Merge branch 'dev' into f-spss-value-label-name

CHANGELOG.md

@@ -19,6 +19,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Fixed ###
 
+- Handling of Pandas/Numpy Integers in Identifiable class
+
 ### Security ###
 
 ### Documentation ###

src/caoscrawler/identifiable.py

@@ -23,6 +23,7 @@ from __future__ import annotations
 
 import json
 import logging
+import numpy as np
 from datetime import datetime
 from hashlib import sha256
 from typing import Optional, Union
@@ -97,10 +98,10 @@ class Identifiable():
         elif isinstance(value, list):
             return "[" + ", ".join([Identifiable._value_representation(el) for el in value]) + "]"
         elif (isinstance(value, str) or isinstance(value, int) or isinstance(value, float)
-              or isinstance(value, datetime)):
+              or isinstance(value, datetime) or np.issubdtype(type(value), np.number)):
             return str(value)
         else:
-            raise ValueError(f"Unknown datatype of the value: {value}")
+            raise ValueError(f"Unknown datatype of the value: {value} with type {type(value)}.")
 
     @staticmethod
     def _create_hashable_string(identifiable: Identifiable) -> str:

unittests/test_identifiable.py

@@ -25,6 +25,7 @@ test identifiable module
 """
 
 import linkahead as db
+import numpy as np
 import pytest
 
 from caoscrawler.identifiable import Identifiable
@@ -86,3 +87,13 @@ def test_equality():
         record_id=12, properties={"a": 0}) == Identifiable(properties={"a": 0})
     assert Identifiable(properties={"a": 0}) == Identifiable(properties={"a": 0})
     assert Identifiable(properties={"a": 0}) != Identifiable(properties={"a": 1})
+
+
+def test_numpy_int():
+    val_int = int(123)
+    val_64 = np.int64(123)
+
+    id_int = Identifiable(name="A", record_type="B", properties={'a': val_int})
+    id_64 = Identifiable(name="A", record_type="B", properties={'a': val_64})
+
+    assert id_int.get_representation() == id_64.get_representation()