ClassCastException in Node.getString

[jon-bell]

We ran into an unexpected ClassCastException when attempting to compile the following string using Rhino 1.7.12: ["".h]=y

Output:

	at org.mozilla.javascript.Node.getString(Node.java:537)
	at org.mozilla.javascript.optimizer.BodyCodegen.generateExpression(Codegen.java:2262)
	at org.mozilla.javascript.optimizer.BodyCodegen.visitSetProp(Codegen.java:5132)
	at org.mozilla.javascript.optimizer.BodyCodegen.generateExpression(Codegen.java:2610)
	at org.mozilla.javascript.optimizer.BodyCodegen.generateExpression(Codegen.java:2309)
	at org.mozilla.javascript.optimizer.BodyCodegen.generateExpression(Codegen.java:2807)
	at org.mozilla.javascript.optimizer.BodyCodegen.generateExpression(Codegen.java:2309)
	at org.mozilla.javascript.optimizer.BodyCodegen.generateExpression(Codegen.java:2807)
	at org.mozilla.javascript.optimizer.BodyCodegen.generateStatement(Codegen.java:2059)
	at org.mozilla.javascript.optimizer.BodyCodegen.generateStatement(Codegen.java:1870)
	at org.mozilla.javascript.optimizer.BodyCodegen.generateBodyCode(Codegen.java:1292)
	at org.mozilla.javascript.optimizer.Codegen.generateCode(Codegen.java:307)
	at org.mozilla.javascript.optimizer.Codegen.compileToClassFile(Codegen.java:166)
	at org.mozilla.javascript.optimizer.Codegen.compile(Codegen.java:75)
	at org.mozilla.javascript.Context.compileImpl(Context.java:2551)
	at org.mozilla.javascript.Context.compileString(Context.java:1507)
	at org.mozilla.javascript.Context.compileString(Context.java:1496)
	at edu.berkeley.cs.jqf.examples.rhino.CompilerTest.main(CompilerTest.java:123)

[rbri]

Can you please prepare a simple test case to make it easy for us to reproduce this.

You can have a look at the packages org.mozilla.javascript.tests.es5 or org.mozilla.javascript.tests.es6 for sample tests.

[tuchida]

["".h]=""
// Exception in thread "main" java.lang.ClassCastException: org.mozilla.javascript.ast.StringLiteral cannot be cast to org.mozilla.javascript.ast.Name

[1..h]="" // no error

[0..toString.h] = [123] // // expect: Number.prototype.toString.h === 123
// Exception in thread "main" java.lang.NullPointerException

Perhaps it has something to do with this code.

rhino/src/org/mozilla/javascript/Parser.java

Lines 4196 to 4205 in c275c35

	if (nodeType == Token.GETPROP) {
	type = Token.SETPROP;
	// TODO(stevey) - see https://bugzilla.mozilla.org/show_bug.cgi?id=492036
	// The new AST code generates NAME tokens for GETPROP ids where the old
	// parser
	// generated STRING nodes. If we don't set the type to STRING below, this
	// will
	// cause java.lang.VerifyError in codegen for code like
	// "var obj={p:3};[obj.p]=[9];"
	id.setType(Token.STRING);

[p-bakker]

You sure those double dots (..) are correct in your examples? As I think that makes it E4X-specific syntax and thus completely different codePaths, no?

[tuchida]

Yes. The first dot is a decimal point. The second dot is a member operator.
http://adripofjavascript.com/blog/drips/numbers-and-javascripts-dot-notation.html

[p-bakker]

es. The first dot is a decimal point. The second dot is a member operator.

Right... learned something new about JavaScript :-)