airbyte-source-google-search-console 1.6.0.dev202503282247__py3-none-any.whl → 1.7.0__py3-none-any.whl

{airbyte_source_google_search_console-1.6.0.dev202503282247.dist-info → airbyte_source_google_search_console-1.7.0.dist-info}/METADATA

@@ -1,19 +1,19 @@
 Metadata-Version: 2.3
 Name: airbyte-source-google-search-console
-Version: 1.6.0.dev202503282247
+Version: 1.7.0
 Summary: Source implementation for Google Search Console.
 License: Elv2
 Author: Airbyte
 Author-email: contact@airbyte.io
-Requires-Python: >=3.9,<3.12
+Requires-Python: >=3.10,<3.12
 Classifier: License :: Other/Proprietary License
 Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
-Requires-Dist: airbyte-cdk (>=1,<2)
+Requires-Dist: airbyte-cdk (>=6,<7)
 Requires-Dist: google-api-python-client (==2.105.0)
 Requires-Dist: google-auth (==2.23.3)
+Requires-Dist: pendulum (>=3.1.0,<4.0.0)
 Project-URL: Documentation, https://docs.airbyte.com/integrations/sources/google-search-console
 Project-URL: Homepage, https://airbyte.com
 Project-URL: Repository, https://github.com/airbytehq/airbyte

airbyte_source_google_search_console-1.7.0.dist-info/RECORD

@@ -0,0 +1,22 @@
+source_google_search_console/__init__.py,sha256=HQCPu-CK7XmVDtP9rmTdB2XyraVCc6pv9pw38-O8y48,1191
+source_google_search_console/components.py,sha256=5o8kH2xwYUvk3yjnSd6okJVF2KBUqUIF4V97xRpPAyI,2803
+source_google_search_console/config_migrations.py,sha256=Cl4SUdJpAf6wMM_vVhqjjU89NfUq9LIGJ9zNrWiBk-A,4235
+source_google_search_console/exceptions.py,sha256=iD3jYC4WxVCEKGsqQ7Vaj1tbjhJZ4S5mnSDnwFJdsIQ,1097
+source_google_search_console/manifest.yaml,sha256=m6IhF4bRwF9nAivJev2ebB8QPvoKRbYixF-5ZKSBkms,25882
+source_google_search_console/run.py,sha256=TBkPlseTERarkj6wL8AMEKgm5Xsb2drnltPVH6257-M,2195
+source_google_search_console/schemas/search_analytics_all_fields.json,sha256=iQxRh_c_yz3uGofqpo1KX571TMmzYjKScb0PtI6SN_Q,1729
+source_google_search_console/schemas/search_analytics_by_date.json,sha256=meCbWDayc1y0q-Lu-CAdjQVnsM8xZBX3BdF129UC1P8,1388
+source_google_search_console/schemas/search_analytics_by_device.json,sha256=VtoFjmmv9rx-uhSFaRn0wm4LeSxRIaexrxg2Spvbneo,1525
+source_google_search_console/schemas/search_analytics_by_page.json,sha256=KyUojZc4Lv3hPswxIJzUL5QDNsbvSugGjl_uHGF7Am4,1473
+source_google_search_console/schemas/search_analytics_by_query.json,sha256=mYc1Fu1A5TWLZCApZSjjfs_urW5BRUCPQI3rw3yQjt4,1439
+source_google_search_console/schemas/search_analytics_page_report.json,sha256=-b0Y0LenTchS0q9A2aQ4hIjUjXkYF8erOtyrTMhf6MM,1776
+source_google_search_console/schemas/search_analytics_site_report_by_page.json,sha256=hWKHkm1reqGGu1dNcWBe6_XkZ5tK-UaiymrYRVgxRxI,1515
+source_google_search_console/schemas/search_analytics_site_report_by_site.json,sha256=rAh6LuNy7nCrrNM9MTd0qxAVc886ecQaqWRgV63OfyA,1408
+source_google_search_console/service_account_authenticator.py,sha256=pAWKAXfwfTY3xkXvQJH0EyFphFULdCIcC47YXYTO9X8,1307
+source_google_search_console/source.py,sha256=ddqYq3ZSNrVtGvuQQ5WnHgMLuf92AtumzdDKWGffwRo,8721
+source_google_search_console/spec.json,sha256=SensIrWvoie2aGwxmgoMYpq2jb3XP3F2myQWCVQ4db8,8980
+source_google_search_console/streams.py,sha256=yb78hGaKPDCJoFqPSun3lDhr8VhSxcLgXTLtX2ouVpY,15804
+airbyte_source_google_search_console-1.7.0.dist-info/METADATA,sha256=R-pJQZUfWv2duwzSFWta7vbn8V01TeESsaoRO5728mo,5621
+airbyte_source_google_search_console-1.7.0.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
+airbyte_source_google_search_console-1.7.0.dist-info/entry_points.txt,sha256=DMcgc9bCX-Vt6hm_68pa77qS3eGdeMhg-UdlFc-XKUM,85
+airbyte_source_google_search_console-1.7.0.dist-info/RECORD,,

{airbyte_source_google_search_console-1.6.0.dev202503282247.dist-info → airbyte_source_google_search_console-1.7.0.dist-info}/WHEEL

@@ -1,4 +1,4 @@
 Wheel-Version: 1.0
-Generator: poetry-core 2.1.1
+Generator: poetry-core 2.1.3
 Root-Is-Purelib: true
 Tag: py3-none-any

source_google_search_console/components.py

@@ -0,0 +1,81 @@
+#
+# Copyright (c) 2025 Airbyte, Inc., all rights reserved.
+#
+
+from dataclasses import dataclass
+from typing import Any, Mapping, Optional
+
+from airbyte_cdk.sources.declarative.migrations.state_migration import StateMigration
+
+
+@dataclass
+class NestedSubstreamStateMigration(StateMigration):
+    """
+    We require a custom state migration because SearchAnalytics streams contain two nested levels of
+    substreams. The existing LegacyToPerPartitionStateMigration only handles one level.
+
+    Legacy state format is as follows:
+    {
+        "date": "2025-05-28",
+        "https://www.example.com/": {
+            "web": {
+                "date": "2025-05-25"
+            },
+            "news": {
+                "date": "2023-05-22"
+            }
+        }
+    }
+
+    The resulting migrated per-partition state is:
+    {
+        "use_global_cursor": false,
+        "states": [
+        {
+            "partition": {
+                "search_type": "web",
+                "site_url": "https://www.example.com/"
+            },
+            "cursor": {
+                "date": "2025-05-25"
+            }
+        },
+        {
+            "partition": {
+                "search_type": "news",
+                "site_url": "https://www.example.com/"
+            },
+            "cursor": {
+                "date": "2023-05-22"
+            }
+        }],
+        "state": {
+            "date": "2025-05-25"
+        }
+    }
+    """
+
+    def should_migrate(self, stream_state: Mapping[str, Any]) -> bool:
+        return len(stream_state) > 0 and "states" not in stream_state
+
+    def migrate(self, stream_state: Mapping[str, Any]) -> Mapping[str, Any]:
+        global_state: Optional[Mapping[str, Any]] = None
+        per_partition_state = []
+        for site_url_key, search_type_state in stream_state.items():
+            if site_url_key == "date":
+                # The legacy state also contains a global cursor value under the `date` key which equates
+                # to global state.
+                #
+                # However, the Python implementation does not appear to be implemented
+                # correctly and simply saves the state of the last seen partition. Since I don't trust the
+                # legacy value and in the current implementation global state is applied to partitions
+                # without an existing value, I'm making a conscious choice to not migrate the global value.
+                continue
+            else:
+                site_url = site_url_key
+                for search_type_key, cursor in search_type_state.items():
+                    per_partition_state.append({"partition": {"site_url": site_url, "search_type": search_type_key}, "cursor": cursor})
+        return {
+            "use_global_cursor": False,
+            "states": per_partition_state,
+        }

source_google_search_console/config_migrations.py

@@ -6,8 +6,11 @@
 import logging
 from typing import Any, List, Mapping
 
+import orjson
+
 from airbyte_cdk.config_observation import create_connector_config_control_message
 from airbyte_cdk.entrypoint import AirbyteEntrypoint
+from airbyte_cdk.models import AirbyteMessageSerializer
 from airbyte_cdk.sources import Source
 from airbyte_cdk.sources.message import InMemoryMessageRepository, MessageRepository
 
@@ -77,7 +80,7 @@ class MigrateCustomReports:
         cls.message_repository.emit_message(create_connector_config_control_message(migrated_config))
         # emit the Airbyte Control Message from message queue to stdout
         for message in cls.message_repository._message_queue:
-            print(message.json(exclude_unset=True))
+            print(orjson.dumps(AirbyteMessageSerializer.dump(message)).decode())
 
     @classmethod
     def migrate(cls, args: List[str], source: Source) -> None: