new RegExp("\\A", "umi");
+new RegExp("\\A", "us");
+new RegExp("\\A", "usi");
+new RegExp("\\A", "usm");
+new RegExp("\\A", "usmi");
+new RegExp("\\z", "u");
+new RegExp("\\z", "um");
+new RegExp("\\z", "umi");
+new RegExp("\\z", "us");
+new RegExp("\\z", "usi");
+new RegExp("\\z", "usm");
+new RegExp("\\z", "usmi");

test/built-ins/RegExp/buffer-boundaries/syntax/upper-z-escape-reserved-in-unicode-modes-literal-u-mode.js

@@ -0,0 +1,23 @@
+// Copyright 2026 Ron Buckton. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+author: Ron Buckton
+description: >
+  `\Z` (upper-case Z) is reserved in Unicode ('u') mode.
+info: |
+  Patterns
+
+  Note: While the sequence `\Z` is already an error when either the `u` (unicode) or `v` (unicode sets) modes are
+  present, it is reserved for possible future use as an extension of the `\A` and `\z` assertions.
+
+esid: sec-patterns
+features: [regexp-buffer-boundaries]
+negative:
+  phase: parse
+  type: SyntaxError
+---*/
+
+$DONOTEVALUATE();
+
+/\Z/u;

test/built-ins/RegExp/buffer-boundaries/syntax/upper-z-escape-reserved-in-unicode-modes-literal-v-mode.js

@@ -0,0 +1,23 @@
+// Copyright 2026 Ron Buckton. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+author: Ron Buckton
+description: >
+  `\Z` (upper-case Z) is reserved in Unicode Sets ('v') mode.
+info: |
+  Patterns
+
+  Note: While the sequence `\Z` is already an error when either the `u` (unicode) or `v` (unicode sets) modes are
+  present, it is reserved for possible future use as an extension of the `\A` and `\z` assertions.
+
+esid: sec-patterns
+features: [regexp-buffer-boundaries]
+negative:
+  phase: parse
+  type: SyntaxError
+---*/
+
+$DONOTEVALUATE();
+
+/\Z/v;

test/built-ins/RegExp/buffer-boundaries/syntax/upper-z-escape-reserved-in-unicode-modes.js

@@ -0,0 +1,19 @@
+// Copyright 2026 Ron Buckton. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+author: Ron Buckton
+description: >
+  `\Z` (upper-case Z) is reserved in any unicode mode.
+info: |
+  Patterns
+
+  Note: While the sequence `\Z` is already an error when either the `u` (unicode) or `v` (unicode sets) modes are
+  present, it is reserved for possible future use as an extension of the `\A` and `\z` assertions.
+
+esid: sec-patterns
+features: [regexp-buffer-boundaries]
+---*/
+
+assert.throws(SyntaxError, function() { new RegExp("\\Z", "u"); }, 'new RegExp("\\Z", "u")');
+assert.throws(SyntaxError, function() { new RegExp("\\Z", "v"); }, 'new RegExp("\\Z", "v")');

test/built-ins/RegExp/buffer-boundaries/syntax/v-mode.js

@@ -0,0 +1,48 @@
+// Copyright 2026 Ron Buckton. All rights reserved.
+// This code is governed by the BSD license found in the LICENSE file.
+
+/*---
+author: Ron Buckton
+description: >
+  `\A` and `\z` are parsed successfully in v-mode
+info: |
+  Patterns
+
+  Assertion[UnicodeMode, UnicodeSetsMode, NamedCaptureGroups] :
+    ...
+    [+UnicodeMode] `\` `A`
+    [+UnicodeMode] `\` `z`
+    ...
+
+esid: sec-patterns
+features: [regexp-buffer-boundaries]
+---*/
+
+/\A/v;
+/\A/vm;
+/\A/vmi;
+/\A/vs;
+/\A/vsi;
+/\A/vsm;
+/\A/vsmi;
+/\z/v;
+/\z/vm;
+/\z/vmi;
+/\z/vs;
+/\z/vsi;
+/\z/vsm;
+/\z/vsmi;
+new RegExp("\\A", "v");
+new RegExp("\\A", "vm");
+new RegExp("\\A", "vmi");
+new RegExp("\\A", "vs");
+new RegExp("\\A", "vsi");
+new RegExp("\\A", "vsm");
+new RegExp("\\A", "vsmi");
+new RegExp("\\z", "v");
+new RegExp("\\z", "vm");
+new RegExp("\\z", "vmi");
+new RegExp("\\z", "vs");
+new RegExp("\\z", "vsi");
+new RegExp("\\z", "vsm");
+new RegExp("\\z", "vsmi");

test/staging/sm/RegExp/unicode-disallow-extended.js

@@ -13,7 +13,6 @@ esid: pending
 
 assert.compareArray(/\^\$\\\.\*\+\?\(\)\[\]\{\}\|/u.exec("^$\\.*+?()[]{}|"),
               ["^$\\.*+?()[]{}|"]);
-assert.throws(SyntaxError, () => eval(`/\\A/u`));
 assert.throws(SyntaxError, () => eval(`/\\-/u`));
 assert.throws(SyntaxError, () => eval(`/\\U{10}/u`));
 assert.throws(SyntaxError, () => eval(`/\\U0000/u`));