Fixed a bug in the methods to delete name, given name and last name of instances of the ResponsibleAgent class.

[arcangelo7]

I found and fixed a small bug in the remove_name, remove_given_name and remove_family_name methods of the ResponsibleAgent class. I'll report the code with and without the bug regarding remove_family_name, but the same goes for the other two functions:

# Code with the bug: self.g is not a subject
def remove_family_name(self) -> None:
    self.g.remove((self.g, GraphEntity.iri_family_name, None))

# Code without the bug: self.g -> self.res
def remove_family_name(self) -> None:
    self.g.remove((self.res, GraphEntity.iri_family_name, None))

[iosonopersia]

Hi @arcangelo7 , thank you for this submission.
I'm going to merge your fix right now.
Are you aware of other bugs that could be solved before publishing a new version of the library?

[arcangelo7]

Hi @arcangelo7 , thank you for this submission.
I'm going to merge your fix right now.
Are you aware of other bugs that could be solved before publishing a new version of the library?

Yes, I am aware of another bug, but I only have a temporary solution to it, unconvincing if I have to be honest.
The bug occurs by running the upload_all function of the Storer class and occurs when the update_query returned from get_update_query is None instead of "". In that case if update_query == "" does not work and the function crashes when trying to concatenate query_string += " ; " + update_query at row 279, because it cannot concatenate a string with None.

oc_ocdm/oc_ocdm/storer.py

Line 279 in 61e5e4e

query_string += " ; " + update_query

My temporary solution was to check whether update_query is None as well as "", this way:

if update_query == "" or update_query is None:
    skipped_queries += 1

However, I do not consider it a solution, because it should be understood why sometimes the update_query is None instead of "".

[iosonopersia]

get_update_query() returns None when the given entity is to_be_deleted and its preexisting_graph is empty (either because it's a new entity or because the preexisting_graph wasn't provided). Obviously, this is an unwanted behaviour.

@arcangelo7 : thank you again for spotting all these issues... I'm going to fix this and then a new version will be published!