| Index: pkg/analyzer/lib/src/generated/parser.dart
|
| diff --git a/pkg/analyzer/lib/src/generated/parser.dart b/pkg/analyzer/lib/src/generated/parser.dart
|
| index d325fd1eaf3235f63bd193fcbce244d25bf3fcb6..4e99c1be378465678fd7bc9becaeec36cf2b20ec 100644
|
| --- a/pkg/analyzer/lib/src/generated/parser.dart
|
| +++ b/pkg/analyzer/lib/src/generated/parser.dart
|
| @@ -224,6 +224,10 @@ class IncrementalParseDispatcher implements ASTVisitor<ASTNode> {
|
|
|
| ASTNode visitAssignmentExpression(AssignmentExpression node) {
|
| if (identical(_oldNode, node.leftHandSide)) {
|
| + // TODO(brianwilkerson) If the assignment is part of a cascade section, then we don't have a
|
| + // single parse method that will work. Otherwise, we can parse a conditional expression, but
|
| + // need to ensure that the resulting expression is assignable.
|
| + // return parser.parseConditionalExpression();
|
| throw new InsufficientContextException();
|
| } else if (identical(_oldNode, node.rightHandSide)) {
|
| if (isCascadeAllowed(node)) {
|
| @@ -509,6 +513,7 @@ class IncrementalParseDispatcher implements ASTVisitor<ASTNode> {
|
| }
|
|
|
| ASTNode visitFormalParameterList(FormalParameterList node) {
|
| + // We don't know which kind of parameter to parse.
|
| throw new InsufficientContextException();
|
| }
|
|
|
| @@ -756,6 +761,7 @@ class IncrementalParseDispatcher implements ASTVisitor<ASTNode> {
|
| }
|
| return _parser.parseSimpleIdentifier();
|
| } else if (identical(_oldNode, node.body)) {
|
| + //return parser.parseFunctionBody();
|
| throw new InsufficientContextException();
|
| }
|
| return notAChild(node);
|
| @@ -1076,6 +1082,7 @@ class IncrementalParseDispatcher implements ASTVisitor<ASTNode> {
|
| * @return `true` if the right-hand side can be a cascade expression
|
| */
|
| bool isCascadeAllowed(AssignmentExpression node) {
|
| + // TODO(brianwilkerson) Implement this method.
|
| throw new InsufficientContextException();
|
| }
|
|
|
| @@ -1086,6 +1093,7 @@ class IncrementalParseDispatcher implements ASTVisitor<ASTNode> {
|
| * @return `true` if the expression can be a cascade expression
|
| */
|
| bool isCascadeAllowed2(ThrowExpression node) {
|
| + // TODO(brianwilkerson) Implement this method.
|
| throw new InsufficientContextException();
|
| }
|
|
|
| @@ -1187,20 +1195,33 @@ class IncrementalParser {
|
| ASTNode reparse(ASTNode originalStructure, Token leftToken, Token rightToken, int originalStart, int originalEnd) {
|
| ASTNode oldNode = null;
|
| ASTNode newNode = null;
|
| + //
|
| + // Find the first token that needs to be re-parsed.
|
| + //
|
| Token firstToken = leftToken.next;
|
| if (identical(firstToken, rightToken)) {
|
| + // If there are no new tokens, then we need to include at least one copied node in the range.
|
| firstToken = leftToken;
|
| }
|
| + //
|
| + // Find the smallest AST node that encompasses the range of re-scanned tokens.
|
| + //
|
| if (originalEnd < originalStart) {
|
| oldNode = new NodeLocator.con1(originalStart).searchWithin(originalStructure);
|
| } else {
|
| oldNode = new NodeLocator.con2(originalStart, originalEnd).searchWithin(originalStructure);
|
| }
|
| + //
|
| + // Find the token at which parsing is to begin.
|
| + //
|
| int originalOffset = oldNode.offset;
|
| Token parseToken = findTokenAt(firstToken, originalOffset);
|
| if (parseToken == null) {
|
| return null;
|
| }
|
| + //
|
| + // Parse the appropriate AST structure starting at the appropriate place.
|
| + //
|
| Parser parser = new Parser(_source, _errorListener);
|
| parser.currentToken = parseToken;
|
| while (newNode == null) {
|
| @@ -1214,6 +1235,9 @@ class IncrementalParser {
|
| try {
|
| IncrementalParseDispatcher dispatcher = new IncrementalParseDispatcher(parser, oldNode);
|
| newNode = parent.accept(dispatcher);
|
| + //
|
| + // Validate that the new node can replace the old node.
|
| + //
|
| Token mappedToken = _tokenMap.get(oldNode.endToken.next);
|
| if (mappedToken == null || mappedToken.offset != newNode.endToken.next.offset || newNode.offset != oldNode.offset) {
|
| advanceToParent = true;
|
| @@ -1232,7 +1256,11 @@ class IncrementalParser {
|
| }
|
| }
|
| _updatedNode = newNode;
|
| + //
|
| + // Replace the old node with the new node in a copy of the original AST structure.
|
| + //
|
| if (identical(oldNode, originalStructure)) {
|
| + // We ended up re-parsing the whole structure, so there's no need for a copy.
|
| ResolutionCopier.copyResolutionData(oldNode, newNode);
|
| return newNode;
|
| }
|
| @@ -1476,6 +1504,10 @@ class Parser {
|
| * @return the argument that was parsed
|
| */
|
| Expression parseArgument() {
|
| + //
|
| + // Both namedArgument and expression can start with an identifier, but only namedArgument can
|
| + // have an identifier followed by a colon.
|
| + //
|
| if (matchesIdentifier() && matches4(peek(), TokenType.COLON)) {
|
| return new NamedExpression(parseLabel(), parseExpression3());
|
| } else {
|
| @@ -1503,6 +1535,10 @@ class Parser {
|
| if (matches5(TokenType.CLOSE_PAREN)) {
|
| return new ArgumentList(leftParenthesis, arguments, andAdvance);
|
| }
|
| + //
|
| + // Even though unnamed arguments must all appear before any named arguments, we allow them to
|
| + // appear in any order so that we can recover faster.
|
| + //
|
| Expression argument = parseArgument();
|
| arguments.add(argument);
|
| bool foundNamedArgument = argument is NamedExpression;
|
| @@ -1512,6 +1548,7 @@ class Parser {
|
| arguments.add(argument);
|
| if (foundNamedArgument) {
|
| if (!generatedError && argument is! NamedExpression) {
|
| + // Report the error, once, but allow the arguments to be in any order in the AST.
|
| reportError11(ParserErrorCode.POSITIONAL_AFTER_NAMED_ARGUMENT, []);
|
| generatedError = true;
|
| }
|
| @@ -1519,6 +1556,9 @@ class Parser {
|
| foundNamedArgument = true;
|
| }
|
| }
|
| + // TODO(brianwilkerson) Recovery: Look at the left parenthesis to see whether there is a
|
| + // matching right parenthesis. If there is, then we're more likely missing a comma and should
|
| + // go back to parsing arguments.
|
| Token rightParenthesis = expect2(TokenType.CLOSE_PAREN);
|
| return new ArgumentList(leftParenthesis, arguments, rightParenthesis);
|
| }
|
| @@ -1568,6 +1608,7 @@ class Parser {
|
| statements.add(statement);
|
| }
|
| if (identical(_currentToken, statementStart)) {
|
| + // Ensure that we are making progress and report an error if we're not.
|
| reportError12(ParserErrorCode.UNEXPECTED_TOKEN, _currentToken, [_currentToken.lexeme]);
|
| advance();
|
| }
|
| @@ -1611,13 +1652,22 @@ class Parser {
|
| validateModifiersForGetterOrSetterOrMethod(modifiers);
|
| return parseMethodDeclaration(commentAndMetadata, modifiers.externalKeyword, modifiers.staticKeyword, returnType);
|
| } else {
|
| + //
|
| + // We have found an error of some kind. Try to recover.
|
| + //
|
| if (matchesIdentifier()) {
|
| if (matchesAny(peek(), [TokenType.EQ, TokenType.COMMA, TokenType.SEMICOLON])) {
|
| + //
|
| + // We appear to have a variable declaration with a type of "void".
|
| + //
|
| reportError10(ParserErrorCode.VOID_VARIABLE, returnType, []);
|
| return parseInitializedIdentifierList(commentAndMetadata, modifiers.staticKeyword, validateModifiersForField(modifiers), returnType);
|
| }
|
| }
|
| if (isOperator(_currentToken)) {
|
| + //
|
| + // We appear to have found an operator declaration without the 'operator' keyword.
|
| + //
|
| validateModifiersForOperator(modifiers);
|
| return parseOperator(commentAndMetadata, modifiers.externalKeyword, returnType);
|
| }
|
| @@ -1635,6 +1685,9 @@ class Parser {
|
| return parseOperator(commentAndMetadata, modifiers.externalKeyword, null);
|
| } else if (!matchesIdentifier()) {
|
| if (isOperator(_currentToken)) {
|
| + //
|
| + // We appear to have found an operator declaration without the 'operator' keyword.
|
| + //
|
| validateModifiersForOperator(modifiers);
|
| return parseOperator(commentAndMetadata, modifiers.externalKeyword, null);
|
| }
|
| @@ -1669,12 +1722,25 @@ class Parser {
|
| return parseOperator(commentAndMetadata, modifiers.externalKeyword, type);
|
| } else if (!matchesIdentifier()) {
|
| if (matches5(TokenType.CLOSE_CURLY_BRACKET)) {
|
| + //
|
| + // We appear to have found an incomplete declaration at the end of the class. At this point
|
| + // it consists of a type name, so we'll treat it as a field declaration with a missing
|
| + // field name and semicolon.
|
| + //
|
| return parseInitializedIdentifierList(commentAndMetadata, modifiers.staticKeyword, validateModifiersForField(modifiers), type);
|
| }
|
| if (isOperator(_currentToken)) {
|
| + //
|
| + // We appear to have found an operator declaration without the 'operator' keyword.
|
| + //
|
| validateModifiersForOperator(modifiers);
|
| return parseOperator(commentAndMetadata, modifiers.externalKeyword, type);
|
| }
|
| + //
|
| + // We appear to have found an incomplete declaration before another declaration.
|
| + // At this point it consists of a type name, so we'll treat it as a field declaration
|
| + // with a missing field name and semicolon.
|
| + //
|
| reportError12(ParserErrorCode.EXPECTED_CLASS_MEMBER, _currentToken, []);
|
| try {
|
| lockErrorListener();
|
| @@ -1724,6 +1790,10 @@ class Parser {
|
| if (matches5(TokenType.SCRIPT_TAG)) {
|
| scriptTag = new ScriptTag(andAdvance);
|
| }
|
| + //
|
| + // Even though all directives must appear before declarations and must occur in a given order,
|
| + // we allow directives and declarations to occur in any order so that we can recover better.
|
| + //
|
| bool libraryDirectiveFound = false;
|
| bool partOfDirectiveFound = false;
|
| bool partDirectiveFound = false;
|
| @@ -1855,6 +1925,11 @@ class Parser {
|
| } else if (matches(Keyword.RETHROW)) {
|
| return parseRethrowExpression();
|
| }
|
| + //
|
| + // assignableExpression is a subset of conditionalExpression, so we can parse a conditional
|
| + // expression and then determine whether it is followed by an assignmentOperator, checking for
|
| + // conformance to the restricted grammar after making that determination.
|
| + //
|
| Expression expression = parseConditionalExpression();
|
| TokenType tokenType = _currentToken.type;
|
| if (identical(tokenType, TokenType.PERIOD_PERIOD)) {
|
| @@ -1893,6 +1968,11 @@ class Parser {
|
| } else if (matches(Keyword.RETHROW)) {
|
| return parseRethrowExpression();
|
| }
|
| + //
|
| + // assignableExpression is a subset of conditionalExpression, so we can parse a conditional
|
| + // expression and then determine whether it is followed by an assignmentOperator, checking for
|
| + // conformance to the restricted grammar after making that determination.
|
| + //
|
| Expression expression = parseConditionalExpression();
|
| if (_currentToken.type.isAssignmentOperator) {
|
| Token operator = andAdvance;
|
| @@ -1948,6 +2028,11 @@ class Parser {
|
| if (matches5(TokenType.CLOSE_PAREN)) {
|
| return new FormalParameterList(leftParenthesis, null, null, null, andAdvance);
|
| }
|
| + //
|
| + // Even though it is invalid to have default parameters outside of brackets, required parameters
|
| + // inside of brackets, or multiple groups of default and named parameters, we allow all of these
|
| + // cases so that we can recover better.
|
| + //
|
| List<FormalParameter> parameters = new List<FormalParameter>();
|
| List<FormalParameter> normalParameters = new List<FormalParameter>();
|
| List<FormalParameter> positionalParameters = new List<FormalParameter>();
|
| @@ -1968,6 +2053,7 @@ class Parser {
|
| if (firstParameter) {
|
| firstParameter = false;
|
| } else if (!optional(TokenType.COMMA)) {
|
| + // TODO(brianwilkerson) The token is wrong, we need to recover from this case.
|
| if (getEndToken(leftParenthesis) != null) {
|
| reportError11(ParserErrorCode.EXPECTED_TOKEN, [TokenType.COMMA.lexeme]);
|
| } else {
|
| @@ -1976,6 +2062,9 @@ class Parser {
|
| }
|
| }
|
| initialToken = _currentToken;
|
| + //
|
| + // Handle the beginning of parameter groups.
|
| + //
|
| if (matches5(TokenType.OPEN_SQUARE_BRACKET)) {
|
| wasOptionalParameter = true;
|
| if (leftSquareBracket != null && !reportedMuliplePositionalGroups) {
|
| @@ -2003,12 +2092,19 @@ class Parser {
|
| currentParameters = namedParameters;
|
| kind = ParameterKind.NAMED;
|
| }
|
| + //
|
| + // Parse and record the parameter.
|
| + //
|
| FormalParameter parameter = parseFormalParameter(kind);
|
| parameters.add(parameter);
|
| currentParameters.add(parameter);
|
| if (identical(kind, ParameterKind.REQUIRED) && wasOptionalParameter) {
|
| reportError10(ParserErrorCode.NORMAL_BEFORE_OPTIONAL_PARAMETERS, parameter, []);
|
| }
|
| + //
|
| + // Handle the end of parameter groups.
|
| + //
|
| + // TODO(brianwilkerson) Improve the detection and reporting of missing and mismatched delimiters.
|
| if (matches5(TokenType.CLOSE_SQUARE_BRACKET)) {
|
| rightSquareBracket = andAdvance;
|
| currentParameters = normalParameters;
|
| @@ -2038,12 +2134,18 @@ class Parser {
|
| }
|
| } while (!matches5(TokenType.CLOSE_PAREN) && initialToken != _currentToken);
|
| Token rightParenthesis = expect2(TokenType.CLOSE_PAREN);
|
| + //
|
| + // Check that the groups were closed correctly.
|
| + //
|
| if (leftSquareBracket != null && rightSquareBracket == null) {
|
| reportError11(ParserErrorCode.MISSING_TERMINATOR_FOR_PARAMETER_GROUP, ["]"]);
|
| }
|
| if (leftCurlyBracket != null && rightCurlyBracket == null) {
|
| reportError11(ParserErrorCode.MISSING_TERMINATOR_FOR_PARAMETER_GROUP, ["}"]);
|
| }
|
| + //
|
| + // Build the parameter list.
|
| + //
|
| if (leftSquareBracket == null) {
|
| leftSquareBracket = leftCurlyBracket;
|
| }
|
| @@ -2542,15 +2644,20 @@ class Parser {
|
| */
|
| bool couldBeStartOfCompilationUnitMember() {
|
| if ((matches(Keyword.IMPORT) || matches(Keyword.EXPORT) || matches(Keyword.LIBRARY) || matches(Keyword.PART)) && !matches4(peek(), TokenType.PERIOD) && !matches4(peek(), TokenType.LT)) {
|
| + // This looks like the start of a directive
|
| return true;
|
| } else if (matches(Keyword.CLASS)) {
|
| + // This looks like the start of a class definition
|
| return true;
|
| } else if (matches(Keyword.TYPEDEF) && !matches4(peek(), TokenType.PERIOD) && !matches4(peek(), TokenType.LT)) {
|
| + // This looks like the start of a typedef
|
| return true;
|
| } else if (matches(Keyword.VOID) || ((matches(Keyword.GET) || matches(Keyword.SET)) && matchesIdentifier2(peek())) || (matches(Keyword.OPERATOR) && isOperator(peek()))) {
|
| + // This looks like the start of a function
|
| return true;
|
| } else if (matchesIdentifier()) {
|
| if (matches4(peek(), TokenType.OPEN_PAREN)) {
|
| + // This looks like the start of a function
|
| return true;
|
| }
|
| Token token = skipReturnType(_currentToken);
|
| @@ -2572,6 +2679,10 @@ class Parser {
|
| SimpleIdentifier createSyntheticIdentifier() {
|
| Token syntheticToken;
|
| if (identical(_currentToken.type, TokenType.KEYWORD)) {
|
| + // Consider current keyword token as an identifier.
|
| + // It is not always true, e.g. "^is T" where "^" is place the place for synthetic identifier.
|
| + // By creating SyntheticStringToken we can distinguish a real identifier from synthetic.
|
| + // In the code completion behavior will depend on a cursor position - before or on "is".
|
| syntheticToken = injectToken(new SyntheticStringToken(TokenType.IDENTIFIER, _currentToken.lexeme, _currentToken.offset));
|
| } else {
|
| syntheticToken = createSyntheticToken2(TokenType.IDENTIFIER);
|
| @@ -2633,6 +2744,8 @@ class Parser {
|
| if (matches(keyword)) {
|
| return andAdvance;
|
| }
|
| + // Remove uses of this method in favor of matches?
|
| + // Pass in the error code to use to report the error?
|
| reportError11(ParserErrorCode.EXPECTED_TOKEN, [keyword.syntax]);
|
| return _currentToken;
|
| }
|
| @@ -2648,6 +2761,8 @@ class Parser {
|
| if (matches5(type)) {
|
| return andAdvance;
|
| }
|
| + // Remove uses of this method in favor of matches?
|
| + // Pass in the error code to use to report the error?
|
| if (identical(type, TokenType.SEMICOLON)) {
|
| reportError12(ParserErrorCode.EXPECTED_TOKEN, _currentToken.previous, [type.lexeme]);
|
| } else {
|
| @@ -2786,10 +2901,13 @@ class Parser {
|
| }
|
| Token afterReturnType = skipTypeName(_currentToken);
|
| if (afterReturnType == null) {
|
| + // There was no return type, but it is optional, so go back to where we started.
|
| afterReturnType = _currentToken;
|
| }
|
| Token afterIdentifier = skipSimpleIdentifier(afterReturnType);
|
| if (afterIdentifier == null) {
|
| + // It's possible that we parsed the function name as if it were a type name, so see whether
|
| + // it makes sense if we assume that there is no type.
|
| afterIdentifier = skipSimpleIdentifier(_currentToken);
|
| }
|
| if (afterIdentifier == null) {
|
| @@ -2798,6 +2916,8 @@ class Parser {
|
| if (isFunctionExpression(afterIdentifier)) {
|
| return true;
|
| }
|
| + // It's possible that we have found a getter. While this isn't valid at this point we test for
|
| + // it in order to recover better.
|
| if (matches(Keyword.GET)) {
|
| Token afterName = skipSimpleIdentifier(_currentToken.next);
|
| if (afterName == null) {
|
| @@ -2860,17 +2980,22 @@ class Parser {
|
| */
|
| bool isInitializedVariableDeclaration() {
|
| if (matches(Keyword.FINAL) || matches(Keyword.VAR)) {
|
| + // An expression cannot start with a keyword other than 'const', 'rethrow', or 'throw'.
|
| return true;
|
| }
|
| if (matches(Keyword.CONST)) {
|
| + // Look to see whether we might be at the start of a list or map literal, otherwise this
|
| + // should be the start of a variable declaration.
|
| return !matchesAny(peek(), [
|
| TokenType.LT,
|
| TokenType.OPEN_CURLY_BRACKET,
|
| TokenType.OPEN_SQUARE_BRACKET,
|
| TokenType.INDEX]);
|
| }
|
| + // We know that we have an identifier, and need to see whether it might be a type name.
|
| Token token = skipTypeName(_currentToken);
|
| if (token == null) {
|
| + // There was no type name, so this can't be a declaration.
|
| return false;
|
| }
|
| token = skipSimpleIdentifier(token);
|
| @@ -2920,16 +3045,20 @@ class Parser {
|
| * @return `true` if the given token appears to be the beginning of an operator declaration
|
| */
|
| bool isOperator(Token startToken) {
|
| + // Accept any operator here, even if it is not user definable.
|
| if (!startToken.isOperator) {
|
| return false;
|
| }
|
| + // Token "=" means that it is actually field initializer.
|
| if (identical(startToken.type, TokenType.EQ)) {
|
| return false;
|
| }
|
| + // Consume all operator tokens.
|
| Token token = startToken.next;
|
| while (token.isOperator) {
|
| token = token.next;
|
| }
|
| + // Formal parameter list is expect now.
|
| return matches4(token, TokenType.OPEN_PAREN);
|
| }
|
|
|
| @@ -3220,6 +3349,11 @@ class Parser {
|
| if (matches(Keyword.SUPER)) {
|
| return parseAssignableSelector(new SuperExpression(andAdvance), false);
|
| }
|
| + //
|
| + // A primary expression can start with an identifier. We resolve the ambiguity by determining
|
| + // whether the primary consists of anything other than an identifier and/or is followed by an
|
| + // assignableSelector.
|
| + //
|
| Expression expression = parsePrimaryExpression();
|
| bool isOptional = primaryAllowed || expression is SimpleIdentifier;
|
| while (true) {
|
| @@ -3277,6 +3411,7 @@ class Parser {
|
| return new PropertyAccess(prefix, period, parseSimpleIdentifier());
|
| } else {
|
| if (!optional) {
|
| + // Report the missing selector.
|
| reportError11(ParserErrorCode.MISSING_ASSIGNABLE_SELECTOR, []);
|
| }
|
| return prefix;
|
| @@ -3396,6 +3531,7 @@ class Parser {
|
| period = null;
|
| functionName = null;
|
| } else if (expression == null) {
|
| + // It should not be possible to get here.
|
| expression = new MethodInvocation(expression, period, createSyntheticIdentifier(), parseArgumentList());
|
| } else {
|
| expression = new FunctionExpressionInvocation(expression, parseArgumentList());
|
| @@ -3463,6 +3599,10 @@ class Parser {
|
| if (matches5(TokenType.LT)) {
|
| typeParameters = parseTypeParameterList();
|
| }
|
| + //
|
| + // Parse the clauses. The parser accepts clauses in any order, but will generate errors if they
|
| + // are not in the order required by the specification.
|
| + //
|
| ExtendsClause extendsClause = null;
|
| WithClause withClause = null;
|
| ImplementsClause implementsClause = null;
|
| @@ -3504,10 +3644,16 @@ class Parser {
|
| if (withClause != null && extendsClause == null) {
|
| reportError12(ParserErrorCode.WITH_WITHOUT_EXTENDS, withClause.withKeyword, []);
|
| }
|
| + //
|
| + // Look for and skip over the extra-lingual 'native' specification.
|
| + //
|
| NativeClause nativeClause = null;
|
| if (matches2(_NATIVE) && matches4(peek(), TokenType.STRING)) {
|
| nativeClause = parseNativeClause();
|
| }
|
| + //
|
| + // Parse the body of the class.
|
| + //
|
| Token leftBracket = null;
|
| List<ClassMember> members = null;
|
| Token rightBracket = null;
|
| @@ -3677,6 +3823,7 @@ class Parser {
|
| * @return the comment reference that was parsed, or `null` if no reference could be found
|
| */
|
| CommentReference parseCommentReference(String referenceSource, int sourceOffset) {
|
| + // TODO(brianwilkerson) The errors are not getting the right offset/length and are being duplicated.
|
| if (referenceSource.length == 0) {
|
| return null;
|
| }
|
| @@ -3710,6 +3857,10 @@ class Parser {
|
| }
|
| return new CommentReference(newKeyword, identifier);
|
| } else if (matches3(firstToken, Keyword.THIS) || matches3(firstToken, Keyword.NULL) || matches3(firstToken, Keyword.TRUE) || matches3(firstToken, Keyword.FALSE)) {
|
| + // TODO(brianwilkerson) If we want to support this we will need to extend the definition
|
| + // of CommentReference to take an expression rather than an identifier. For now we just
|
| + // ignore it to reduce the number of errors produced, but that's probably not a valid
|
| + // long term approach.
|
| return null;
|
| }
|
| } on JavaException catch (exception) {
|
| @@ -3808,8 +3959,14 @@ class Parser {
|
| validateModifiersForTopLevelFunction(modifiers);
|
| return parseFunctionDeclaration(commentAndMetadata, modifiers.externalKeyword, returnType);
|
| } else {
|
| + //
|
| + // We have found an error of some kind. Try to recover.
|
| + //
|
| if (matchesIdentifier()) {
|
| if (matchesAny(peek(), [TokenType.EQ, TokenType.COMMA, TokenType.SEMICOLON])) {
|
| + //
|
| + // We appear to have a variable declaration with a type of "void".
|
| + //
|
| reportError10(ParserErrorCode.VOID_VARIABLE, returnType, []);
|
| return new TopLevelVariableDeclaration(commentAndMetadata.comment, commentAndMetadata.metadata, parseVariableDeclarationList2(null, validateModifiersForTopLevelVariable(modifiers), null), expect2(TokenType.SEMICOLON));
|
| }
|
| @@ -3845,6 +4002,7 @@ class Parser {
|
| } else if (matches5(TokenType.AT)) {
|
| return new TopLevelVariableDeclaration(commentAndMetadata.comment, commentAndMetadata.metadata, parseVariableDeclarationList2(null, validateModifiersForTopLevelVariable(modifiers), returnType), expect2(TokenType.SEMICOLON));
|
| } else if (!matchesIdentifier()) {
|
| + // TODO(brianwilkerson) Generalize this error. We could also be parsing a top-level variable at this point.
|
| reportError12(ParserErrorCode.EXPECTED_EXECUTABLE, _currentToken, []);
|
| Token semicolon;
|
| if (matches5(TokenType.SEMICOLON)) {
|
| @@ -4028,6 +4186,8 @@ class Parser {
|
| } else if (matches(Keyword.PART)) {
|
| return parsePartDirective(commentAndMetadata);
|
| } else {
|
| + // Internal error: this method should not have been invoked if the current token was something
|
| + // other than one of the above.
|
| throw new IllegalStateException("parseDirective invoked in an invalid state; currentToken = ${_currentToken}");
|
| }
|
| }
|
| @@ -4296,6 +4456,7 @@ class Parser {
|
| DeclaredIdentifier loopVariable = null;
|
| SimpleIdentifier identifier = null;
|
| if (variableList == null) {
|
| + // We found: <expression> 'in'
|
| reportError11(ParserErrorCode.MISSING_VARIABLE_IN_FOR_EACH, []);
|
| } else {
|
| NodeList<VariableDeclaration> variables = variableList.variables;
|
| @@ -4399,6 +4560,7 @@ class Parser {
|
| }
|
| return new NativeFunctionBody(nativeToken, stringLiteral, expect2(TokenType.SEMICOLON));
|
| } else {
|
| + // Invalid function body
|
| reportError11(emptyErrorCode, []);
|
| return new EmptyFunctionBody(createSyntheticToken2(TokenType.SEMICOLON));
|
| }
|
| @@ -4452,6 +4614,11 @@ class Parser {
|
| } else {
|
| body = new EmptyFunctionBody(expect2(TokenType.SEMICOLON));
|
| }
|
| + // if (!isStatement && matches(TokenType.SEMICOLON)) {
|
| + // // TODO(brianwilkerson) Improve this error message.
|
| + // reportError(ParserErrorCode.UNEXPECTED_TOKEN, currentToken.getLexeme());
|
| + // advance();
|
| + // }
|
| return new FunctionDeclaration(commentAndMetadata.comment, commentAndMetadata.metadata, externalKeyword, returnType, keyword, name, new FunctionExpression(parameters, body));
|
| }
|
|
|
| @@ -4529,6 +4696,9 @@ class Parser {
|
| return new FunctionTypeAlias(commentAndMetadata.comment, commentAndMetadata.metadata, keyword, returnType, name, typeParameters, parameters, semicolon);
|
| } else if (!matches5(TokenType.OPEN_PAREN)) {
|
| reportError11(ParserErrorCode.MISSING_TYPEDEF_PARAMETERS, []);
|
| + // TODO(brianwilkerson) Recover from this error. At the very least we should skip to the start
|
| + // of the next valid compilation unit member, allowing for the possibility of finding the
|
| + // typedef parameters before that point.
|
| return new FunctionTypeAlias(commentAndMetadata.comment, commentAndMetadata.metadata, keyword, returnType, name, typeParameters, new FormalParameterList(createSyntheticToken2(TokenType.OPEN_PAREN), null, null, null, createSyntheticToken2(TokenType.CLOSE_PAREN)), createSyntheticToken2(TokenType.SEMICOLON));
|
| }
|
| FormalParameterList parameters = parseFormalParameterList();
|
| @@ -4721,6 +4891,8 @@ class Parser {
|
| if (matchesIdentifier()) {
|
| return parseLibraryIdentifier();
|
| } else if (matches5(TokenType.STRING)) {
|
| + // TODO(brianwilkerson) Recovery: This should be extended to handle arbitrary tokens until we
|
| + // can find a token that can start a compilation unit member.
|
| StringLiteral string = parseStringLiteral();
|
| reportError10(ParserErrorCode.NON_IDENTIFIER_LIBRARY_NAME, string, []);
|
| } else {
|
| @@ -4746,6 +4918,7 @@ class Parser {
|
| * @return the list literal that was parsed
|
| */
|
| ListLiteral parseListLiteral(Token modifier, TypeArgumentList typeArguments) {
|
| + // may be empty list literal
|
| if (matches5(TokenType.INDEX)) {
|
| BeginToken leftBracket = new BeginToken(TokenType.OPEN_SQUARE_BRACKET, _currentToken.offset);
|
| Token rightBracket = new Token(TokenType.CLOSE_SQUARE_BRACKET, _currentToken.offset + 1);
|
| @@ -4756,6 +4929,7 @@ class Parser {
|
| _currentToken = _currentToken.next;
|
| return new ListLiteral(modifier, typeArguments, leftBracket, null, rightBracket);
|
| }
|
| + // open
|
| Token leftBracket = expect2(TokenType.OPEN_SQUARE_BRACKET);
|
| if (matches5(TokenType.CLOSE_SQUARE_BRACKET)) {
|
| return new ListLiteral(modifier, typeArguments, leftBracket, null, andAdvance);
|
| @@ -5057,6 +5231,7 @@ class Parser {
|
| * @return the non-labeled statement that was parsed
|
| */
|
| Statement parseNonLabeledStatement() {
|
| + // TODO(brianwilkerson) Pass the comment and metadata on where appropriate.
|
| CommentAndMetadata commentAndMetadata = parseCommentAndMetadata();
|
| if (matches5(TokenType.OPEN_CURLY_BRACKET)) {
|
| if (matches4(peek(), TokenType.STRING)) {
|
| @@ -5068,6 +5243,7 @@ class Parser {
|
| return parseBlock();
|
| } else if (matches5(TokenType.KEYWORD) && !(_currentToken as KeywordToken).keyword.isPseudoKeyword) {
|
| Keyword keyword = (_currentToken as KeywordToken).keyword;
|
| + // TODO(jwren) compute some metrics to figure out a better order for this if-then sequence to optimize performance
|
| if (identical(keyword, Keyword.ASSERT)) {
|
| return parseAssertStatement();
|
| } else if (identical(keyword, Keyword.BREAK)) {
|
| @@ -5102,15 +5278,26 @@ class Parser {
|
| TokenType.FUNCTION])) {
|
| return parseFunctionDeclarationStatement2(commentAndMetadata, returnType);
|
| } else {
|
| + //
|
| + // We have found an error of some kind. Try to recover.
|
| + //
|
| if (matchesIdentifier()) {
|
| if (matchesAny(peek(), [TokenType.EQ, TokenType.COMMA, TokenType.SEMICOLON])) {
|
| + //
|
| + // We appear to have a variable declaration with a type of "void".
|
| + //
|
| reportError10(ParserErrorCode.VOID_VARIABLE, returnType, []);
|
| return parseVariableDeclarationStatement(commentAndMetadata);
|
| }
|
| } else if (matches5(TokenType.CLOSE_CURLY_BRACKET)) {
|
| + //
|
| + // We appear to have found an incomplete statement at the end of a block. Parse it as a
|
| + // variable declaration.
|
| + //
|
| return parseVariableDeclarationStatement2(commentAndMetadata, null, returnType);
|
| }
|
| reportError11(ParserErrorCode.MISSING_STATEMENT, []);
|
| + // TODO(brianwilkerson) Recover from this error.
|
| return new EmptyStatement(createSyntheticToken2(TokenType.SEMICOLON));
|
| }
|
| } else if (identical(keyword, Keyword.CONST)) {
|
| @@ -5132,6 +5319,9 @@ class Parser {
|
| } else if (identical(keyword, Keyword.NEW) || identical(keyword, Keyword.TRUE) || identical(keyword, Keyword.FALSE) || identical(keyword, Keyword.NULL) || identical(keyword, Keyword.SUPER) || identical(keyword, Keyword.THIS)) {
|
| return new ExpressionStatement(parseExpression3(), expect2(TokenType.SEMICOLON));
|
| } else {
|
| + //
|
| + // We have found an error of some kind. Try to recover.
|
| + //
|
| reportError11(ParserErrorCode.MISSING_STATEMENT, []);
|
| return new EmptyStatement(createSyntheticToken2(TokenType.SEMICOLON));
|
| }
|
| @@ -5350,6 +5540,16 @@ class Parser {
|
| } else if (matches5(TokenType.OPEN_SQUARE_BRACKET) || matches5(TokenType.INDEX)) {
|
| return parseListLiteral(null, null);
|
| } else if (matchesIdentifier()) {
|
| + // TODO(brianwilkerson) The code below was an attempt to recover from an error case, but it
|
| + // needs to be applied as a recovery only after we know that parsing it as an identifier
|
| + // doesn't work. Leaving the code as a reminder of how to recover.
|
| + // if (isFunctionExpression(peek())) {
|
| + // //
|
| + // // Function expressions were allowed to have names at one point, but this is now illegal.
|
| + // //
|
| + // reportError(ParserErrorCode.NAMED_FUNCTION_EXPRESSION, getAndAdvance());
|
| + // return parseFunctionExpression();
|
| + // }
|
| return parsePrefixedIdentifier();
|
| } else if (matches(Keyword.NEW)) {
|
| return parseNewExpression();
|
| @@ -5368,6 +5568,10 @@ class Parser {
|
| } else if (matches5(TokenType.QUESTION)) {
|
| return parseArgumentDefinitionTest();
|
| } else if (matches(Keyword.VOID)) {
|
| + //
|
| + // Recover from having a return type of "void" where a return type is not expected.
|
| + //
|
| + // TODO(brianwilkerson) Improve this error message.
|
| reportError11(ParserErrorCode.UNEXPECTED_TOKEN, [_currentToken.lexeme]);
|
| advance();
|
| return parsePrimaryExpression();
|
| @@ -5662,6 +5866,8 @@ class Parser {
|
| Token colon = expect2(TokenType.COLON);
|
| members.add(new SwitchDefault(labels, defaultKeyword, colon, parseStatements2()));
|
| } else {
|
| + // We need to advance, otherwise we could end up in an infinite loop, but this could be a
|
| + // lot smarter about recovering from the error.
|
| reportError11(ParserErrorCode.EXPECTED_CASE_OR_DEFAULT, []);
|
| while (!matches5(TokenType.EOF) && !matches5(TokenType.CLOSE_CURLY_BRACKET) && !matches(Keyword.CASE) && !matches(Keyword.DEFAULT)) {
|
| advance();
|
| @@ -5878,6 +6084,10 @@ class Parser {
|
| Token operator = andAdvance;
|
| if (matches(Keyword.SUPER)) {
|
| if (matches4(peek(), TokenType.OPEN_SQUARE_BRACKET) || matches4(peek(), TokenType.PERIOD)) {
|
| + // "prefixOperator unaryExpression"
|
| + // --> "prefixOperator postfixExpression"
|
| + // --> "prefixOperator primary selector*"
|
| + // --> "prefixOperator 'super' assignableSelector selector*"
|
| return new PrefixExpression(operator, parseUnaryExpression());
|
| }
|
| return new PrefixExpression(operator, new SuperExpression(andAdvance));
|
| @@ -5887,8 +6097,14 @@ class Parser {
|
| Token operator = andAdvance;
|
| if (matches(Keyword.SUPER)) {
|
| if (matches4(peek(), TokenType.OPEN_SQUARE_BRACKET) || matches4(peek(), TokenType.PERIOD)) {
|
| + // --> "prefixOperator 'super' assignableSelector selector*"
|
| return new PrefixExpression(operator, parseUnaryExpression());
|
| }
|
| + //
|
| + // Even though it is not valid to use an incrementing operator ('++' or '--') before 'super',
|
| + // we can (and therefore must) interpret "--super" as semantically equivalent to "-(-super)".
|
| + // Unfortunately, we cannot do the same for "++super" because "+super" is also not valid.
|
| + //
|
| if (identical(operator.type, TokenType.MINUS_MINUS)) {
|
| int offset = operator.offset;
|
| Token firstOperator = new Token(TokenType.MINUS, offset);
|
| @@ -5898,6 +6114,7 @@ class Parser {
|
| operator.previous.setNext(firstOperator);
|
| return new PrefixExpression(firstOperator, new PrefixExpression(secondOperator, new SuperExpression(andAdvance)));
|
| } else {
|
| + // Invalid operator before 'super'
|
| reportError11(ParserErrorCode.INVALID_OPERATOR_FOR_SUPER, [operator.lexeme]);
|
| return new PrefixExpression(operator, new SuperExpression(andAdvance));
|
| }
|
| @@ -5989,7 +6206,14 @@ class Parser {
|
| * @return the variable declaration statement that was parsed
|
| */
|
| VariableDeclarationStatement parseVariableDeclarationStatement(CommentAndMetadata commentAndMetadata) {
|
| + // Token startToken = currentToken;
|
| VariableDeclarationList variableList = parseVariableDeclarationList(commentAndMetadata);
|
| + // if (!matches(TokenType.SEMICOLON)) {
|
| + // if (matches(startToken, Keyword.VAR) && isTypedIdentifier(startToken.getNext())) {
|
| + // // TODO(brianwilkerson) This appears to be of the form "var type variable". We should do
|
| + // // a better job of recovering in this case.
|
| + // }
|
| + // }
|
| Token semicolon = expect2(TokenType.SEMICOLON);
|
| return new VariableDeclarationStatement(variableList, semicolon);
|
| }
|
| @@ -6136,9 +6360,11 @@ class Parser {
|
| Token next = startToken.next;
|
| if (matchesIdentifier2(next)) {
|
| Token next2 = next.next;
|
| + // "Type parameter" or "Type<" or "prefix.Type"
|
| if (matchesIdentifier2(next2) || matches4(next2, TokenType.LT) || matches4(next2, TokenType.PERIOD)) {
|
| return skipTypeName(next);
|
| }
|
| + // "parameter"
|
| return next;
|
| }
|
| } else if (matches3(startToken, Keyword.VAR)) {
|
| @@ -6197,18 +6423,29 @@ class Parser {
|
| if (matches4(next, TokenType.CLOSE_PAREN)) {
|
| return next.next;
|
| }
|
| + //
|
| + // Look to see whether the token after the open parenthesis is something that should only occur
|
| + // at the beginning of a parameter list.
|
| + //
|
| if (matchesAny(next, [
|
| TokenType.AT,
|
| TokenType.OPEN_SQUARE_BRACKET,
|
| TokenType.OPEN_CURLY_BRACKET]) || matches3(next, Keyword.VOID) || (matchesIdentifier2(next) && (matchesAny(next.next, [TokenType.COMMA, TokenType.CLOSE_PAREN])))) {
|
| return skipPastMatchingToken(startToken);
|
| }
|
| + //
|
| + // Look to see whether the first parameter is a function typed parameter without a return type.
|
| + //
|
| if (matchesIdentifier2(next) && matches4(next.next, TokenType.OPEN_PAREN)) {
|
| Token afterParameters = skipFormalParameterList(next.next);
|
| if (afterParameters != null && (matchesAny(afterParameters, [TokenType.COMMA, TokenType.CLOSE_PAREN]))) {
|
| return skipPastMatchingToken(startToken);
|
| }
|
| }
|
| + //
|
| + // Look to see whether the first parameter has a type or is a function typed parameter with a
|
| + // return type.
|
| + //
|
| Token afterType = skipFinalConstVarOrType(next);
|
| if (afterType == null) {
|
| return null;
|
| @@ -6328,6 +6565,11 @@ class Parser {
|
| if (identical(type, TokenType.STRING_INTERPOLATION_EXPRESSION)) {
|
| token = token.next;
|
| type = token.type;
|
| + //
|
| + // Rather than verify that the following tokens represent a valid expression, we simply skip
|
| + // tokens until we reach the end of the interpolation, being careful to handle nested string
|
| + // literals.
|
| + //
|
| int bracketNestingLevel = 1;
|
| while (bracketNestingLevel > 0) {
|
| if (identical(type, TokenType.EOF)) {
|
| @@ -6484,6 +6726,10 @@ class Parser {
|
| if (!matches4(startToken, TokenType.LT)) {
|
| return null;
|
| }
|
| + //
|
| + // We can't skip a type parameter because it can be preceeded by metadata, so we just assume
|
| + // that everything before the matching end token is valid.
|
| + //
|
| int depth = 1;
|
| Token next = startToken.next;
|
| while (depth > 0) {
|
| @@ -6532,9 +6778,18 @@ class Parser {
|
| builder.appendChar(currentChar);
|
| return index + 1;
|
| }
|
| + //
|
| + // We have found an escape sequence, so we parse the string to determine what kind of escape
|
| + // sequence and what character to add to the builder.
|
| + //
|
| int length = lexeme.length;
|
| int currentIndex = index + 1;
|
| if (currentIndex >= length) {
|
| + // Illegal escape sequence: no char after escape
|
| + // This cannot actually happen because it would require the escape character to be the last
|
| + // character in the string, but if it were it would escape the closing quote, leaving the
|
| + // string unclosed.
|
| + // reportError(ParserErrorCode.MISSING_CHAR_IN_ESCAPE_SEQUENCE);
|
| return length;
|
| }
|
| currentChar = lexeme.codeUnitAt(currentIndex);
|
| @@ -6552,12 +6807,14 @@ class Parser {
|
| builder.appendChar(0xB);
|
| } else if (currentChar == 0x78) {
|
| if (currentIndex + 2 >= length) {
|
| + // Illegal escape sequence: not enough hex digits
|
| reportError11(ParserErrorCode.INVALID_HEX_ESCAPE, []);
|
| return length;
|
| }
|
| int firstDigit = lexeme.codeUnitAt(currentIndex + 1);
|
| int secondDigit = lexeme.codeUnitAt(currentIndex + 2);
|
| if (!isHexDigit(firstDigit) || !isHexDigit(secondDigit)) {
|
| + // Illegal escape sequence: invalid hex digit
|
| reportError11(ParserErrorCode.INVALID_HEX_ESCAPE, []);
|
| } else {
|
| builder.appendChar(((Character.digit(firstDigit, 16) << 4) + Character.digit(secondDigit, 16)));
|
| @@ -6566,6 +6823,7 @@ class Parser {
|
| } else if (currentChar == 0x75) {
|
| currentIndex++;
|
| if (currentIndex >= length) {
|
| + // Illegal escape sequence: not enough hex digits
|
| reportError11(ParserErrorCode.INVALID_UNICODE_ESCAPE, []);
|
| return length;
|
| }
|
| @@ -6573,6 +6831,7 @@ class Parser {
|
| if (currentChar == 0x7B) {
|
| currentIndex++;
|
| if (currentIndex >= length) {
|
| + // Illegal escape sequence: incomplete escape
|
| reportError11(ParserErrorCode.INVALID_UNICODE_ESCAPE, []);
|
| return length;
|
| }
|
| @@ -6581,6 +6840,7 @@ class Parser {
|
| int value = 0;
|
| while (currentChar != 0x7D) {
|
| if (!isHexDigit(currentChar)) {
|
| + // Illegal escape sequence: invalid hex digit
|
| reportError11(ParserErrorCode.INVALID_UNICODE_ESCAPE, []);
|
| currentIndex++;
|
| while (currentIndex < length && lexeme.codeUnitAt(currentIndex) != 0x7D) {
|
| @@ -6592,18 +6852,21 @@ class Parser {
|
| value = (value << 4) + Character.digit(currentChar, 16);
|
| currentIndex++;
|
| if (currentIndex >= length) {
|
| + // Illegal escape sequence: incomplete escape
|
| reportError11(ParserErrorCode.INVALID_UNICODE_ESCAPE, []);
|
| return length;
|
| }
|
| currentChar = lexeme.codeUnitAt(currentIndex);
|
| }
|
| if (digitCount < 1 || digitCount > 6) {
|
| + // Illegal escape sequence: not enough or too many hex digits
|
| reportError11(ParserErrorCode.INVALID_UNICODE_ESCAPE, []);
|
| }
|
| appendScalarValue(builder, lexeme.substring(index, currentIndex + 1), value, index, currentIndex);
|
| return currentIndex + 1;
|
| } else {
|
| if (currentIndex + 3 >= length) {
|
| + // Illegal escape sequence: not enough hex digits
|
| reportError11(ParserErrorCode.INVALID_UNICODE_ESCAPE, []);
|
| return length;
|
| }
|
| @@ -6612,6 +6875,7 @@ class Parser {
|
| int thirdDigit = lexeme.codeUnitAt(currentIndex + 2);
|
| int fourthDigit = lexeme.codeUnitAt(currentIndex + 3);
|
| if (!isHexDigit(firstDigit) || !isHexDigit(secondDigit) || !isHexDigit(thirdDigit) || !isHexDigit(fourthDigit)) {
|
| + // Illegal escape sequence: invalid hex digits
|
| reportError11(ParserErrorCode.INVALID_UNICODE_ESCAPE, []);
|
| } else {
|
| appendScalarValue(builder, lexeme.substring(index, currentIndex + 1), (((((Character.digit(firstDigit, 16) << 4) + Character.digit(secondDigit, 16)) << 4) + Character.digit(thirdDigit, 16)) << 4) + Character.digit(fourthDigit, 16), index, currentIndex + 3);
|
| @@ -8169,6 +8433,9 @@ class ResolutionCopier implements ASTVisitor<bool> {
|
| this._toNode = toNode;
|
| return fromNode.accept(this);
|
| }
|
| + //
|
| + // Check for a simple transformation caused by entering a period.
|
| + //
|
| if (toNode is PrefixedIdentifier) {
|
| SimpleIdentifier prefix = toNode.prefix;
|
| if (fromNode.runtimeType == prefix.runtimeType) {
|
| @@ -8257,6 +8524,8 @@ class ResolutionCopier implements ASTVisitor<bool> {
|
| * AST node (and all of it's children) to a writer.
|
| */
|
| class ToFormattedSourceVisitor implements ASTVisitor<Object> {
|
| + static String COMMENTS_KEY = "List of comments before statement";
|
| +
|
| /**
|
| * The writer to which the source is to be written.
|
| */
|
| @@ -8448,9 +8717,11 @@ class ToFormattedSourceVisitor implements ASTVisitor<Object> {
|
| ScriptTag scriptTag = node.scriptTag;
|
| NodeList<Directive> directives = node.directives;
|
| visit(scriptTag);
|
| + // directives
|
| String prefix = scriptTag == null ? "" : " ";
|
| visitList7(prefix, directives, "\n");
|
| nl();
|
| + // declarations
|
| prefix = scriptTag == null && directives.isEmpty ? "" : "\n";
|
| visitList7(prefix, node.declarations, "\n\n");
|
| return null;
|
| @@ -9164,6 +9435,18 @@ class ToFormattedSourceVisitor implements ASTVisitor<Object> {
|
| indent();
|
| }
|
|
|
| + void printLeadingComments(Statement statement) {
|
| + List<String> comments = statement.getProperty(COMMENTS_KEY) as List<String>;
|
| + if (comments == null) {
|
| + return;
|
| + }
|
| + for (String comment in comments) {
|
| + _writer.print(comment);
|
| + _writer.print("\n");
|
| + indent();
|
| + }
|
| + }
|
| +
|
| /**
|
| * Safely visit the given node.
|
| *
|
| @@ -9269,10 +9552,12 @@ class ToFormattedSourceVisitor implements ASTVisitor<Object> {
|
| if (nodes != null) {
|
| int size = nodes.length;
|
| if (size != 0) {
|
| + // prefix
|
| _writer.print(prefix);
|
| if (prefix.endsWith("\n")) {
|
| indent();
|
| }
|
| + // nodes
|
| bool newLineSeparator = separator.endsWith("\n");
|
| for (int i = 0; i < size; i++) {
|
| if (i > 0) {
|
| @@ -9281,8 +9566,13 @@ class ToFormattedSourceVisitor implements ASTVisitor<Object> {
|
| indent();
|
| }
|
| }
|
| - nodes[i].accept(this);
|
| + ASTNode node = nodes[i];
|
| + if (node is Statement) {
|
| + printLeadingComments(node);
|
| + }
|
| + node.accept(this);
|
| }
|
| + // suffix
|
| _writer.print(suffix);
|
| }
|
| }
|
|
|
Chromium Code Reviews
Issue 137863002:
Issue 8742. Preserve leading line comments during java2dart translation. (Closed)
Base URL: https://dart.googlecode.com/svn/branches/bleeding_edge/dart