[GR-44464] Implement Well-Formed Unicode Strings proposal.

PullRequest: js/2738

Author: woess

CHANGELOG.md

@@ -9,6 +9,7 @@ See [release calendar](https://www.graalvm.org/release-calendar/) for release da
 * Implemented the [WebAssembly threads](https://github.com/WebAssembly/threads) proposal.
 * Implemented the [Promise.withResolvers](https://github.com/tc39/proposal-promise-with-resolvers) proposal. It is available in ECMAScript staging mode (`--js.ecmascript-version=staging`).
 * Implementation of [Async Iterator Helpers](https://github.com/tc39/proposal-async-iterator-helpers) proposal (that was split out from Iterator Helpers proposal) was moved behind the experimental option `--js.async-iterator-helpers`.
+* Implemented the [Well-Formed Unicode Strings](https://github.com/tc39/proposal-is-usv-string) proposal. It is available in ECMAScript staging mode (`--js.ecmascript-version=staging`).
 
 ## Version 23.1.0
 * NOTE: GraalVM no longer ships with a "js" ScriptEngine. Please either use the Maven dependency or explicitly put `js-scriptengine.jar` on the module path. See [ScriptEngine documentation](docs/user/ScriptEngine.md) for details.

graal-js/src/com.oracle.truffle.js.test.external/src/com/oracle/truffle/js/test/external/test262/Test262Runnable.java

@@ -149,8 +149,10 @@ public class Test262Runnable extends TestRunnable {
                     "String.prototype.at",
                     "String.prototype.endsWith",
                     "String.prototype.includes",
+                    "String.prototype.isWellFormed",
                     "String.prototype.matchAll",
                     "String.prototype.replaceAll",
+                    "String.prototype.toWellFormed",
                     "String.prototype.trimEnd",
                     "String.prototype.trimStart",
                     "Symbol",
@@ -256,8 +258,6 @@ public class Test262Runnable extends TestRunnable {
                     "Array.fromAsync",
                     "Intl.DurationFormat",
                     "IsHTMLDDA",
-                    "String.prototype.isWellFormed",
-                    "String.prototype.toWellFormed",
                     "arbitrary-module-namespace-names",
                     "resizable-arraybuffer",
                     "tail-call-optimization",
@@ -267,6 +267,8 @@ public class Test262Runnable extends TestRunnable {
                     "FinalizationRegistry.prototype.cleanupSome",
                     "Intl.Locale-info",
                     "ShadowRealm",
+                    "String.prototype.isWellFormed",
+                    "String.prototype.toWellFormed",
                     "array-grouping",
                     "arraybuffer-transfer",
                     "decorators",

graal-js/src/com.oracle.truffle.js.test/js/string_wellformed.js

@@ -0,0 +1,30 @@
+/*
+ * Copyright (c) 2023, 2023, Oracle and/or its affiliates. All rights reserved.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * Licensed under the Universal Permissive License v 1.0 as shown at http://oss.oracle.com/licenses/upl.
+ */
+/**
+ * Test for String.prototype.isWellFormed() and String.prototype.toWellFormed().
+ *
+ * @option ecmascript-version=staging
+ */
+load("assert.js");
+
+assertTrue("\u0000".isWellFormed());
+assertTrue("\ud000".isWellFormed());
+assertTrue("\ue000".isWellFormed());
+assertTrue("\uffff".isWellFormed());
+
+assertFalse("\ud800".isWellFormed());
+assertFalse("\udbff".isWellFormed());
+assertFalse("\udc00".isWellFormed());
+assertFalse("\udfff".isWellFormed());
+
+assertSame("\ufffd", "\ud800".toWellFormed());
+assertSame("\ufffd", "\udbff".toWellFormed());
+assertSame("\ufffd", "\udc00".toWellFormed());
+assertSame("\ufffd", "\udfff".toWellFormed());
+
+assertTrue("\ud83d\ude00".isWellFormed());
+assertSame("\ud83d\ude00", "\ud83d\ude00".toWellFormed());

graal-js/src/com.oracle.truffle.js/src/com/oracle/truffle/js/builtins/StringPrototypeBuiltins.java

@@ -83,6 +83,7 @@
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringEndsWithNodeGen;
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringIncludesNodeGen;
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringIndexOfNodeGen;
+import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringIsWellFormedNodeGen;
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringLastIndexOfNodeGen;
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringLocaleCompareIntlNodeGen;
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringLocaleCompareNodeGen;
@@ -106,6 +107,7 @@
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringToLowerCaseNodeGen;
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringToStringNodeGen;
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringToUpperCaseNodeGen;
+import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringToWellFormedNodeGen;
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringTrimLeftNodeGen;
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringTrimNodeGen;
 import com.oracle.truffle.js.builtins.StringPrototypeBuiltinsFactory.JSStringTrimRightNodeGen;
@@ -235,7 +237,11 @@ public enum StringPrototype implements BuiltinEnum<StringPrototype> {
         replaceAll(2),
 
         // ES2022
-        at(1);
+        at(1),
+
+        // ES2024
+        isWellFormed(0),
+        toWellFormed(0);
 
         private final int length;
 
@@ -265,6 +271,8 @@ public int getECMAScriptVersion() {
                 return JSConfig.ECMAScript2021;
             } else if (at == this) {
                 return JSConfig.ECMAScript2022;
+            } else if (EnumSet.of(isWellFormed, toWellFormed).contains(this)) {
+                return JSConfig.ECMAScript2024;
             }
             return BuiltinEnum.super.getECMAScriptVersion();
         }
@@ -396,6 +404,11 @@ protected Object createNode(JSContext context, JSBuiltin builtin, boolean constr
 
             case at:
                 return JSStringAtNodeGen.create(context, builtin, args().withThis().fixedArgs(1).createArgumentNodes(context));
+
+            case isWellFormed:
+                return JSStringIsWellFormedNodeGen.create(context, builtin, args().withThis().createArgumentNodes(context));
+            case toWellFormed:
+                return JSStringToWellFormedNodeGen.create(context, builtin, args().withThis().createArgumentNodes(context));
         }
         return null;
     }
@@ -2971,6 +2984,46 @@ protected Object pad(Object thisObj, Object[] args,
 
     }
 
+    public abstract static class JSStringIsWellFormedNode extends JSStringOperation {
+
+        public JSStringIsWellFormedNode(JSContext context, JSBuiltin builtin) {
+            super(context, builtin);
+        }
+
+        @Specialization
+        protected static Object doString(TruffleString thisStr,
+                        @Shared @Cached TruffleString.IsValidNode isValidNode) {
+            return isValidNode.execute(thisStr, TruffleString.Encoding.UTF_16);
+        }
+
+        @Specialization(guards = "!isString(thisObj)")
+        protected final Object doOther(Object thisObj,
+                        @Shared @Cached TruffleString.IsValidNode isValidNode) {
+            requireObjectCoercible(thisObj);
+            return doString(toString(thisObj), isValidNode);
+        }
+    }
+
+    public abstract static class JSStringToWellFormedNode extends JSStringOperation {
+
+        public JSStringToWellFormedNode(JSContext context, JSBuiltin builtin) {
+            super(context, builtin);
+        }
+
+        @Specialization
+        protected static TruffleString doString(TruffleString thisStr,
+                        @Shared @Cached TruffleString.ToValidStringNode toValidNode) {
+            return toValidNode.execute(thisStr, TruffleString.Encoding.UTF_16);
+        }
+
+        @Specialization(guards = "!isString(thisObj)")
+        protected final TruffleString doOther(Object thisObj,
+                        @Shared @Cached TruffleString.ToValidStringNode toValidNode) {
+            requireObjectCoercible(thisObj);
+            return doString(toString(thisObj), toValidNode);
+        }
+    }
+
     /**
      * Implementation of the CreateRegExpStringIterator abstract operation as specified by the
      * String.prototype.matchAll draft proposal.