Feat: add datasource Dropbox (#11488)

### What problem does this PR solve?

Add datasource Dropbox.

### Type of change

- [x] New Feature (non-breaking change which adds functionality)
This commit is contained in:
Yongteng Lei
2025-11-25 09:40:03 +08:00
committed by GitHub
parent d5f8548200
commit d1744aaaf3
9 changed files with 197 additions and 37 deletions

View File

@ -119,6 +119,7 @@ class FileSource(StrEnum):
SLACK = "slack"
TEAMS = "teams"
MOODLE = "moodle"
DROPBOX = "dropbox"
class PipelineTaskType(StrEnum):

View File

@ -50,6 +50,7 @@ class DocumentSource(str, Enum):
DISCORD = "discord"
MOODLE = "moodle"
S3_COMPATIBLE = "s3_compatible"
DROPBOX = "dropbox"
class FileOrigin(str, Enum):

View File

@ -1,13 +1,24 @@
"""Dropbox connector"""
import logging
from datetime import timezone
from typing import Any
from dropbox import Dropbox
from dropbox.exceptions import ApiError, AuthError
from dropbox.files import FileMetadata, FolderMetadata
from common.data_source.config import INDEX_BATCH_SIZE
from common.data_source.exceptions import ConnectorValidationError, InsufficientPermissionsError, ConnectorMissingCredentialError
from common.data_source.config import INDEX_BATCH_SIZE, DocumentSource
from common.data_source.exceptions import (
ConnectorMissingCredentialError,
ConnectorValidationError,
InsufficientPermissionsError,
)
from common.data_source.interfaces import LoadConnector, PollConnector, SecondsSinceUnixEpoch
from common.data_source.models import Document, GenerateDocumentsOutput
from common.data_source.utils import get_file_ext
logger = logging.getLogger(__name__)
class DropboxConnector(LoadConnector, PollConnector):
@ -19,29 +30,29 @@ class DropboxConnector(LoadConnector, PollConnector):
def load_credentials(self, credentials: dict[str, Any]) -> dict[str, Any] | None:
"""Load Dropbox credentials"""
try:
access_token = credentials.get("dropbox_access_token")
if not access_token:
raise ConnectorMissingCredentialError("Dropbox access token is required")
access_token = credentials.get("dropbox_access_token")
if not access_token:
raise ConnectorMissingCredentialError("Dropbox access token is required")
self.dropbox_client = Dropbox(access_token)
return None
except Exception as e:
raise ConnectorMissingCredentialError(f"Dropbox: {e}")
self.dropbox_client = Dropbox(access_token)
return None
def validate_connector_settings(self) -> None:
"""Validate Dropbox connector settings"""
if not self.dropbox_client:
if self.dropbox_client is None:
raise ConnectorMissingCredentialError("Dropbox")
try:
# Test connection by getting current account info
self.dropbox_client.users_get_current_account()
except (AuthError, ApiError) as e:
if "invalid_access_token" in str(e).lower():
raise InsufficientPermissionsError("Invalid Dropbox access token")
else:
raise ConnectorValidationError(f"Dropbox validation error: {e}")
self.dropbox_client.files_list_folder(path="", limit=1)
except AuthError as e:
logger.exception("[Dropbox]: Failed to validate Dropbox credentials")
raise ConnectorValidationError(f"Dropbox credential is invalid: {e}")
except ApiError as e:
if e.error is not None and "insufficient_permissions" in str(e.error).lower():
raise InsufficientPermissionsError("Your Dropbox token does not have sufficient permissions.")
raise ConnectorValidationError(f"Unexpected Dropbox error during validation: {e.user_message_text or e}")
except Exception as e:
raise ConnectorValidationError(f"Unexpected error during Dropbox settings validation: {e}")
def _download_file(self, path: str) -> bytes:
"""Download a single file from Dropbox."""
@ -56,24 +67,103 @@ class DropboxConnector(LoadConnector, PollConnector):
raise ConnectorMissingCredentialError("Dropbox")
try:
# Try to get existing shared links first
shared_links = self.dropbox_client.sharing_list_shared_links(path=path)
if shared_links.links:
return shared_links.links[0].url
# Create a new shared link
link_settings = self.dropbox_client.sharing_create_shared_link_with_settings(path)
return link_settings.url
except Exception:
# Fallback to basic link format
return f"https://www.dropbox.com/home{path}"
link_metadata = self.dropbox_client.sharing_create_shared_link_with_settings(path)
return link_metadata.url
except ApiError as err:
logger.exception(f"[Dropbox]: Failed to create a shared link for {path}: {err}")
return ""
def poll_source(self, start: SecondsSinceUnixEpoch, end: SecondsSinceUnixEpoch) -> Any:
def _yield_files_recursive(
self,
path: str,
start: SecondsSinceUnixEpoch | None,
end: SecondsSinceUnixEpoch | None,
) -> GenerateDocumentsOutput:
"""Yield files in batches from a specified Dropbox folder, including subfolders."""
if self.dropbox_client is None:
raise ConnectorMissingCredentialError("Dropbox")
result = self.dropbox_client.files_list_folder(
path,
limit=self.batch_size,
recursive=False,
include_non_downloadable_files=False,
)
while True:
batch: list[Document] = []
for entry in result.entries:
if isinstance(entry, FileMetadata):
modified_time = entry.client_modified
if modified_time.tzinfo is None:
modified_time = modified_time.replace(tzinfo=timezone.utc)
else:
modified_time = modified_time.astimezone(timezone.utc)
time_as_seconds = modified_time.timestamp()
if start is not None and time_as_seconds <= start:
continue
if end is not None and time_as_seconds > end:
continue
try:
downloaded_file = self._download_file(entry.path_display)
except Exception:
logger.exception(f"[Dropbox]: Error downloading file {entry.path_display}")
continue
batch.append(
Document(
id=f"dropbox:{entry.id}",
blob=downloaded_file,
source=DocumentSource.DROPBOX,
semantic_identifier=entry.name,
extension=get_file_ext(entry.name),
doc_updated_at=modified_time,
size_bytes=entry.size if getattr(entry, "size", None) is not None else len(downloaded_file),
)
)
elif isinstance(entry, FolderMetadata):
yield from self._yield_files_recursive(entry.path_lower, start, end)
if batch:
yield batch
if not result.has_more:
break
result = self.dropbox_client.files_list_folder_continue(result.cursor)
def poll_source(self, start: SecondsSinceUnixEpoch, end: SecondsSinceUnixEpoch) -> GenerateDocumentsOutput:
"""Poll Dropbox for recent file changes"""
# Simplified implementation - in production this would handle actual polling
return []
if self.dropbox_client is None:
raise ConnectorMissingCredentialError("Dropbox")
def load_from_state(self) -> Any:
for batch in self._yield_files_recursive("", start, end):
yield batch
def load_from_state(self) -> GenerateDocumentsOutput:
"""Load files from Dropbox state"""
# Simplified implementation
return []
return self._yield_files_recursive("", None, None)
if __name__ == "__main__":
import os
logging.basicConfig(level=logging.DEBUG)
connector = DropboxConnector()
connector.load_credentials({"dropbox_access_token": os.environ.get("DROPBOX_ACCESS_TOKEN")})
connector.validate_connector_settings()
document_batches = connector.load_from_state()
try:
first_batch = next(document_batches)
print(f"Loaded {len(first_batch)} documents in first batch.")
for doc in first_batch:
print(f"- {doc.semantic_identifier} ({doc.size_bytes} bytes)")
except StopIteration:
print("No documents available in Dropbox.")

View File

@ -37,7 +37,7 @@ from api.db.services.connector_service import ConnectorService, SyncLogsService
from api.db.services.knowledgebase_service import KnowledgebaseService
from common import settings
from common.config_utils import show_configs
from common.data_source import BlobStorageConnector, NotionConnector, DiscordConnector, GoogleDriveConnector, MoodleConnector, JiraConnector
from common.data_source import BlobStorageConnector, NotionConnector, DiscordConnector, GoogleDriveConnector, MoodleConnector, JiraConnector, DropboxConnector
from common.constants import FileSource, TaskStatus
from common.data_source.config import INDEX_BATCH_SIZE
from common.data_source.confluence_connector import ConfluenceConnector
@ -211,6 +211,27 @@ class Gmail(SyncBase):
pass
class Dropbox(SyncBase):
SOURCE_NAME: str = FileSource.DROPBOX
async def _generate(self, task: dict):
self.connector = DropboxConnector(batch_size=self.conf.get("batch_size", INDEX_BATCH_SIZE))
self.connector.load_credentials(self.conf["credentials"])
if task["reindex"] == "1" or not task["poll_range_start"]:
document_generator = self.connector.load_from_state()
begin_info = "totally"
else:
poll_start = task["poll_range_start"]
document_generator = self.connector.poll_source(
poll_start.timestamp(), datetime.now(timezone.utc).timestamp()
)
begin_info = f"from {poll_start}"
logging.info(f"[Dropbox] Connect to Dropbox {begin_info}")
return document_generator
class GoogleDrive(SyncBase):
SOURCE_NAME: str = FileSource.GOOGLE_DRIVE
@ -454,7 +475,8 @@ func_factory = {
FileSource.SHAREPOINT: SharePoint,
FileSource.SLACK: Slack,
FileSource.TEAMS: Teams,
FileSource.MOODLE: Moodle
FileSource.MOODLE: Moodle,
FileSource.DROPBOX: Dropbox,
}

View File

@ -0,0 +1 @@
<svg xmlns="http://www.w3.org/2000/svg" viewBox="89.9 347.3 32 32" width="64" height="64" fill="#007ee5"><path d="M99.337 348.42L89.9 354.5l6.533 5.263 9.467-5.837m-16 11l9.437 6.2 6.563-5.505-9.467-5.868m9.467 5.868l6.594 5.505 9.406-6.14-6.503-5.233m6.503-5.203l-9.406-6.14-6.594 5.505 9.497 5.837m-9.467 7.047l-6.594 5.474-2.843-1.845v2.087l9.437 5.656 9.437-5.656v-2.087l-2.843 1.845"/></svg>

After

Width:  |  Height:  |  Size: 396 B

View File

@ -742,6 +742,10 @@ Example: https://fsn1.your-objectstorage.com`,
'Comma-separated emails whose "My Drive" contents should be indexed (include the primary admin).',
google_driveSharedFoldersTip:
'Comma-separated Google Drive folder links to crawl.',
dropboxDescription:
'Connect your Dropbox to sync files and folders from a chosen account.',
dropboxAccessTokenTip:
'Generate a long-lived access token in the Dropbox App Console with files.metadata.read, files.content.read, and sharing.read scopes.',
moodleDescription:
'Connect to your Moodle LMS to sync course content, forums, and resources.',
moodleUrlTip:

View File

@ -722,6 +722,9 @@ General实体和关系提取提示来自 GitHub - microsoft/graphrag基于
'需要索引其 “我的云端硬盘” 的邮箱,多个邮箱用逗号分隔(建议包含管理员)。',
google_driveSharedFoldersTip:
'需要同步的 Google Drive 文件夹链接,多个链接用逗号分隔。',
dropboxDescription: '连接 Dropbox同步指定账号下的文件与文件夹。',
dropboxAccessTokenTip:
'请在 Dropbox App Console 生成 Access Token并勾选 files.metadata.read、files.content.read、sharing.read 等必要权限。',
jiraDescription: '接入 Jira 工作区持续同步Issues、评论与附件。',
jiraBaseUrlTip:
'Jira 的 Base URL例如https://your-domain.atlassian.net。',

View File

@ -12,6 +12,7 @@ export enum DataSourceKey {
MOODLE = 'moodle',
// GMAIL = 'gmail',
JIRA = 'jira',
DROPBOX = 'dropbox',
// SHAREPOINT = 'sharepoint',
// SLACK = 'slack',
// TEAMS = 'teams',
@ -53,6 +54,11 @@ export const DataSourceInfo = {
description: t(`setting.${DataSourceKey.JIRA}Description`),
icon: <SvgIcon name={'data-source/jira'} width={38} />,
},
[DataSourceKey.DROPBOX]: {
name: 'Dropbox',
description: t(`setting.${DataSourceKey.DROPBOX}Description`),
icon: <SvgIcon name={'data-source/dropbox'} width={38} />,
},
};
export const DataSourceFormBaseFields = [
@ -408,6 +414,22 @@ export const DataSourceFormFields = {
tooltip: t('setting.jiraPasswordTip'),
},
],
[DataSourceKey.DROPBOX]: [
{
label: 'Access Token',
name: 'config.credentials.dropbox_access_token',
type: FormFieldType.Password,
required: true,
tooltip: t('setting.dropboxAccessTokenTip'),
},
{
label: 'Batch Size',
name: 'config.batch_size',
type: FormFieldType.Number,
required: false,
placeholder: 'Defaults to 2',
},
],
};
export const DataSourceFormDefaultValues = {
@ -508,4 +530,14 @@ export const DataSourceFormDefaultValues = {
},
},
},
[DataSourceKey.DROPBOX]: {
name: '',
source: DataSourceKey.DROPBOX,
config: {
batch_size: 2,
credentials: {
dropbox_access_token: '',
},
},
},
};

View File

@ -56,6 +56,12 @@ const dataSourceTemplates = [
description: DataSourceInfo[DataSourceKey.JIRA].description,
icon: DataSourceInfo[DataSourceKey.JIRA].icon,
},
{
id: DataSourceKey.DROPBOX,
name: DataSourceInfo[DataSourceKey.DROPBOX].name,
description: DataSourceInfo[DataSourceKey.DROPBOX].description,
icon: DataSourceInfo[DataSourceKey.DROPBOX].icon,
},
];
const DataSource = () => {
const { t } = useTranslation();