Compare commits

...

3 Commits

Author SHA1 Message Date
pablonyx
a473f5b0d5 simplify slack check 2025-02-19 11:36:00 -08:00
pablonyx
601a42575d add slack validation 2025-02-19 11:35:12 -08:00
pablonyx
2e462af29b k 2025-02-19 11:35:12 -08:00
2 changed files with 120 additions and 0 deletions

View File

@@ -7,15 +7,22 @@ from typing import Optional
import boto3 # type: ignore
from botocore.client import Config # type: ignore
from botocore.exceptions import ClientError
from botocore.exceptions import NoCredentialsError
from botocore.exceptions import PartialCredentialsError
from mypy_boto3_s3 import S3Client # type: ignore
from onyx.configs.app_configs import INDEX_BATCH_SIZE
from onyx.configs.constants import BlobType
from onyx.configs.constants import DocumentSource
from onyx.connectors.interfaces import ConnectorValidationError
from onyx.connectors.interfaces import CredentialExpiredError
from onyx.connectors.interfaces import GenerateDocumentsOutput
from onyx.connectors.interfaces import InsufficientPermissionsError
from onyx.connectors.interfaces import LoadConnector
from onyx.connectors.interfaces import PollConnector
from onyx.connectors.interfaces import SecondsSinceUnixEpoch
from onyx.connectors.interfaces import UnexpectedError
from onyx.connectors.models import ConnectorMissingCredentialError
from onyx.connectors.models import Document
from onyx.connectors.models import Section
@@ -240,6 +247,73 @@ class BlobStorageConnector(LoadConnector, PollConnector):
return None
def validate_connector_settings(self) -> None:
if self.s3_client is None:
raise ConnectorMissingCredentialError(
"Blob storage credentials not loaded."
)
if not self.bucket_name:
raise ConnectorValidationError(
"No bucket name was provided in connector settings."
)
try:
# We only fetch one object/page as a light-weight validation step.
# This ensures we trigger typical S3 permission checks (ListObjectsV2, etc.).
self.s3_client.list_objects_v2(
Bucket=self.bucket_name, Prefix=self.prefix, MaxKeys=1
)
except NoCredentialsError:
raise ConnectorMissingCredentialError(
"No valid blob storage credentials found or provided to boto3."
)
except PartialCredentialsError:
raise ConnectorMissingCredentialError(
"Partial or incomplete blob storage credentials provided to boto3."
)
except ClientError as e:
error_code = e.response["Error"].get("Code", "")
status_code = e.response["ResponseMetadata"].get("HTTPStatusCode")
# Most common S3 error cases
if error_code in [
"AccessDenied",
"InvalidAccessKeyId",
"SignatureDoesNotMatch",
]:
if status_code == 403 or error_code == "AccessDenied":
raise InsufficientPermissionsError(
f"Insufficient permissions to list objects in bucket '{self.bucket_name}'. "
"Please check your bucket policy and/or IAM policy."
)
if status_code == 401 or error_code == "SignatureDoesNotMatch":
raise CredentialExpiredError(
"Provided blob storage credentials appear invalid or expired."
)
raise CredentialExpiredError(
f"Credential issue encountered ({error_code})."
)
if error_code == "NoSuchBucket" or status_code == 404:
raise ConnectorValidationError(
f"Bucket '{self.bucket_name}' does not exist or cannot be found."
)
raise ConnectorValidationError(
f"Unexpected S3 client error (code={error_code}, status={status_code}): {e}"
)
except Exception as e:
# Catch-all for anything not captured by the above
# Since we are unsure of the error and it may not disable the connector,
# raise an unexpected error (does not disable connector)
raise UnexpectedError(
f"Unexpected error during blob storage settings validation: {e}"
)
if __name__ == "__main__":
credentials_dict = {

View File

@@ -20,9 +20,13 @@ from onyx.configs.app_configs import INDEX_BATCH_SIZE
from onyx.configs.constants import DocumentSource
from onyx.connectors.interfaces import CheckpointConnector
from onyx.connectors.interfaces import CheckpointOutput
from onyx.connectors.interfaces import ConnectorValidationError
from onyx.connectors.interfaces import CredentialExpiredError
from onyx.connectors.interfaces import GenerateSlimDocumentOutput
from onyx.connectors.interfaces import InsufficientPermissionsError
from onyx.connectors.interfaces import SecondsSinceUnixEpoch
from onyx.connectors.interfaces import SlimConnector
from onyx.connectors.interfaces import UnexpectedError
from onyx.connectors.models import BasicExpertInfo
from onyx.connectors.models import ConnectorCheckpoint
from onyx.connectors.models import ConnectorFailure
@@ -666,6 +670,48 @@ class SlackConnector(SlimConnector, CheckpointConnector):
)
return checkpoint
def validate_connector_settings(self) -> None:
if self.client is None:
raise ConnectorMissingCredentialError("Slack credentials not loaded.")
try:
# Minimal API call to confirm we can list channels
# We set limit=1 for a lightweight check
response = self.client.conversations_list(limit=1, types=["public_channel"])
# Just ensure Slack responded "ok: True"
if not response.get("ok", False):
error_msg = response.get("error", "Unknown error from Slack")
if error_msg == "invalid_auth":
raise ConnectorValidationError(
f"Invalid or expired Slack bot token ({error_msg})."
)
elif error_msg == "not_authed":
raise CredentialExpiredError(
f"Invalid or expired Slack bot token ({error_msg})."
)
raise UnexpectedError(f"Slack API returned a failure: {error_msg}")
except SlackApiError as e:
slack_error = e.response.get("error", "")
if slack_error == "missing_scope":
# The needed scope is typically "channels:read" or "groups:read"
# for viewing channels. The error message might also contain the
# specific scope needed vs. provided.
raise InsufficientPermissionsError(
"Slack bot token lacks the necessary scope to list channels. "
"Please ensure your Slack app has 'channels:read' (or 'groups:read' for private channels) enabled."
)
else:
# Generic Slack error
raise UnexpectedError(
f"Unexpected Slack error '{slack_error}' during settings validation."
)
except Exception as e:
# Catch-all for unexpected exceptions
raise UnexpectedError(
f"Unexpected error during Slack settings validation: {e}"
)
if __name__ == "__main__":
import os