airbyte-source-google-drive 0.4.3.dev202504071123__tar.gz → 0.4.4__tar.gz

{airbyte_source_google_drive-0.4.3.dev202504071123 → airbyte_source_google_drive-0.4.4}/PKG-INFO

@@ -1,7 +1,8 @@
-Metadata-Version: 2.3
+Metadata-Version: 2.1
 Name: airbyte-source-google-drive
-Version: 0.4.3.dev202504071123
+Version: 0.4.4
 Summary: Source implementation for Google Drive.
+Home-page: https://airbyte.com
 License: ELv2
 Author: Airbyte
 Author-email: contact@airbyte.io
@@ -10,13 +11,12 @@ Classifier: License :: Other/Proprietary License
 Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
-Requires-Dist: airbyte-cdk[file-based] (==6.45.0.dev04101)
+Requires-Dist: airbyte-cdk[file-based] (>=6.45.10,<7.0.0)
 Requires-Dist: google-api-python-client (==2.104.0)
 Requires-Dist: google-api-python-client-stubs (==1.18.0)
 Requires-Dist: google-auth-httplib2 (==0.1.1)
 Requires-Dist: google-auth-oauthlib (==1.1.0)
 Project-URL: Documentation, https://docs.airbyte.com/integrations/sources/google-drive
-Project-URL: Homepage, https://airbyte.com
 Project-URL: Repository, https://github.com/airbytehq/airbyte
 Description-Content-Type: text/markdown
 

{airbyte_source_google_drive-0.4.3.dev202504071123 → airbyte_source_google_drive-0.4.4}/pyproject.toml

@@ -1,27 +1,19 @@
 [build-system]
-requires = [
-    "poetry-core>=1.0.0",
-]
+requires = [ "poetry-core>=1.0.0",]
 build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
-version = "0.4.3.dev202504071123"
+version = "0.4.4"
 name = "airbyte-source-google-drive"
 description = "Source implementation for Google Drive."
-authors = [
-    "Airbyte <contact@airbyte.io>",
-]
+authors = [ "Airbyte <contact@airbyte.io>",]
 license = "ELv2"
 readme = "README.md"
 documentation = "https://docs.airbyte.com/integrations/sources/google-drive"
 homepage = "https://airbyte.com"
 repository = "https://github.com/airbytehq/airbyte"
-packages = [
-    { include = "source_google_drive" },
-]
-source = [
-    { name = "testpypi", url = "https://test.pypi.org/simple/", priority = "supplemental" },
-]
+[[tool.poetry.packages]]
+include = "source_google_drive"
 
 [tool.poetry.dependencies]
 python = "^3.10,<3.12"
@@ -29,12 +21,8 @@ google-api-python-client = "==2.104.0"
 google-auth-httplib2 = "==0.1.1"
 google-auth-oauthlib = "==1.1.0"
 google-api-python-client-stubs = "==1.18.0"
+airbyte-cdk = {extras = ["file-based"], version = "^6.45.10"}
 
-[tool.poetry.dependencies.airbyte-cdk]
-extras = [
-    "file-based",
-]
-version = "6.45.0.dev04101"
 
 [tool.poetry.scripts]
 source-google-drive = "source_google_drive.run:run"
@@ -43,7 +31,10 @@ source-google-drive = "source_google_drive.run:run"
 pytest-mock = "^3.12.0"
 pytest = "^8.0.0"
 
+
 [tool.poe]
 include = [
+    # Shared tasks definition file(s) can be imported here.
+    # Run `poe` or `poe --help` to see the list of available tasks.
     "${POE_GIT_DIR}/poe-tasks/poetry-connector-tasks.toml",
 ]

{airbyte_source_google_drive-0.4.3.dev202504071123 → airbyte_source_google_drive-0.4.4}/source_google_drive/stream_reader.py

@@ -63,6 +63,16 @@ class GoogleDriveRemoteFile(RemoteFile):
     # The mime type of the file as returned by the Google Drive API
     # This is not the same as the mime type when opened by the parser (e.g. google docs is exported as docx)
     original_mime_type: str
+    view_link: str
+    # Only populated for items in shared drives.
+    drive_id: Optional[str] = None
+    created_at: datetime
+
+    @property
+    def url(self) -> str:
+        if self.drive_id:
+            return f"https://drive.google.com/open?id={self.id}&driveId={self.drive_id}"
+        return self.view_link
 
 
 class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
@@ -134,10 +144,11 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
             (path, folder_id) = folder_id_queue.pop()
             # fetch all files in this folder (1000 is the max page size)
             # supportsAllDrives and includeItemsFromAllDrives are required to access files in shared drives
+            # ref https://developers.google.com/workspace/drive/api/reference/rest/v3/files#File
             request = service.files().list(
                 q=f"'{folder_id}' in parents",
                 pageSize=1000,
-                fields="nextPageToken, files(id, name, modifiedTime, mimeType)",
+                fields="nextPageToken, files(id, name, modifiedTime, mimeType, webViewLink, driveId, createdTime)",
                 supportsAllDrives=True,
                 includeItemsFromAllDrives=True,
             )
@@ -160,6 +171,7 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
                         continue
                     else:
                         last_modified = datetime.strptime(new_file["modifiedTime"], "%Y-%m-%dT%H:%M:%S.%fZ")
+                        created_at = datetime.strptime(new_file["createdTime"], "%Y-%m-%dT%H:%M:%S.%fZ")
                         original_mime_type = new_file["mimeType"]
                         mime_type = (
                             self._get_export_mime_type(original_mime_type)
@@ -169,9 +181,12 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
                         remote_file = GoogleDriveRemoteFile(
                             uri=file_name,
                             last_modified=last_modified,
+                            created_at=created_at,
                             id=new_file["id"],
                             original_mime_type=original_mime_type,
                             mime_type=mime_type,
+                            drive_id=new_file.get("driveId"),
+                            view_link=new_file.get("webViewLink"),
                         )
                         if self.file_matches_globs(remote_file, globs):
                             yield remote_file
@@ -266,7 +281,7 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
             raise FileSizeLimitError(message=message, internal_message=message, failure_type=FailureType.config_error)
 
         try:
-            file_paths = self._get_file_transfer_paths(file, local_directory)
+            file_paths = self._get_file_transfer_paths(source_file_relative_path=file.uri, staging_directory=local_directory)
             local_file_path = file_paths[self.LOCAL_FILE_PATH]
             file_relative_path = file_paths[self.FILE_RELATIVE_PATH]
             file_name = file_paths[self.FILE_NAME]
@@ -289,16 +304,19 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
                     progress = status.resumable_progress / status.total_size * 100 if status.total_size else 0
                     logger.info(f"Processing file {file.uri}, progress: {progress:.2f}%")
 
+            logger.info(f"Finished uploading file {file.uri} to {local_file_path}")
             # native google objects seems to be reporting lower size through the api than the final download size
             file_size = getsize(local_file_path)
 
             file_record_data = FileRecordData(
                 folder=file_paths[self.FILE_FOLDER],
-                filename=file_name,
+                file_name=file_name,
                 bytes=file_size,
                 id=file.id,
                 mime_type=file.mime_type,
-                updated_at=int(file.last_modified.timestamp() * 1000),
+                created_at=file.created_at.strftime("%Y-%m-%dT%H:%M:%S.%fZ"),
+                updated_at=file.last_modified.strftime("%Y-%m-%dT%H:%M:%S.%fZ"),
+                source_uri=file.url,
             )
             file_reference = AirbyteRecordMessageFileReference(
                 staging_file_url=local_file_path,