airbyte-source-google-drive 0.2.0.dev202502031727__tar.gz → 0.2.0.dev202502032236__tar.gz

{airbyte_source_google_drive-0.2.0.dev202502031727 → airbyte_source_google_drive-0.2.0.dev202502032236}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: airbyte-source-google-drive
-Version: 0.2.0.dev202502031727
+Version: 0.2.0.dev202502032236
 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] (==6.26.0.dev04106)
+Requires-Dist: airbyte-cdk[file-based] (==6.26.0.dev04107)
 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-0.2.0.dev202502031727 → airbyte_source_google_drive-0.2.0.dev202502032236}/pyproject.toml

@@ -5,7 +5,7 @@ requires = [
 build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
-version = "0.2.0.dev202502031727"
+version = "0.2.0.dev202502032236"
 name = "airbyte-source-google-drive"
 description = "Source implementation for Google Drive."
 authors = [
@@ -31,7 +31,7 @@ google-api-python-client-stubs = "==1.18.0"
 extras = [
     "file-based",
 ]
-version = "6.26.0.dev04106"
+version = "6.26.0.dev04107"
 
 [tool.poetry.scripts]
 source-google-drive = "source_google_drive.run:run"

{airbyte_source_google_drive-0.2.0.dev202502031727 → airbyte_source_google_drive-0.2.0.dev202502032236}/source_google_drive/spec.py

@@ -1,7 +1,9 @@
 #
 # Copyright (c) 2023 Airbyte, Inc., all rights reserved.
 #
-
+import uuid
+from datetime import datetime
+from enum import Enum
 from typing import Any, Dict, Literal, Optional, Union
 
 import dpath.util
@@ -11,22 +13,40 @@ from airbyte_cdk import OneOfOptionConfig
 from airbyte_cdk.sources.file_based.config.abstract_file_based_spec import (
     AbstractFileBasedSpec,
     DeliverRawFiles,
+    DeliverRecords,
 )
 from airbyte_cdk.sources.file_based.config.abstract_file_based_spec import (
-    DeliverRecords as DeliverRecordsBase,
+    DeliverPermissions as DeliverPermissionsBase,
 )
 
 
-class DeliverRecords(DeliverRecordsBase):
-    # Overriding to make visible with airbyte_hidden=False
-    sync_acl_permissions: bool = Field(
-        title="Include ACL Permissions",
-        description="Joins Document allowlists to each stream.",
-        default=False,
-        airbyte_hidden=False,
-        order=0,
-    )
-    domain: Optional[str] = Field(title="Domain", description="The domain of the identities.", airbyte_hidden=False, order=1)
+class RemoteIdentityType(Enum):
+    USER = "user"
+    GROUP = "group"
+
+
+class RemoteIdentity(BaseModel):
+    id: uuid.UUID
+    remote_id: str
+    parent_id: str | None = None
+    name: str | None = None
+    description: str | None = None
+    email_address: str | None = None
+    member_email_addresses: list[str] | None = None
+    type: RemoteIdentityType
+    modified_at: datetime
+
+
+class RemotePermissions(BaseModel):
+    id: str
+    file_path: str
+    allowed_identity_remote_ids: list[str] | None = None
+    denied_identity_remote_ids: list[str] | None = None
+    publicly_accessible: bool = False
+
+
+class DeliverPermissions(DeliverPermissionsBase):
+    domain: Optional[str] = Field(title="Domain", description="The Google domain of the identities.", airbyte_hidden=False, order=1)
 
 
 class OAuthCredentials(BaseModel):
@@ -77,7 +97,7 @@ class SourceGoogleDriveSpec(AbstractFileBasedSpec, BaseModel):
         pattern_descriptor="https://drive.google.com/drive/folders/MY-FOLDER-ID",
     )
 
-    delivery_method: DeliverRecords | DeliverRawFiles = Field(
+    delivery_method: DeliverRecords | DeliverRawFiles | DeliverPermissions = Field(
         title="Delivery Method",
         discriminator="delivery_type",
         type="object",

{airbyte_source_google_drive-0.2.0.dev202502031727 → airbyte_source_google_drive-0.2.0.dev202502032236}/source_google_drive/stream_reader.py

@@ -18,14 +18,13 @@ from googleapiclient.discovery import build
 from googleapiclient.http import MediaIoBaseDownload
 
 from airbyte_cdk import AirbyteTracedException, FailureType
-from airbyte_cdk.sources.file_based.config.permissions import RemoteFileIdentity, RemoteFileIdentityType, RemoteFilePermissions
 from airbyte_cdk.sources.file_based.exceptions import FileSizeLimitError
 from airbyte_cdk.sources.file_based.file_based_stream_reader import AbstractFileBasedStreamReader, FileReadMode
 from airbyte_cdk.sources.file_based.remote_file import RemoteFile
 from source_google_drive.utils import get_folder_id
 
 from .exceptions import ErrorDownloadingFile, ErrorFetchingMetadata
-from .spec import SourceGoogleDriveSpec
+from .spec import RemoteIdentity, RemoteIdentityType, RemotePermissions, SourceGoogleDriveSpec
 
 
 FOLDER_MIME_TYPE = "application/vnd.google-apps.folder"
@@ -122,7 +121,7 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
                 if self.config.credentials.auth_type == "Client":
                     creds = credentials.Credentials.from_authorized_user_info(self.config.credentials.dict())
                 else:
-                    scopes = PERMISSIONS_API_SCOPES if self.sync_acl_permissions() else None
+                    scopes = PERMISSIONS_API_SCOPES if self.include_identities_stream() else None
                     creds = service_account.Credentials.from_service_account_info(
                         json.loads(self.config.credentials.service_account_info), scopes=scopes
                     )
@@ -310,7 +309,7 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
         except Exception as e:
             raise ErrorDownloadingFile(f"There was an error while trying to download the file {file.uri}: {str(e)}")
 
-    def get_file_permissions(self, file_id: str, file_name: str, logger: logging.Logger) -> Tuple[List[RemoteFileIdentity], bool]:
+    def get_file_permissions(self, file_id: str, file_name: str, logger: logging.Logger) -> Tuple[List[RemoteIdentity], bool]:
         """
         Retrieves the permissions of a file in Google Drive and checks for public access.
 
@@ -346,13 +345,13 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
         except Exception as e:
             raise ErrorFetchingMetadata(f"An error occurred while retrieving file permissions: {str(e)}")
 
-    def _to_remote_file_identity(self, identity: dict[str, Any]) -> RemoteFileIdentity | None:
+    def _to_remote_file_identity(self, identity: dict[str, Any]) -> RemoteIdentity | None:
         if identity.get("id") in PUBLIC_PERMISSION_IDS:
             return None
         if identity.get("deleted") is True:
             return None
 
-        return RemoteFileIdentity(
+        return RemoteIdentity(
             modified_at=datetime.now(),
             id=uuid.uuid4(),
             remote_id=identity.get("emailAddress"),
@@ -364,7 +363,7 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
 
     def get_file_acl_permissions(self, file: GoogleDriveRemoteFile, logger: logging.Logger) -> Dict[str, Any]:
         remote_identities, is_public = self.get_file_permissions(file.id, file_name=file.uri, logger=logger)
-        return RemoteFilePermissions(
+        return RemotePermissions(
             id=file.id,
             file_path=file.uri,
             allowed_identity_remote_ids=[p.remote_id for p in remote_identities],
@@ -410,24 +409,24 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
         members_api = directory_service.members()
 
         for user in self._get_looping_google_api_list_response(users_api, "users", {"domain": domain}, logger):
-            rfp = RemoteFileIdentity(
+            rfp = RemoteIdentity(
                 id=uuid.uuid4(),
                 remote_id=user["primaryEmail"],
                 name=user["name"]["fullName"] if user["name"] is not None else None,
                 email_address=user["primaryEmail"],
                 member_email_addresses=[x["address"] for x in user["emails"]],
-                type=RemoteFileIdentityType.USER,
+                type=RemoteIdentityType.USER,
                 modified_at=datetime_now(),
             )
             yield rfp.dict()
 
         for group in self._get_looping_google_api_list_response(groups_api, "groups", {"domain": domain}, logger):
-            rfp = RemoteFileIdentity(
+            rfp = RemoteIdentity(
                 id=uuid.uuid4(),
                 remote_id=group["email"],
                 name=group["name"],
                 email_address=group["email"],
-                type=RemoteFileIdentityType.GROUP,
+                type=RemoteIdentityType.GROUP,
                 modified_at=datetime_now(),
             )