* As each node has to be handled differently. * Also we can attach many information for some nodes to the token, to make linting eaiser. But that's specific per node and those goes to an own class. Relates #67
* As official project does no longer modify the file to tokenize. * Add tests to test integration.