[23] ECJ rejects 'module' as top level package in an import statement

[jarthana]

Given the two classes as below:

package module;
public class Z {}

and

import module.Z; // Error here - Syntax error on token ".", delete this token
public class X {} 

However, we accept module.Z z = null;

The relevant spec that allows this:

During the reduction of input characters to input elements ([3.5](https://docs.oracle.com/javase/specs/jls/se22/html/jls-3.html#jls-3.5)), a sequence of input characters that notionally matches a contextual keyword is reduced to a contextual keyword if and only if both of the following conditions hold:

   1.  The sequence is recognized as a terminal specified in a suitable context of the syntactic grammar ([2.3](https://docs.oracle.com/javase/specs/jls/se22/html/jls-2.html#jls-2.3)), as follows:
        For module and open, when recognized as a terminal in a ModuleDeclaration ([7.7](https://docs.oracle.com/javase/specs/jls/se22/html/jls-7.html#jls-7.7)).

        For module, when recognized as a terminal in a SingleModuleImportDeclaration ([7.5.5](https://cr.openjdk.org/~gbierman/jep476/jep476-20240604/specs/module-import-declarations-jls.html#jls-7.5.5)) or a ModuleDeclaration ([7.7](https://docs.oracle.com/javase/specs/jls/se22/html/jls-7.html#jls-7.7)).

As per this, the module in import module.Z shouldn't be recognized as as a contextual keyword. Copying @stephan-herrmann for weighing in.

[stephan-herrmann]

Thanks, @jarthana for catching this.

In fact I had prepared handling of "import module." but then forgot two things:

• write tests for it
• make sure it works not only in module-info.java but also in regular CU (where normally no disambiguation would happen)