Multiple disks issue on windows and latest rdflib release.

[urbanmatthias]

In rdflib, they use urllib to parse file paths. In the latest release 5.0 from April 2020 they strip off the drive letter on windows machines with multiple drives: See RDFLib/rdflib#1172 (comment)
The reason for this is the use of urldecode, which confuses the drive letter with a protocol like http://

Currently the user must have the corresponding fix of rdflib installed, which is unlikely since it is not part of the latest release.

Proposed solutions:

• Catch the error tell the user to install rdflib from source. In the latest master branch the fix is available.
• Try to avoid the error even in old rdflib versions by explicitly providing the protocol file:// to rdflib's serialize. See Fix serialize with multiple disks on windows RDFLib/rdflib#1172 (comment)

If possible, the second is the preferred option

[pablo-de-andres]

I just ran into this issue again with @eich-2020 on his windows machine
He is using the latest OSP-core (v3.5.5) with rdflib v5.0.0. We attempted to upgrade rdflib, but OSP-core now has a constraint that it should be lower than v6, so that is not an option

[kysrpex]

I just ran into this issue again with @eich-2020 on his windows machine
He is using the latest OSP-core (v3.5.5) with rdflib v5.0.0. We attempted to upgrade rdflib, but OSP-core now has a constraint that it should be lower than v6, so that is not an option

For reference, I paste the traceback here.

INFO 2021-08-02 12:11:47,533 [osp.core.ontology.installation]: Will install the following namespaces: ['emmo']
INFO 2021-08-02 12:11:47,539 [osp.core.ontology.namespace_registry]: Loading namespaces emmo, http://emmo.info/emmo#.
ERROR 2021-08-02 12:11:47,865 [osp.core.pico]: An Exception occurred during installation.
Traceback (most recent call last):
  File "C:\Anaconda3\lib\site-packages\osp\core\pico.py", line 65, in install_from_terminal
    ontology_installer.install(args.files)
  File "C:\Anaconda3\lib\site-packages\osp\core\ontology\installation.py", line 35, in install
    self.install(files, self.get_new_packages, False)
  File "C:\Anaconda3\lib\site-packages\osp\core\ontology\installation.py", line 151, in _install
    self.namespace_registry.load_parser(parser)
  File "C:\Anaconda3\lib\site-packages\osp\core\ontology\namespace_registry.py", line 375, in load_parser
    ontology = Ontology(from_parser=parser)
  File "C:\Anaconda3\lib\site-packages\osp\core\ontology\ontology.py", line 120, in __init
    self._ontology_graph = deepcopy(parser.graph)
  File "C:\Anaconda3\lib\site-packages\osp\core\ontology\parser\owl\parser.py", line 87, in graph
    self._graph = self._read_ontology_graph(self._yaml_config,
  File "C:\Anaconda3\lib\site-packages\osp\core\ontology\parser\owl\parser.py", line 209, in _read_ontology_graph
    graph.parse(file_like,
  File "C:\Anaconda3\lib\site-packages\rdflib\graph.py", line 1078, in parse
    parser.parse(source, self, *args)
  File "C:\Anaconda3\lib\site-packages\rdflib\plugins\parsers\rdfxml.py", line 579, in parse
    self._parser.parse(source)
  File "C:\Anaconda3\lib\xml\sax\expatreader.py", line 111, in parse
    xmlreader.IncrementalParser.parse(self, source)
  File "C:\Anaconda3\lib\xml\sax\xmlreader.py", line 126, in parse
    buffer = file.read(self._bufsize)
  File "C:\Anaconda3\lib\encodings\cp1252.py", line 23, in decode
    return codecs.charmap_decode(input,self.errors,decoding_table)[0]
UnicodeDecodeError: 'charmap' codec can't decode byte 0x81 in position 16461: character maps to <undefined>
ERROR 2021-08-02 12:11:47,898 [osp.core.pico]: Consider running 'pico --log-level debug install ...'

[kysrpex]

I just ran into this issue again with @eich-2020 on his windows machine
He is using the latest OSP-core (v3.5.5) with rdflib v5.0.0. We attempted to upgrade rdflib, but OSP-core now has a constraint that it should be lower than v6, so that is not an option

This does not look like the same issue to me. More likely it has something to do with the encoding of the ontology file.

[kysrpex]

I am closing this and opening a new issue as the issue is unlikely to be related to this one.

EDIT: the new issue.