Improved WM element (ID) handling

[matthias-mayr]

When testing, I ran into two issues that are addressed in this PR.

1. The lightstring2uri method could return a string instead of an rdflib URIRef. This caused errors of this kind:

/wm: AauSpatialReasoner  warn: Adding relation to http://www.inf.ufrgs.br/phi-group/ontologies/cora.owl#Robot-1-skiros:test_robot
/wm: statement to add ('http://www.inf.ufrgs.br/phi-group/ontologies/cora.owl#Robot-1', rdflib.term.URIRef('http://www.w3.org/1999/02/22-rdf-syntax-ns#type'), rdflib.term.URIRef('http://www.w3.org/2002/07/owl#Named
  ~  Individual'))
/test_robot: Service call failed: service [/wm/modify] responded with an error: b'error processing request: Subject http://www.inf.ufrgs.br/phi-group/ontologies/cora.owl#Robot-1must be an rdflib term'

2. When obtaining the ID number from an element, it was assumed that there is no dash "-" other than the separator between type and ID in the URI.
   Since we have URIs like http://www.inf.ufrgs.br/phi-group/ontologies/cora.owl#Robot-1, this could lead to errors like:

File "/home/matthias/Workspaces/skiros2_noetic_ws/src/skiros2/skiros2_world_model/src/skiros2_world_model/core/world_model.py", line 557, in _uri2id\n    return int(uri.split
   ~  (\'-\')[1])\n', "ValueError: invalid literal for int() with base 10: 'group/ontologies/cora.owl#Robot'\n"]
/test_robot: Service call failed: service [/wm/modify] responded with an error: b"error processing request: invalid literal for int() with base 10: 'group/ontologies/cora.owl#Robot'"

This two commits address both. However, I am not sure if they would cause side effects.
@frovida: If you have sth to comment, please do so.

I will merge them into the ROS 2 branch, so they would get some sort of testing.
This also came up when "reviving" the skiros2_test_lib. Once it's fully up, we can also make changes with some more confidence.

[matthias-mayr]

wrong mention @frvd

[frvd]

@matthias-mayr I think it would be just easier to replace find with rfind :)

[matthias-mayr]

rfind for the "-" or the "#"? Assume the former since the latter does not seem to make any sense to me

However just using rfind for the "-" does not necessarily yield to the expected result.
Example:
A malformed URI like http://www.inf.ufrgs.br/phi-group/ontologies/cora.owl#Robot that does not have an ID should lead to a return value of "-1" according to the previous/other code.
That's why I wanted to search explicitly after the hash if there's one

[frvd]

Right, what about the following?

hash_pos = self._id.rfind('#')
dash_pos = self._id.rfind("-")
return -1 if dash_pos == -1 or dash_pos < hash_pos else int(self._id.split('-')[1])