Opcua-tools

There was a need to add functions which will subset for reference types which are a subtype of the "NonHierarchicalReferences" in the OPC UA standard. Major changes include:

adding the function non_hierarchical_references() which finds the subset of references which are the references which are subtypes of "NonHierarchicalReferences".
adding the function non_hierarchical_references_trg_has_no_modelling_rule() which finds the subset of references which are the references which are subtypes of "NonHierarchicalReferences" that point to target nodes which have no ModellingRule references.

The major changes in this include:

- Adding `NullHandler` logger and removing other logging config which is recommended for logging of python packages. See README for more information
- Added `create_nodes_path()` as a member function to UAGraph

This release simply rolls back to pandas version under `1.5.0` and removes a FutureDepricationWarning, but casting a numeric column to int instead of allowing it to be `dtype=object`.

An UAEURange class was added which resembles the UAEngineeringUnits class. The `XML` parser will also create the UAEURange from EURange tags in the `XML` when following the convention of EURange objects in the OPC UA standard. The appropiate tests were added for parsing and encoding or exporting to `XML` from a UAGraph.

Additionally, `print()` functions were also changed to use the `logging` Python module. The format of the logging was also changed.

Dynamic handling of `xmlns` prefixes in Enumeration definitions. Instead of exactly matching the string of the tag name, a substring match is performed.

This patch allows for handling of enums such that trailing and leading whitespaces are ignored when referring to an enum value. Additionally, it forces enums strings to follow the 'title' convention with spaces and the first of every word capitalized.