airbyte-source-github 1.8.1__tar.gz → 1.8.3__tar.gz

{airbyte_source_github-1.8.1 → airbyte_source_github-1.8.3}/PKG-INFO

@@ -1,18 +1,17 @@
 Metadata-Version: 2.1
 Name: airbyte-source-github
-Version: 1.8.1
+Version: 1.8.3
 Summary: Source implementation for GitHub.
 Home-page: https://airbyte.com
 License: MIT
 Author: Airbyte
 Author-email: contact@airbyte.io
-Requires-Python: >=3.9,<3.12
+Requires-Python: >=3.10,<3.12
 Classifier: License :: OSI Approved :: MIT 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 (>=3,<4)
+Requires-Dist: airbyte-cdk (>=4,<5)
 Requires-Dist: sgqlc (==16.3)
 Project-URL: Documentation, https://docs.airbyte.com/integrations/sources/github
 Project-URL: Repository, https://github.com/airbytehq/airbyte

{airbyte_source_github-1.8.1 → airbyte_source_github-1.8.3}/pyproject.toml

@@ -5,7 +5,7 @@ requires = [
 build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
-version = "1.8.1"
+version = "1.8.3"
 name = "airbyte-source-github"
 description = "Source implementation for GitHub."
 authors = [
@@ -21,8 +21,8 @@ packages = [
 ]
 
 [tool.poetry.dependencies]
-python = "^3.9,<3.12"
-airbyte-cdk = "^3"
+python = "^3.10,<3.12"
+airbyte-cdk = "^4"
 sgqlc = "==16.3"
 
 [tool.poetry.scripts]

{airbyte_source_github-1.8.1 → airbyte_source_github-1.8.3}/source_github/errors_handlers.py

@@ -90,6 +90,15 @@ class GithubStreamABCErrorHandler(HttpStatusErrorHandler):
 
 
 class ContributorActivityErrorHandler(HttpStatusErrorHandler):
+    """
+    This custom error handler is needed for streams based on repository statistics endpoints like ContributorActivity because
+    when requesting data that hasn't been cached yet when the request is made, you'll receive a 202 response. And these requests
+    need to retried to get the actual results.
+
+    See the docs for more info:
+    https://docs.github.com/en/rest/metrics/statistics?apiVersion=2022-11-28#a-word-about-caching
+    """
+
     def interpret_response(self, response_or_exception: Optional[Union[requests.Response, Exception]] = None) -> ErrorResolution:
         if isinstance(response_or_exception, requests.Response) and response_or_exception.status_code == requests.codes.ACCEPTED:
             return ErrorResolution(

{airbyte_source_github-1.8.1 → airbyte_source_github-1.8.3}/source_github/streams.py

@@ -9,10 +9,11 @@ from urllib import parse
 
 import pendulum
 import requests
-from airbyte_cdk import BackoffStrategy
+from airbyte_cdk import BackoffStrategy, StreamSlice
 from airbyte_cdk.models import AirbyteLogMessage, AirbyteMessage, Level, SyncMode
 from airbyte_cdk.models import Type as MessageType
 from airbyte_cdk.sources.streams.availability_strategy import AvailabilityStrategy
+from airbyte_cdk.sources.streams.checkpoint.substream_resumable_full_refresh_cursor import SubstreamResumableFullRefreshCursor
 from airbyte_cdk.sources.streams.core import CheckpointMixin, Stream
 from airbyte_cdk.sources.streams.http import HttpStream
 from airbyte_cdk.sources.streams.http.error_handlers import ErrorHandler, ErrorResolution, HttpStatusErrorHandler, ResponseAction
@@ -57,6 +58,9 @@ class GithubStreamABC(HttpStream, ABC):
         self.api_url = api_url
         self.state = {}
 
+        if not self.supports_incremental:
+            self.cursor = SubstreamResumableFullRefreshCursor()
+
     @property
     def url_base(self) -> str:
         return self.api_url
@@ -1613,7 +1617,8 @@ class ContributorActivity(GithubStream):
         return record
 
     def get_error_handler(self) -> Optional[ErrorHandler]:
-        return ContributorActivityErrorHandler(logger=self.logger, max_retries=self.max_retries, error_mapping=GITHUB_DEFAULT_ERROR_MAPPING)
+
+        return ContributorActivityErrorHandler(logger=self.logger, max_retries=5, error_mapping=GITHUB_DEFAULT_ERROR_MAPPING)
 
     def get_backoff_strategy(self) -> Optional[Union[BackoffStrategy, List[BackoffStrategy]]]:
         return ContributorActivityBackoffStrategy()
@@ -1645,6 +1650,13 @@ class ContributorActivity(GithubStream):
                         message=f"Syncing `{self.__class__.__name__}` " f"stream isn't available for repository `{repository}`.",
                     ),
                 )
+
+                # In order to retain the existing stream behavior before we added RFR to this stream, we need to close out the
+                # partition after we give up the maximum number of retries on the 202 response. This does lead to the question
+                # of if we should prematurely exit in the first place, but for now we're going to aim for feature parity
+                partition_obj = stream_slice.get("partition")
+                if self.cursor and partition_obj:
+                    self.cursor.close_slice(StreamSlice(cursor_slice={}, partition=partition_obj))
             else:
                 raise e