@Simulant87 Can you include unit tests for the new behavior?

@stleary Sure, I added tests for the new exception messages.

I am not sure, if you like my latest change, explainging the numbers in the syntax error with (zero based) "index" and (1 based) "number in line". I think this makes it easier to find the error when counting for the wrong character in the input string. But it requires a lot of tests to be updated.

If you don't like it, I can also revert it again. But I think it makes the error messages easier to read.

What problem does this code solve?
Some tokener parsing error messages are not sufficiently descriptive

Does the code still compile with Java6?
Yes

Risks
Low. It's possible that some users' unit tests may fail due to this change. I think the value of the new messages outweighs this risk.

Changes to the API?
No, just changes to some parsing error messages.

Will this require a new release?
No

Should the documentation be updated?
No

Does it break the unit tests?
No, there was no previous coverage for these errors. New tests were added.

Was any code refactored in this commit?
No

Review status
APPROVED

Starting 3-day comment window.