Destination Pinecone: Add source_tag for attribution + unit tests (#3…

…8151)

Co-authored-by: Aaron ("AJ") Steers <[email protected]>

airbyte-integrations/connectors/destination-pinecone/destination_pinecone/indexer.py

@@ -2,6 +2,7 @@
 # Copyright (c) 2023 Airbyte, Inc., all rights reserved.
 #
 
+import os
 import uuid
 from typing import Optional
 
@@ -25,14 +26,17 @@
 
 MAX_IDS_PER_DELETE = 1000
 
+AIRBYTE_TAG = "airbyte"
+AIRBYTE_TEST_TAG = "airbyte_test"
+
 
 class PineconeIndexer(Indexer):
     config: PineconeIndexingModel
 
     def __init__(self, config: PineconeIndexingModel, embedding_dimensions: int):
         super().__init__(config)
         try:
-            self.pc = PineconeGRPC(api_key=config.pinecone_key, threaded=True)
+            self.pc = PineconeGRPC(api_key=config.pinecone_key, source_tag=self.get_source_tag, threaded=True)
         except PineconeException as e:
             return AirbyteConnectionStatus(status=Status.FAILED, message=str(e))
 
@@ -62,6 +66,10 @@ def pre_sync(self, catalog: ConfiguredAirbyteCatalog):
     def post_sync(self):
         return []
 
+    def get_source_tag(self):
+        is_test = "PYTEST_CURRENT_TEST" in os.environ or "RUN_IN_AIRBYTE_CI" in os.environ
+        return AIRBYTE_TEST_TAG if is_test else AIRBYTE_TAG
+
     def delete_vectors(self, filter, namespace=None, prefix=None):
         if self._pod_type == "starter":
             # Starter pod types have a maximum of 100000 rows

airbyte-integrations/connectors/destination-pinecone/integration_tests/pinecone_integration_test.py

@@ -4,6 +4,7 @@
 
 import json
 import logging
+import os
 import time
 
 from airbyte_cdk.destinations.vector_db_based.embedder import OPEN_AI_VECTOR_SIZE
@@ -48,6 +49,8 @@ def tearDown(self):
             else :
                 print("Noting to delete. No data in the index/namespace.")
 
+    def test_integration_test_flag_is_set(self):
+        assert "PYTEST_CURRENT_TEST" in os.environ
 
     def test_check_valid_config(self):
         outcome = DestinationPinecone().check(logging.getLogger("airbyte"), self.config)        

airbyte-integrations/connectors/destination-pinecone/metadata.yaml

@@ -13,7 +13,7 @@ data:
   connectorSubtype: vectorstore
   connectorType: destination
   definitionId: 3d2b6f84-7f0d-4e3f-a5e5-7c7d4b50eabd
-  dockerImageTag: 0.1.0
+  dockerImageTag: 0.1.1
   dockerRepository: airbyte/destination-pinecone
   documentationUrl: https://docs.airbyte.com/integrations/destinations/pinecone
   githubIssueLabel: destination-pinecone

airbyte-integrations/connectors/destination-pinecone/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "airbyte-destination-pinecone"
-version = "0.1.0"
+version = "0.1.1"
 description = "Airbyte destination implementation for Pinecone."
 authors = ["Airbyte <[email protected]>"]
 license = "MIT"

airbyte-integrations/connectors/destination-pinecone/unit_tests/destination_test.py

@@ -66,6 +66,26 @@ def test_check_with_errors(self, MockedEmbedder, MockedPineconeIndexer):
         mock_embedder.check.assert_called_once()
         mock_indexer.check.assert_called_once()
 
+    def test_check_with_config_errors(self):
+        bad_config = {
+            "processing": {"text_fields": ["str_col"], "metadata_fields": [], "chunk_size": 1000},
+            "embedding_2": {"mode": "openai", "openai_key": "mykey"},
+            "indexing": {
+                "pinecone_key": "mykey",
+                "pinecone_environment": "myenv",
+                "index": "myindex",
+            },
+        }
+        destination = DestinationPinecone()
+        result = destination.check(self.logger, bad_config)
+        self.assertEqual(result.status, Status.FAILED)
+
+    def test_check_with_init_indexer_errors(self):
+        destination = DestinationPinecone()
+        with patch("destination_pinecone.destination.PineconeIndexer", side_effect=Exception("Indexer Error")):
+            result = destination.check(self.logger, self.config)
+        self.assertEqual(result.status, Status.FAILED)
+
     @patch("destination_pinecone.destination.Writer")
     @patch("destination_pinecone.destination.PineconeIndexer")
     @patch("destination_pinecone.destination.create_from_config")

airbyte-integrations/connectors/destination-pinecone/unit_tests/pinecone_indexer_test.py

@@ -2,6 +2,7 @@
 # Copyright (c) 2023 Airbyte, Inc., all rights reserved.
 #
 
+import os
 from unittest.mock import ANY, MagicMock, Mock, call, patch
 
 import pytest
@@ -55,6 +56,36 @@ def mock_determine_spec_type():
         mock.return_value = "pod"
         yield mock
 
+
+def test_get_source_tag_default():
+    # case when no test env variables are set
+    os.environ.pop("PYTEST_CURRENT_TEST", None)
+    os.environ.pop("RUN_IN_AIRBYTE_CI", None)
+    indexer = create_pinecone_indexer()
+    assert indexer.get_source_tag() == "airbyte"
+
+
+def test_get_source_tag_with_pytest():
+    # pytest is running by default here
+    indexer = create_pinecone_indexer()
+    assert indexer.get_source_tag() == "airbyte_test"
+
+    # pytest plus ci is running
+    with patch.dict("os.environ", {"RUN_IN_AIRBYTE_CI": "value does not matter"}):
+        assert indexer.get_source_tag() == "airbyte_test"
+
+
+@patch.dict("os.environ", {"RUN_IN_AIRBYTE_CI": "Value does not matter"})
+def test_get_source_tag_with_ci():
+    # CI and pytest is running 
+    indexer = create_pinecone_indexer()
+    assert indexer.get_source_tag() == "airbyte_test"
+
+    # CI is running but pytest is not
+    with patch.dict("os.environ", {"PYTEST_CURRENT_TEST": "Value does not matter"}):
+        assert indexer.get_source_tag() == "airbyte_test"
+
+
 def test_pinecone_index_upsert_and_delete(mock_describe_index):
     indexer = create_pinecone_indexer()
     indexer._pod_type = "p1"

docs/integrations/destinations/pinecone.md

@@ -76,6 +76,7 @@ OpenAI and Fake embeddings produce vectors with 1536 dimensions, and the Cohere
 
 | Version | Date       | Pull Request                                              | Subject                                                                                                                      |
 | :------ | :--------- | :-------------------------------------------------------- | :--------------------------------------------------------------------------------------------------------------------------- |
+| 0.1.1   | 2023-05-14 | [#38151](https://github.com/airbytehq/airbyte/pull/38151) | Add airbyte source tag for attribution
 | 0.1.0   | 2023-05-06 | [#37756](https://github.com/airbytehq/airbyte/pull/37756) | Add support for Pinecone Serverless                                                                                          |
 | 0.0.24  | 2023-04-15 | [#37333](https://github.com/airbytehq/airbyte/pull/37333) | Update CDK & pytest version to fix security vulnerabilities.                                                                 |
 | 0.0.23  | 2023-03-22 | [#35911](https://github.com/airbytehq/airbyte/pull/35911) | Bump versions to latest, resolves test failures.                                                                             |