JavaParser

Consume one token of the specified type, or signal an error if it is not there.

If the next token is the identifier "record", convert it into the RECORD token. This makes it easier to handle records in various parts of the code, in particular when a parentToken is passed to some functions.

Annotation ::= TypeName [( [AnnotationArgument {, AnnotationArgument}] )] AnnotationArgument ::= ElementValuePair | ELementValue ElementValuePair ::= Identifier = ElementValue ElementValue ::= ConstExpressionSubset | ElementValueArrayInitializer | Annotation ElementValueArrayInitializer ::= { [ElementValue {, ElementValue}] [,] } ConstExpressionSubset ::= Literal | QualifiedName | ClassLiteral

We support only subset of const expressions expected in this context by java. If we encounter expression that we cannot parse, we do not raise parsing error, but instead we skip entire annotation silently.

CompilationUnit ::= [package QualId semi] TopStatSeq

Convert (qual)ident to type identifier

Translate names in Select/Ident nodes to type names.

Parse a sequence of field declarations, separated by commas. This one is tricky because a comma might also appear in an initializer. Since we don't parse initializers we don't know what the comma signifies. We solve this with a second list buffer maybe which contains potential variable definitions. Once we have reached the end of the statement, we know whether these potential definitions are real or not.

Tree representing java.lang.Object

Tree representing java.lang.Record

This is the general parse entry point. Overridden by ScriptParser

skip parent or brace enclosed sequence of things

If the last read offset is strictly greater than start, assign tree the span from start to last read offset, with given point. If the last offset is less than or equal to start, the tree t did not consume any source for its construction. In this case, don't assign a span yet, but wait for its span to be determined by setChildSpans when the parent node is positioned.

Positions tree. If t does not have a span yet, set its span to the given one.

in.offset, except if this is at a new line, in which case lastOffset is preferred.

Unconditionally issue an error at given span, without updating lastErrorOffset.

Issue an error at given offset if beyond last error offset and update lastErrorOffset.

The offset where the last syntax error was reported, or if a skip to a safepoint occurred afterwards, the offset of the safe point.