airbyte-source-google-drive 0.4.0.dev202504091818__py3-none-any.whl → 0.4.0.dev202504161719__py3-none-any.whl

{airbyte_source_google_drive-0.4.0.dev202504091818.dist-info → airbyte_source_google_drive-0.4.0.dev202504161719.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: airbyte-source-google-drive
-Version: 0.4.0.dev202504091818
+Version: 0.4.0.dev202504161719
 Summary: Source implementation for Google Drive.
 License: ELv2
 Author: Airbyte
@@ -10,7 +10,7 @@ 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.dev04107)
+Requires-Dist: airbyte-cdk[file-based] (==6.45.4.post49.dev14495925594)
 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)

{airbyte_source_google_drive-0.4.0.dev202504091818.dist-info → airbyte_source_google_drive-0.4.0.dev202504161719.dist-info}/RECORD

@@ -6,9 +6,9 @@ source_google_drive/schemas/identities.json,sha256=JXBR_v0wpfDKiWVzLoc8bPs33x5CG
 source_google_drive/source.py,sha256=Hmz60uMCQZyA9AhCN0bOUSW2iL3GW6VTwI5vrK0D1RY,4140
 source_google_drive/spec.py,sha256=-WkA2zGuQtf3G7uK8uq9BnimUlQh0s3vsqROmIHOgzI,4718
 source_google_drive/stream_permissions_reader.py,sha256=88bSJBC5n2svioTDwgVqf0x5yXRxityk_i0xokWEJcQ,9026
-source_google_drive/stream_reader.py,sha256=r7eSUJ1CTgaZsMg3TljUDjSNidTyAyfRGfv2vxhLdlQ,14909
+source_google_drive/stream_reader.py,sha256=L9NwJWsPspY0_2w3ipRcYU6dH_xqCysc_mMylX09b4E,15531
 source_google_drive/utils.py,sha256=ewR-kBKLmtD-s7zqCfGECfzWYF43tpQdscAQIlUEkR8,1022
-airbyte_source_google_drive-0.4.0.dev202504091818.dist-info/METADATA,sha256=3kVi44z3GI6xpjSQPUhi44MAUSk6FoldVXQlYvhoK5w,5536
-airbyte_source_google_drive-0.4.0.dev202504091818.dist-info/WHEEL,sha256=fGIA9gx4Qxk2KDKeNJCbOEwSrmLtjWCwzBz351GyrPQ,88
-airbyte_source_google_drive-0.4.0.dev202504091818.dist-info/entry_points.txt,sha256=YgpJf0nA5Mn0B7YC9VOFI847vz1jI6U4q7BeLUOXa54,67
-airbyte_source_google_drive-0.4.0.dev202504091818.dist-info/RECORD,,
+airbyte_source_google_drive-0.4.0.dev202504161719.dist-info/METADATA,sha256=3dZkqQBqApBxzvVb8BhMfh8uWTT-aB-iqveVw4i-_d8,5549
+airbyte_source_google_drive-0.4.0.dev202504161719.dist-info/WHEEL,sha256=fGIA9gx4Qxk2KDKeNJCbOEwSrmLtjWCwzBz351GyrPQ,88
+airbyte_source_google_drive-0.4.0.dev202504161719.dist-info/entry_points.txt,sha256=YgpJf0nA5Mn0B7YC9VOFI847vz1jI6U4q7BeLUOXa54,67
+airbyte_source_google_drive-0.4.0.dev202504161719.dist-info/RECORD,,

source_google_drive/stream_reader.py

@@ -63,6 +63,15 @@ 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
+
+    @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 +143,11 @@ 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)",
                 supportsAllDrives=True,
                 includeItemsFromAllDrives=True,
             )
@@ -172,6 +182,8 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
                             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 +278,7 @@ 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]
@@ -300,6 +312,7 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
                 id=file.id,
                 mime_type=file.mime_type,
                 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,