airbyte-source-google-drive 0.3.3__tar.gz → 0.4.0__tar.gz

{airbyte_source_google_drive-0.3.3 → airbyte_source_google_drive-0.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: airbyte-source-google-drive
-Version: 0.3.3
+Version: 0.4.0
 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] (>=6.38.5,<7.0.0)
+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)

{airbyte_source_google_drive-0.3.3 → airbyte_source_google_drive-0.4.0}/pyproject.toml

@@ -5,7 +5,7 @@ requires = [
 build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
-version = "0.3.3"
+version = "0.4.0"
 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 = "^6.38.5"
+version = "^6.45.10"
 
 [tool.poetry.scripts]
 source-google-drive = "source_google_drive.run:run"

{airbyte_source_google_drive-0.3.3 → airbyte_source_google_drive-0.4.0}/source_google_drive/stream_reader.py

@@ -6,27 +6,26 @@
 import io
 import json
 import logging
-import uuid
+import os
 from datetime import datetime
 from io import IOBase
 from os.path import getsize
-from typing import Any, Dict, Iterable, Iterator, List, Optional, Set, Tuple
+from typing import Dict, Iterable, List, Optional, Set, Tuple
 
-import pytz
 from google.oauth2 import credentials, service_account
 from googleapiclient.discovery import build
 from googleapiclient.http import MediaIoBaseDownload
 
 from airbyte_cdk import AirbyteTracedException, FailureType
+from airbyte_cdk.models import AirbyteRecordMessageFileReference
 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.file_record_data import FileRecordData
 from airbyte_cdk.sources.file_based.remote_file import RemoteFile
-from airbyte_cdk.sources.streams.core import package_name_from_class
-from airbyte_cdk.sources.utils.schema_helpers import InternalConfig, ResourceSchemaLoader
 from source_google_drive.utils import get_folder_id
 
 from .exceptions import ErrorDownloadingFile, ErrorFetchingMetadata
-from .spec import RemoteIdentity, RemoteIdentityType, RemotePermissions, SourceGoogleDriveSpec
+from .spec import SourceGoogleDriveSpec
 
 
 FOLDER_MIME_TYPE = "application/vnd.google-apps.folder"
@@ -64,6 +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):
@@ -135,10 +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,
             )
@@ -161,6 +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)
@@ -170,9 +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
@@ -245,7 +259,9 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
         except Exception as e:
             raise ErrorFetchingMetadata(f"An error occurred while retrieving file size: {str(e)}")
 
-    def get_file(self, file: GoogleDriveRemoteFile, local_directory: str, logger: logging.Logger) -> Dict[str, str | int]:
+    def upload(
+        self, file: GoogleDriveRemoteFile, local_directory: str, logger: logging.Logger
+    ) -> Tuple[FileRecordData, AirbyteRecordMessageFileReference]:
         """
         Downloads a file from Google Drive to a specified local directory.
 
@@ -265,15 +281,18 @@ class SourceGoogleDriveStreamReader(AbstractFileBasedStreamReader):
             raise FileSizeLimitError(message=message, internal_message=message, failure_type=FailureType.config_error)
 
         try:
-            file_relative_path, local_file_path, absolute_file_path = 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]
 
             if self._is_exportable_document(file.original_mime_type):
                 request = self.google_drive_service.files().export_media(fileId=file.id, mimeType=file.mime_type)
 
                 file_extension = DOWNLOADABLE_DOCUMENTS_MIME_TYPES[file.original_mime_type][DOCUMENT_FILE_EXTENSION_KEY]
                 local_file_path += file_extension
-                absolute_file_path += file_extension
                 file_relative_path += file_extension
+                file_name += file_extension
             else:
                 request = self.google_drive_service.files().get_media(fileId=file.id)
 
@@ -285,10 +304,26 @@ 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)
 
-            return {"file_url": absolute_file_path, "bytes": file_size, "file_relative_path": file_relative_path}
+            file_record_data = FileRecordData(
+                folder=file_paths[self.FILE_FOLDER],
+                file_name=file_name,
+                bytes=file_size,
+                id=file.id,
+                mime_type=file.mime_type,
+                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,
+                source_file_relative_path=file_relative_path,
+                file_size_bytes=file_size,
+            )
+            return file_record_data, file_reference
 
         except Exception as e:
             raise ErrorDownloadingFile(f"There was an error while trying to download the file {file.uri}: {str(e)}")