Adds license headers and performs linting

Author: mafaldasv-dev

examples/opik_vs_elastic/helpers/indexing.py

@@ -78,11 +78,7 @@ def get_elasticsearch_client() -> ElasticsearchClient:
     """Build a client from CLOUD_ID/ELASTICSEARCH_API_KEY or ES_URL/ES_HOST env vars."""
     cloud_id = os.getenv("CLOUD_ID")
     api_key = os.getenv("ELASTICSEARCH_API_KEY")
-    host = (
-        os.getenv("ES_URL")
-        or os.getenv("ELASTICSEARCH_ENDPOINT")
-        or os.getenv("ES_HOST")
-    )
+    host = os.getenv("ES_URL") or os.getenv("ELASTICSEARCH_ENDPOINT") or os.getenv("ES_HOST")
 
     if cloud_id and api_key:
         es = Elasticsearch(cloud_id=cloud_id, api_key=api_key)
@@ -91,8 +87,6 @@ def get_elasticsearch_client() -> ElasticsearchClient:
     elif host:
         es = Elasticsearch(hosts=[host])
     else:
-        raise ValueError(
-            "Set CLOUD_ID + ELASTICSEARCH_API_KEY, or ES_URL/ES_HOST, to connect to Elasticsearch."
-        )
+        raise ValueError("Set CLOUD_ID + ELASTICSEARCH_API_KEY, or ES_URL/ES_HOST, to connect to Elasticsearch.")
 
     return ElasticsearchClient(es=es)

examples/opik_vs_elastic/orca_helpers/opik_client.py

@@ -1,3 +1,7 @@
+# Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+# or more contributor license agreements. Licensed under the Elastic License 2.0;
+# you may not use this file except in compliance with the Elastic License 2.0.
+
 """Opik dataset client, mirroring and extending orca_helpers's opik_client.py.
 
 orca_helpers's ``clients/opik_client.py`` only wraps ``get_dataset`` and
@@ -62,9 +66,7 @@ def list_datasets(
         Falls back to the client's default project when ``project_name`` is
         omitted.
         """
-        return self._sdk_client.get_datasets(
-            max_results=max_results, project_name=project_name
-        )
+        return self._sdk_client.get_datasets(max_results=max_results, project_name=project_name)
 
     def add_rows(
         self,
@@ -74,9 +76,7 @@ def add_rows(
         ignore_keys: list[str] | None = None,
     ) -> None:
         """Insert dataframe rows into a dataset, renaming/skipping columns."""
-        dataset.insert_from_pandas(
-            df, keys_mapping=keys_mapping, ignore_keys=ignore_keys
-        )
+        dataset.insert_from_pandas(df, keys_mapping=keys_mapping, ignore_keys=ignore_keys)
 
     def get_rows(
         self,
@@ -88,9 +88,7 @@ def get_rows(
 
     def dataset_url(self, dataset: Dataset) -> str:
         """Build a UI URL for viewing this dataset in Opik."""
-        return url_helpers.get_dataset_url_by_id(
-            dataset_id=dataset.id, url_override=OpikConfig().url_override
-        )
+        return url_helpers.get_dataset_url_by_id(dataset_id=dataset.id, url_override=OpikConfig().url_override)
 
     def project_url(self, project_name: str | None = None) -> str:
         """Build a UI URL for viewing a project in Opik."""

examples/opik_vs_elastic/orca_helpers/scripts/index_kb_with_orca.py

@@ -1,3 +1,7 @@
+# Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+# or more contributor license agreements. Licensed under the Elastic License 2.0;
+# you may not use this file except in compliance with the Elastic License 2.0.
+
 """Index the Wix knowledge base into Elasticsearch using orca_helpers's real pipeline.
 
 This mirrors orca_helpers's own ``orca_helpers.etl.indexing.cli`` so it runs the exact code the
@@ -40,10 +44,7 @@ def main() -> int:
 
     es_client = get_elasticsearch_client()
     if es_client is None:
-        raise RuntimeError(
-            "Could not connect to Elasticsearch. Set CLOUD_ID + ELASTICSEARCH_API_KEY "
-            "or ES_HOST."
-        )
+        raise RuntimeError("Could not connect to Elasticsearch. Set CLOUD_ID + ELASTICSEARCH_API_KEY or ES_HOST.")
 
     indexing_config = IndexingConfig(gcs_root=GCS_ROOT, recreate_indices=True)
 

examples/opik_vs_elastic/orca_helpers/scripts/register_datasets_opik.py

@@ -1,3 +1,7 @@
+# Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+# or more contributor license agreements. Licensed under the Elastic License 2.0;
+# you may not use this file except in compliance with the Elastic License 2.0.
+
 """Register the full Wix QA golden dataset into Opik.
 
 Mirrors orca_helpers's dataset-registration pattern (see