Source GCS: Fix unstructured format (#34158)

airbytehq · Jan 12, 2024 · 94dc686 · 94dc686
1 parent e2146ea
commit 94dc686
Show file tree

Hide file tree

Showing 5 changed files with 10 additions and 10 deletions.
diff --git a/airbyte-integrations/connectors/source-gcs/integration_tests/spec.json b/airbyte-integrations/connectors/source-gcs/integration_tests/spec.json
@@ -58,9 +58,9 @@
             },
             "primary_key": {
               "title": "Primary Key",
-              "description": "The column or columns (for a composite key) that serves as the unique identifier of a record.",
-              "type": "string",
-              "airbyte_hidden": true
+              "description": "The column or columns (for a composite key) that serves as the unique identifier of a record. If empty, the primary key will default to the parser's default primary key.",
+              "airbyte_hidden": true,
+              "type": "string"
             },
             "days_to_sync_if_history_is_full": {
               "title": "Days To Sync If History Is Full",

diff --git a/airbyte-integrations/connectors/source-gcs/metadata.yaml b/airbyte-integrations/connectors/source-gcs/metadata.yaml
@@ -7,7 +7,7 @@ data:
   connectorSubtype: file
   connectorType: source
   definitionId: 2a8c41ae-8c23-4be0-a73f-2ab10ca1a820
-  dockerImageTag: 0.3.3
+  dockerImageTag: 0.3.4
   dockerRepository: airbyte/source-gcs
   documentationUrl: https://docs.airbyte.com/integrations/sources/gcs
   githubIssueLabel: source-gcs

diff --git a/airbyte-integrations/connectors/source-gcs/source_gcs/config.py b/airbyte-integrations/connectors/source-gcs/source_gcs/config.py
@@ -80,3 +80,7 @@ def replace_enum_allOf_and_anyOf(schema):
             objects_to_check["anyOf"] = objects_to_check.pop("allOf")
 
         return super(Config, Config).replace_enum_allOf_and_anyOf(schema)
+
+    @staticmethod
+    def remove_discriminator(schema) -> None:
+        pass
diff --git a/airbyte-integrations/connectors/source-gcs/source_gcs/stream_reader.py b/airbyte-integrations/connectors/source-gcs/source_gcs/stream_reader.py
@@ -5,7 +5,6 @@
 import itertools
 import json
 import logging
-from contextlib import contextmanager
 from datetime import datetime, timedelta
 from io import IOBase
 from typing import Iterable, List, Optional
@@ -94,7 +93,6 @@ def _handle_file_listing_error(self, exc: Exception, prefix: str, logger: loggin
             prefix=prefix,
         ) from exc
 
-    @contextmanager
     def open_file(self, file: RemoteFile, mode: FileReadMode, encoding: Optional[str], logger: logging.Logger) -> IOBase:
         """
         Open and yield a remote file from GCS for reading.
@@ -105,7 +103,4 @@ def open_file(self, file: RemoteFile, mode: FileReadMode, encoding: Optional[str
         except OSError as oe:
             logger.warning(ERROR_MESSAGE_ACCESS.format(uri=file.uri, bucket=self.config.bucket))
             logger.exception(oe)
-        try:
-            yield result
-        finally:
-            result.close()
+        return result
diff --git a/docs/integrations/sources/gcs.md b/docs/integrations/sources/gcs.md
@@ -37,6 +37,7 @@ Use the service account ID from above, grant read access to your target bucket.
 
 | Version | Date       | Pull Request                                             | Subject                                             |
 |:--------|:-----------|:---------------------------------------------------------|:----------------------------------------------------|
+| 0.3.4   | 2024-01-11 | [34158](https://github.com/airbytehq/airbyte/pull/34158) | Fix issue in stream reader for document file type parser |
 | 0.3.3   | 2023-12-06 | [33187](https://github.com/airbytehq/airbyte/pull/33187) | Bump CDK version to hide source-defined primary key |
 | 0.3.2   | 2023-11-16 | [32608](https://github.com/airbytehq/airbyte/pull/32608) | Improve document file type parser                   |
 | 0.3.1   | 2023-11-13 | [32357](https://github.com/airbytehq/airbyte/pull/32357) | Improve spec schema                                 |