fix: handle cases when running inside a kubernetes cluster (#17)

The current setup for the plugin assumes that we have a kubeconfig file
used to connect to the cluster. However, this is not the case if the
code submitting the job is itself running in a pod in the cluster, as we
will need to use the default service account provided by the cluster.
This modifies the init to catch any errors when loading a kubeconfig
file and use the incluster configuration instead

The issue in the python kubernetes client here describes the problem:
kubernetes-client/python#1005

snakemake_executor_plugin_kubernetes/__init__.py

@@ -84,7 +84,10 @@ class ExecutorSettings(ExecutorSettingsBase):
 # Implementation of your executor
 class Executor(RemoteExecutor):
     def __post_init__(self):
-        kubernetes.config.load_kube_config()
+        try:
+            kubernetes.config.load_kube_config()
+        except kubernetes.config.config_exception.ConfigException:
+            kubernetes.config.load_incluster_config()
 
         self.k8s_cpu_scalar = self.workflow.executor_settings.cpu_scalar
         self.k8s_service_account_name = (