#3646 Use os.cpu_count rather than processor affinity

.github/workflows/publish_pypi.yml

@@ -41,8 +41,8 @@ jobs:
         id: get_num_cores
         shell: python
         run: |
-          from os import environ, sched_getaffinity
-          num_cpus = len(sched_getaffinity(0))
+          from os import environ, cpu_count
+          num_cpus = cpu_count()
           output_file = environ['GITHUB_OUTPUT']
           with open(output_file, "a", encoding="utf-8") as output_stream:
               output_stream.write(f"count={num_cpus}\n")
@@ -80,9 +80,9 @@ jobs:
             VCPKG_DEFAULT_TRIPLET=x64-windows-static-md
             VCPKG_FEATURE_FLAGS=manifests,registries
             CMAKE_GENERATOR="Visual Studio 17 2022"
-            CMAKE_GENERATOR_PLATFORM=x64'
-            CMAKE_BUILD_PARALLEL_LEVEL: ${{ steps.get_num_cores.outputs.num_cpus }}
-          CIBW_ARCHS: "AMD64"
+            CMAKE_GENERATOR_PLATFORM=x64
+            CMAKE_BUILD_PARALLEL_LEVEL=${{ steps.get_num_cores.outputs.count }}
+          CIBW_ARCHS: AMD64
           CIBW_BEFORE_BUILD: python -m pip install setuptools wheel # skip CasADi and CMake
           CIBW_TEST_COMMAND: python -c "import pybamm; pybamm.IDAKLUSolver()"