Add error message for keywords with escapes in them (microsoft#32718)

weswigham · web-flow · commit 4ab85bbf350b · 2019-08-12T16:00:38.000-07:00
* Add error message for keywords with escapes in them

* Move check into parser during advance to next token to utilize context for contextual keywords

* git add .

* Add tests for extended escapes

* Better error courtesy of @DanielRossenwaser

* Add test of browser-inconsistent case and alter condition to match spec

* Merge adjacent conditions

* Use seperate functions for checking keywords vs not

* Use flags to track unicode escape presence

* Adjust error text
diff --git a/src/compiler/diagnosticMessages.json b/src/compiler/diagnosticMessages.json
@@ -851,6 +851,10 @@
         "category": "Error",
         "code": 1259
     },
+    "Keywords cannot contain escape characters.": {
+        "category": "Error",
+        "code": 1260
+    },
     "'with' statements are not allowed in an async function block.": {
         "category": "Error",
         "code": 1300
diff --git a/src/compiler/parser.ts b/src/compiler/parser.ts
@@ -1086,10 +1086,19 @@ namespace ts {
             return currentToken;
         }
 
-        function nextToken(): SyntaxKind {
+        function nextTokenWithoutCheck() {
             return currentToken = scanner.scan();
         }
 
+        function nextToken(): SyntaxKind {
+            // if the keyword had an escape
+            if (isKeyword(currentToken) && (scanner.hasUnicodeEscape() || scanner.hasExtendedUnicodeEscape())) {
+                // issue a parse error for the escape
+                parseErrorAt(scanner.getTokenPos(), scanner.getTextPos(), Diagnostics.Keywords_cannot_contain_escape_characters);
+            }
+            return nextTokenWithoutCheck();
+        }
+
         function nextTokenJSDoc(): JSDocSyntaxKind {
             return currentToken = scanner.scanJsDocToken();
         }
@@ -1380,7 +1389,7 @@ namespace ts {
                     node.originalKeywordKind = token();
                 }
                 node.escapedText = escapeLeadingUnderscores(internIdentifier(scanner.getTokenValue()));
-                nextToken();
+                nextTokenWithoutCheck();
                 return finishNode(node);
             }
 
diff --git a/src/compiler/scanner.ts b/src/compiler/scanner.ts
@@ -18,6 +18,7 @@ namespace ts {
         getTokenPos(): number;
         getTokenText(): string;
         getTokenValue(): string;
+        hasUnicodeEscape(): boolean;
         hasExtendedUnicodeEscape(): boolean;
         hasPrecedingLineBreak(): boolean;
         isIdentifier(): boolean;
@@ -884,6 +885,7 @@ namespace ts {
             getTokenPos: () => tokenPos,
             getTokenText: () => text.substring(tokenPos, pos),
             getTokenValue: () => tokenValue,
+            hasUnicodeEscape: () => (tokenFlags & TokenFlags.UnicodeEscape) !== 0,
             hasExtendedUnicodeEscape: () => (tokenFlags & TokenFlags.ExtendedUnicodeEscape) !== 0,
             hasPrecedingLineBreak: () => (tokenFlags & TokenFlags.PrecedingLineBreak) !== 0,
             isIdentifier: () => token === SyntaxKind.Identifier || token > SyntaxKind.LastReservedWord,
@@ -1245,6 +1247,7 @@ namespace ts {
                         return scanExtendedUnicodeEscape();
                     }
 
+                    tokenFlags |= TokenFlags.UnicodeEscape;
                     // '\uDDDD'
                     return scanHexadecimalEscape(/*numDigits*/ 4);
 
@@ -1376,6 +1379,7 @@ namespace ts {
                     if (!(ch >= 0 && isIdentifierPart(ch, languageVersion))) {
                         break;
                     }
+                    tokenFlags |= TokenFlags.UnicodeEscape;
                     result += text.substring(start, pos);
                     result += utf16EncodeAsString(ch);
                     // Valid Unicode escape is always six characters
@@ -1868,6 +1872,7 @@ namespace ts {
                         const cookedChar = peekUnicodeEscape();
                         if (cookedChar >= 0 && isIdentifierStart(cookedChar, languageVersion)) {
                             pos += 6;
+                            tokenFlags |= TokenFlags.UnicodeEscape;
                             tokenValue = String.fromCharCode(cookedChar) + scanIdentifierParts();
                             return token = getIdentifierToken();
                         }
@@ -2156,6 +2161,7 @@ namespace ts {
                     const cookedChar = peekUnicodeEscape();
                     if (cookedChar >= 0 && isIdentifierStart(cookedChar, languageVersion)) {
                         pos += 6;
+                        tokenFlags |= TokenFlags.UnicodeEscape;
                         tokenValue = String.fromCharCode(cookedChar) + scanIdentifierParts();
                         return token = getIdentifierToken();
                     }
diff --git a/src/compiler/types.ts b/src/compiler/types.ts
@@ -1679,6 +1679,8 @@ namespace ts {
         /* @internal */
         ContainsSeparator = 1 << 9, // e.g. `0b1100_0101`
         /* @internal */
+        UnicodeEscape = 1 << 10,
+        /* @internal */
         BinaryOrOctalSpecifier = BinarySpecifier | OctalSpecifier,
         /* @internal */
         NumericLiteralFlags = Scientific | Octal | HexSpecifier | BinaryOrOctalSpecifier | ContainsSeparator
diff --git a/src/compiler/utilities.ts b/src/compiler/utilities.ts
@@ -2682,6 +2682,10 @@ namespace ts {
         return isKeyword(token) && !isContextualKeyword(token);
     }
 
+    export function isFutureReservedKeyword(token: SyntaxKind): boolean {
+        return SyntaxKind.FirstFutureReservedWord <= token && token <= SyntaxKind.LastFutureReservedWord;
+    }
+
     export function isStringANonContextualKeyword(name: string) {
         const token = stringToToken(name);
         return token !== undefined && isNonContextualKeyword(token);
diff --git a/tests/baselines/reference/api/tsserverlibrary.d.ts b/tests/baselines/reference/api/tsserverlibrary.d.ts
@@ -3174,6 +3174,7 @@ declare namespace ts {
         getTokenPos(): number;
         getTokenText(): string;
         getTokenValue(): string;
+        hasUnicodeEscape(): boolean;
         hasExtendedUnicodeEscape(): boolean;
         hasPrecedingLineBreak(): boolean;
         isIdentifier(): boolean;
diff --git a/tests/baselines/reference/api/typescript.d.ts b/tests/baselines/reference/api/typescript.d.ts
@@ -3174,6 +3174,7 @@ declare namespace ts {
         getTokenPos(): number;
         getTokenText(): string;
         getTokenValue(): string;
+        hasUnicodeEscape(): boolean;
         hasExtendedUnicodeEscape(): boolean;
         hasPrecedingLineBreak(): boolean;
         isIdentifier(): boolean;
diff --git a/tests/baselines/reference/scannerUnicodeEscapeInKeyword1.errors.txt b/tests/baselines/reference/scannerUnicodeEscapeInKeyword1.errors.txt
@@ -0,0 +1,7 @@
+tests/cases/conformance/scanner/ecmascript5/scannerUnicodeEscapeInKeyword1.ts(1,1): error TS1260: Keywords cannot contain escape characters.
+
+
+==== tests/cases/conformance/scanner/ecmascript5/scannerUnicodeEscapeInKeyword1.ts (1 errors) ====
+    \u0076ar x = "hello";
+    ~~~~~~~~
+!!! error TS1260: Keywords cannot contain escape characters.
diff --git a/tests/baselines/reference/scannerUnicodeEscapeInKeyword2.errors.txt b/tests/baselines/reference/scannerUnicodeEscapeInKeyword2.errors.txt
@@ -0,0 +1,61 @@
+tests/cases/conformance/scanner/ecmascript5/file1.ts(3,5): error TS1260: Keywords cannot contain escape characters.
+tests/cases/conformance/scanner/ecmascript5/file1.ts(8,5): error TS1260: Keywords cannot contain escape characters.
+tests/cases/conformance/scanner/ecmascript5/file1.ts(13,1): error TS1260: Keywords cannot contain escape characters.
+tests/cases/conformance/scanner/ecmascript5/file2.ts(1,1): error TS1260: Keywords cannot contain escape characters.
+tests/cases/conformance/scanner/ecmascript5/file2.ts(5,5): error TS1260: Keywords cannot contain escape characters.
+tests/cases/conformance/scanner/ecmascript5/file2.ts(10,5): error TS1260: Keywords cannot contain escape characters.
+tests/cases/conformance/scanner/ecmascript5/file2.ts(15,1): error TS1260: Keywords cannot contain escape characters.
+
+
+==== tests/cases/conformance/scanner/ecmascript5/file1.ts (3 errors) ====
+    var \u0061wait = 12; // ok
+    async function main() {
+        \u0061wait 12; // not ok
+        ~~~~~~~~~~
+!!! error TS1260: Keywords cannot contain escape characters.
+    }
+    
+    var \u0079ield = 12; // ok
+    function *gen() {
+        \u0079ield 12; //not ok
+        ~~~~~~~~~~
+!!! error TS1260: Keywords cannot contain escape characters.
+    }
+    
+    type typ\u0065 = 12; // ok
+    
+    typ\u0065 notok = 0; // not ok
+    ~~~~~~~~~
+!!! error TS1260: Keywords cannot contain escape characters.
+    
+    export {};
+==== tests/cases/conformance/scanner/ecmascript5/file2.ts (4 errors) ====
+    \u{0076}ar x = "hello"; // not ok
+    ~~~~~~~~~~
+!!! error TS1260: Keywords cannot contain escape characters.
+    
+    var \u{0061}wait = 12; // ok
+    async function main() {
+        \u{0061}wait 12; // not ok
+        ~~~~~~~~~~~~
+!!! error TS1260: Keywords cannot contain escape characters.
+    }
+    
+    var \u{0079}ield = 12; // ok
+    function *gen() {
+        \u{0079}ield 12; //not ok
+        ~~~~~~~~~~~~
+!!! error TS1260: Keywords cannot contain escape characters.
+    }
+    
+    type typ\u{0065} = 12; // ok
+    
+    typ\u{0065} notok = 0; // not ok
+    ~~~~~~~~~~~
+!!! error TS1260: Keywords cannot contain escape characters.
+    
+    export {};
+    
+    const a = {def\u0061ult: 12}; // OK, `default` not in keyword position
+    // chrome and jsc may still error on this, ref https://bugs.chromium.org/p/chromium/issues/detail?id=993000 and https://bugs.webkit.org/show_bug.cgi?id=200638
+    
diff --git a/tests/baselines/reference/scannerUnicodeEscapeInKeyword2.js b/tests/baselines/reference/scannerUnicodeEscapeInKeyword2.js
@@ -0,0 +1,62 @@
+//// [tests/cases/conformance/scanner/ecmascript5/scannerUnicodeEscapeInKeyword2.ts] ////
+
+//// [file1.ts]
+var \u0061wait = 12; // ok
+async function main() {
+    \u0061wait 12; // not ok
+}
+
+var \u0079ield = 12; // ok
+function *gen() {
+    \u0079ield 12; //not ok
+}
+
+type typ\u0065 = 12; // ok
+
+typ\u0065 notok = 0; // not ok
+
+export {};
+//// [file2.ts]
+\u{0076}ar x = "hello"; // not ok
+
+var \u{0061}wait = 12; // ok
+async function main() {
+    \u{0061}wait 12; // not ok
+}
+
+var \u{0079}ield = 12; // ok
+function *gen() {
+    \u{0079}ield 12; //not ok
+}
+
+type typ\u{0065} = 12; // ok
+
+typ\u{0065} notok = 0; // not ok
+
+export {};
+
+const a = {def\u0061ult: 12}; // OK, `default` not in keyword position
+// chrome and jsc may still error on this, ref https://bugs.chromium.org/p/chromium/issues/detail?id=993000 and https://bugs.webkit.org/show_bug.cgi?id=200638
+
+
+//// [file1.js]
+var \u0061wait = 12; // ok
+async function main() {
+    await 12; // not ok
+}
+var \u0079ield = 12; // ok
+function* gen() {
+    yield 12; //not ok
+}
+//// [file2.js]
+var x = "hello"; // not ok
+var \u{0061}wait = 12; // ok
+async function main() {
+    await 12; // not ok
+}
+var \u{0079}ield = 12; // ok
+function* gen() {
+    yield 12; //not ok
+}
+const a = { def\u0061ult: 12 }; // OK, `default` not in keyword position
+// chrome and jsc may still error on this, ref https://bugs.chromium.org/p/chromium/issues/detail?id=993000 and https://bugs.webkit.org/show_bug.cgi?id=200638
diff --git a/tests/baselines/reference/scannerUnicodeEscapeInKeyword2.symbols b/tests/baselines/reference/scannerUnicodeEscapeInKeyword2.symbols
@@ -0,0 +1,62 @@
+=== tests/cases/conformance/scanner/ecmascript5/file1.ts ===
+var \u0061wait = 12; // ok
+>\u0061wait : Symbol(\u0061wait, Decl(file1.ts, 0, 3))
+
+async function main() {
+>main : Symbol(main, Decl(file1.ts, 0, 20))
+
+    \u0061wait 12; // not ok
+}
+
+var \u0079ield = 12; // ok
+>\u0079ield : Symbol(\u0079ield, Decl(file1.ts, 5, 3))
+
+function *gen() {
+>gen : Symbol(gen, Decl(file1.ts, 5, 20))
+
+    \u0079ield 12; //not ok
+}
+
+type typ\u0065 = 12; // ok
+>typ\u0065 : Symbol(typ\u0065, Decl(file1.ts, 8, 1))
+
+typ\u0065 notok = 0; // not ok
+>notok : Symbol(notok, Decl(file1.ts, 10, 20))
+
+export {};
+=== tests/cases/conformance/scanner/ecmascript5/file2.ts ===
+\u{0076}ar x = "hello"; // not ok
+>x : Symbol(x, Decl(file2.ts, 0, 10))
+
+var \u{0061}wait = 12; // ok
+>\u{0061}wait : Symbol(\u{0061}wait, Decl(file2.ts, 2, 3))
+
+async function main() {
+>main : Symbol(main, Decl(file2.ts, 2, 22))
+
+    \u{0061}wait 12; // not ok
+}
+
+var \u{0079}ield = 12; // ok
+>\u{0079}ield : Symbol(\u{0079}ield, Decl(file2.ts, 7, 3))
+
+function *gen() {
+>gen : Symbol(gen, Decl(file2.ts, 7, 22))
+
+    \u{0079}ield 12; //not ok
+}
+
+type typ\u{0065} = 12; // ok
+>typ\u{0065} : Symbol(typ\u{0065}, Decl(file2.ts, 10, 1))
+
+typ\u{0065} notok = 0; // not ok
+>notok : Symbol(notok, Decl(file2.ts, 12, 22))
+
+export {};
+
+const a = {def\u0061ult: 12}; // OK, `default` not in keyword position
+>a : Symbol(a, Decl(file2.ts, 18, 5))
+>def\u0061ult : Symbol(def\u0061ult, Decl(file2.ts, 18, 11))
+
+// chrome and jsc may still error on this, ref https://bugs.chromium.org/p/chromium/issues/detail?id=993000 and https://bugs.webkit.org/show_bug.cgi?id=200638
+
diff --git a/tests/baselines/reference/scannerUnicodeEscapeInKeyword2.types b/tests/baselines/reference/scannerUnicodeEscapeInKeyword2.types
diff --git a/tests/baselines/reference/switchStatementsWithMultipleDefaults.errors.txt b/tests/baselines/reference/switchStatementsWithMultipleDefaults.errors.txt
diff --git a/tests/cases/conformance/scanner/ecmascript5/scannerUnicodeEscapeInKeyword2.ts b/tests/cases/conformance/scanner/ecmascript5/scannerUnicodeEscapeInKeyword2.ts

Original file line number	Diff line number	Diff line change
`@@ -1086,10 +1086,19 @@ namespace ts {`
`1086`	`1086`	`return currentToken;`
`1087`	`1087`	`}`
`1088`	`1088`
`1089`		`- function nextToken(): SyntaxKind {`
	`1089`	`+ function nextTokenWithoutCheck() {`
`1090`	`1090`	`return currentToken = scanner.scan();`
`1091`	`1091`	`}`
`1092`	`1092`
	`1093`	`+ function nextToken(): SyntaxKind {`
	`1094`	`+ // if the keyword had an escape`
	`1095`	`+ if (isKeyword(currentToken) && (scanner.hasUnicodeEscape() \|\| scanner.hasExtendedUnicodeEscape())) {`
	`1096`	`+ // issue a parse error for the escape`
	`1097`	`+ parseErrorAt(scanner.getTokenPos(), scanner.getTextPos(), Diagnostics.Keywords_cannot_contain_escape_characters);`
	`1098`	`+ }`
	`1099`	`+ return nextTokenWithoutCheck();`
	`1100`	`+ }`
	`1101`	`+`
`1093`	`1102`	`function nextTokenJSDoc(): JSDocSyntaxKind {`
`1094`	`1103`	`return currentToken = scanner.scanJsDocToken();`
`1095`	`1104`	`}`
`@@ -1380,7 +1389,7 @@ namespace ts {`
`1380`	`1389`	`node.originalKeywordKind = token();`
`1381`	`1390`	`}`
`1382`	`1391`	`node.escapedText = escapeLeadingUnderscores(internIdentifier(scanner.getTokenValue()));`
`1383`		`- nextToken();`
	`1392`	`+ nextTokenWithoutCheck();`
`1384`	`1393`	`return finishNode(node);`
`1385`	`1394`	`}`
`1386`	`1395`