Detect keywords Vs Identifiers ambiguity.

[bd82]

We have the strings of the keywords. either those strings are passed directly as strings
or as simple regExps which we can extract the string from.

So we can check if any pattern could match these keywords
And if so validate that the longer_alt property is used correctly.

[bd82]

I think this should be split up into several different validations.

1. Unreachable Token - e.g: an Identifier appearing before a keyword.

2. Missing longer_alt property - e.g:

   • a keyword appearing before an identifier.
   • An Integer Token appearing before a decimal Token (although we cannot easily detect this).

The missing longer_alt property can be automatically fixed in most cases simply by adding it.

• If multiple longer_alt properties must be added than it is an error.
• If the token already has a longer_alt property we should not validate it as we assume the user knows what they are doing.
• It should also be possible to define the longer_alt property as Lexer.NA to explicitly
  tell the Lexer to ignore this validation, thus providing backward compatible behavior.

[bd82]

detecting Missing longer_alt property requires more in depth regExp analysis.