Replace underscore with dash in source_name (#130)

* Replace underscore with dash in source_name

In both the new and old version of `Kildomaten` we sanitize source names by replacing underscores with dashes, see:
- https://github.com/statisticsnorway/terraform-dapla-source-data-processor/blob/c38815d77401470ca26c576316f02a511a11fe8d/main.tf#L30
- https://github.com/statisticsnorway/terraform-dapla-source-data-automation/blob/f268e4ec03dbd4c6d250348e12a1f7c8d8d7daed/main.tf#L31

Users should not have to concern themselves with this impl detail ,so this PR abstract it away by doing the same sanitization here.

* Fix import sorting error introduced by commit hooks

* Add function that extracts the environment name from project id

* Explicitly type the dictionary

And update `Kuben` test case with valid project id

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "dapla-toolbelt"
-version = "2.0.9"
+version = "2.0.10"
 description = "Dapla Toolbelt"
 authors = ["Dapla Developers <[email protected]>"]
 license = "MIT"

src/dapla/pubsub.py

@@ -158,6 +158,28 @@ def _extract_project_name(project_id: str) -> str:
         )
 
 
+def _extract_env(project_id: str) -> t.Literal["test", "prod"]:
+    """Extracts the environment name from a GCP `Kuben` project ID.
+
+    This function assumes the project ID follows the `Kuben` project naming convention,
+    if it does the character before the last hyphen will represent the environment.
+
+    Args:
+        project_id (str): The project ID of a GCP `Kuben` project.
+
+    Returns:
+        t.Literal['test', 'prod']: the environment name.
+
+    Raises:
+        ValueError: If the project ID does not follow the `Kuben` format
+    """
+    char_to_env_map: dict[str, t.Literal["test", "prod"]] = {"t": "test", "p": "prod"}
+    env_char = project_id.split("-")[-2]
+    if env_char not in char_to_env_map.keys():
+        raise ValueError("Invalid project id")
+    return char_to_env_map[env_char]
+
+
 def trigger_source_data_processing(
     project_id: str, source_name: str, folder_prefix: str, kuben: bool = False
 ) -> None:
@@ -172,10 +194,14 @@ def trigger_source_data_processing(
     project_name = _extract_project_name(project_id)
 
     if kuben:
-        bucket_id = f"ssb-{project_name.rsplit('-', 1)[0]}-data-kilde-test"
+        env = _extract_env(project_id)
+        bucket_id = f"ssb-{project_name.rsplit('-', 1)[0]}-data-kilde-{env}"
     else:
         bucket_id = f"ssb-{project_name}-data-kilde"
 
+    # GCP resources for `Kildomaten` are created with dash as seperator instead of underscore
+    source_name = source_name.replace("_", "-")
+
     _publish_gcs_objects_to_pubsub(
         project_id, bucket_id, folder_prefix, topic_id=f"update-{source_name}"
     )

tests/test_pubsub.py

@@ -9,6 +9,7 @@
 from google.cloud import pubsub_v1
 
 import dapla.pubsub
+from dapla.pubsub import _extract_env
 from dapla.pubsub import _extract_project_name
 from dapla.pubsub import _generate_pubsub_data
 from dapla.pubsub import _get_callback
@@ -83,7 +84,7 @@ def test_trigger_source_data_processing(
     def test_trigger_source_data_processing_kuben(
         self, mock_publish_gcs_objects_to_pubsub: Mock
     ) -> None:
-        kuben_project_id = "my-team-t-jhdfb"
+        kuben_project_id = "dapla-kildomaten-p-zz"
 
         dapla.trigger_source_data_processing(
             kuben_project_id, self.source_folder_name, self.folder_prefix, True
@@ -94,7 +95,7 @@ def test_trigger_source_data_processing_kuben(
         # Check that _publish_gcs_objects_to_pubsub has been called with expected parameters
         mock_publish_gcs_objects_to_pubsub.assert_called_with(
             kuben_project_id,
-            "ssb-my-team-data-kilde-test",
+            "ssb-dapla-kildomaten-data-kilde-prod",
             self.folder_prefix,
             topic_id=self.topic_id,
         )
@@ -123,3 +124,17 @@ def test_extract_project_name(project_id: str, expected_project_name: str) -> No
 def test_invalid_project_id(invalid_project_id: str) -> None:
     with pytest.raises(ValueError):
         _extract_project_name(invalid_project_id)
+
+
+@pytest.mark.parametrize(
+    "project_id, expected_project_id",
+    [("dapla-kildomaten-p-zz", "prod"), ("dapla-t-zz", "test")],
+)
+def test_extract_env(project_id: str, expected_project_id: str) -> None:
+    assert _extract_env(project_id) == expected_project_id
+
+
+def test_extract_env_invalid_project() -> None:
+    project_id = "dapla-kildomaten-p"
+    with pytest.raises(ValueError):
+        _extract_env(project_id)