statnett/cimsparql

cimsparql's Issues

Clean up Prefix/CimModel/GraphDB hirarchy

There are some "problems" with the Prefix/CimModel/GraphDB classes that makes the code difficult to read

CimModel calls setup_client that does not exist

_setup_client is implemented lower in the hirarchy. Thus it is not possible to make an instance of CimModel even though it is not abstract.
Solution: Either make the class abstract and require that children implement _setup_client. It appears like CimModel does not need any of the things in _setup_client, thus the call can probably just be removed, and then the child GraphDBClient can call it.

_prefixes in GraphDBClient initialized with URL???

_prefixes is a dict, and by following the setter method it is a dict. But when a property-setter is called with a URL and the result is a dict, it is a bit confustion.

Suggestion: Create a function fetch_prefixes that returns the dict and write

self.prefixes = fetch_prefixes(url)

_setup_client adds attributes

It would be quite easy to avoid this. In __init__ move

       self.user = os.getenv("GRAPHDB_USER")
        self.passwd = os.getenv("GRAPHDB_USER_PASSWD")
        self.service = service
        self.sparql = SPARQLWrapper(self.service)

into __init__. And add a method _init_sparql that performs the additional initialization of the sparql attribute

Prefix access self._prefixes before initialized

If and instance of Prefix is made, it will crash if the prefixes property is called.

Solution: Add an __init__ method to Prefix and initialize self._prefixes with None or empty dict. If None, it must be handled in the property (e.g. return empty dict if None)

Prefix: cim_version assumes that the "cim" key exist

Suggestion: Check if "cim" exists first. If not, return None

CimModel: TypeMapper only needs the prefixes.

Solution: is it possible to only pass prefixes instead of the entire CimModel instance

CimModel: Initialize self._data_version instead of handling AttributeError in the property

Reduce size of "unused" variables in queries

SNMST-MasterCim15-VERSION-LATEST not in GraphDB repos

Make network_analysis available to user in query

Add docstrings

Remove statnett.no from default

Add authentication

Add query for temperature curve dependent limits

with_sequence_number in load_query and synchronous_machines_query has no effect

Use ApiRegistry for GraphDB LATEST

Add dry-run to all client queries

ac_lines does not return correct temperature correction factors

Use template based predefined queries

It is currently impossible to read predefined queries just by looking at the code, you have to use dry_run to understand the query. We should generate templates for all the queries, making them easier to read, maintain and optimize.

The templates could be defined as variables

BUS_DATA="SELECT ?mrid ?name ?busname ?un Where {?t_mrid ....   FILTER regex(?area, '${area}'}"

or it could be read from files

SELECT  ?mrid ?name ?busname ?un
WHERE {
  ?t_mrid  rdf:type cim:TopologicalNode;
           cim:IdentifiedObject.name ?busname;
           cim:IdentifiedObject.mRID ?mrid;
           cim:TopologicalNode.BaseVoltage/cim:BaseVoltage.nominalVoltage ?un;
           cim:TopologicalNode.ConnectivityNodeContainer ?cont .
  ?cont cim:IdentifiedObject.aliasName|cim:IdentifiedObject.name ?name;
        cim:VoltageLevel.Substation/cim:Substation.Region/cim:SubGeographicalRegion.Region/cim:IdentifiedObject.name ?area .
  FILTER regex(?area, '${area}')
}

where final query is generated using string Template.

from string import Template

with open('bus_data.sparql', 'r') as fid:
    BUS_DATA = fid.read()
    
bus_data_query = Template(BUS_DATA).substitute({'area': 'NO'})

Using queries defined in files makes it possible to use an editor with sparql-mode capabilities for both reading/writing, formatting and testing.

Dependency Dashboard

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Open

These updates have all been created already. Click a checkbox below to force a retry/rebase of any.

• Lock file maintenance

Detected dependencies

github-actions

.github/workflows/codeql-analysis.yml

• actions/checkout v4
• github/codeql-action v3
• github/codeql-action v3
• github/codeql-action v3

.github/workflows/deploy.yml

• actions/checkout v4
• actions/setup-python v5
• abatilo/actions-poetry v2.3.0
• actions/checkout v4
• actions/setup-python v5
• abatilo/actions-poetry v2.3.0
• actions/configure-pages v4
• actions/upload-pages-artifact v3
• actions/deploy-pages v4

.github/workflows/poetry.yml

• actions/checkout v4
• actions/setup-python v5
• abatilo/actions-poetry v2.3.0
• fredrikaverpil/poetry-update v1.2
• peter-evans/create-pull-request v5

.github/workflows/tests.yml

• actions/checkout v4
• actions/setup-python v5
• abatilo/actions-poetry v2.3.0
• actions/checkout v4
• actions/setup-python v5
• abatilo/actions-poetry v2.3.0
• codecov/codecov-action v3
• eclipse/rdf4j-workbench 4.3.8

poetry

pyproject.toml

• python ^3.10

---

• Check this box to trigger a request for Renovate to run again on this repository

Different return values

Fix type-hints

Type-hints are missing some places for example

 @classmethod
    def col_map(cls, data_row, columns) -> Tuple[Dict[str, str]]:
        columns = columns or {}
        col_map = cls._col_map(data_row)
        col_map.update(columns)
        return col_map

and the return type also seems off.