mirror of
https://github.com/langgenius/dify.git
synced 2026-02-13 07:01:23 -05:00
248fa38c34
- Updated storage wrappers to utilize a new base class, StorageWrapper, for better delegation of methods. - Introduced SilentStorage to handle read operations gracefully by returning empty values instead of raising exceptions. - Enhanced CachedPresignStorage to support batch caching of download URLs, improving performance. - Refactored FilePresignStorage to support both presigned URLs and signed proxy URLs for downloads. - Updated AppAssetService to utilize the new storage structure, ensuring consistent asset management.
114 lines
4.1 KiB
Python
114 lines
4.1 KiB
Python
import logging
|
|
from collections.abc import Generator
|
|
|
|
import boto3
|
|
from botocore.client import Config
|
|
from botocore.exceptions import ClientError
|
|
|
|
from configs import dify_config
|
|
from extensions.storage.base_storage import BaseStorage
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
class AwsS3Storage(BaseStorage):
|
|
"""Implementation for Amazon Web Services S3 storage."""
|
|
|
|
def __init__(self):
|
|
super().__init__()
|
|
self.bucket_name = dify_config.S3_BUCKET_NAME
|
|
if dify_config.S3_USE_AWS_MANAGED_IAM:
|
|
logger.info("Using AWS managed IAM role for S3")
|
|
|
|
session = boto3.Session()
|
|
region_name = dify_config.S3_REGION
|
|
self.client = session.client(service_name="s3", region_name=region_name)
|
|
else:
|
|
logger.info("Using ak and sk for S3")
|
|
|
|
self.client = boto3.client(
|
|
"s3",
|
|
aws_secret_access_key=dify_config.S3_SECRET_KEY,
|
|
aws_access_key_id=dify_config.S3_ACCESS_KEY,
|
|
endpoint_url=dify_config.S3_ENDPOINT,
|
|
region_name=dify_config.S3_REGION,
|
|
config=Config(s3={"addressing_style": dify_config.S3_ADDRESS_STYLE}),
|
|
)
|
|
# create bucket
|
|
try:
|
|
self.client.head_bucket(Bucket=self.bucket_name)
|
|
except ClientError as e:
|
|
# if bucket not exists, create it
|
|
if e.response.get("Error", {}).get("Code") == "404":
|
|
self.client.create_bucket(Bucket=self.bucket_name)
|
|
# if bucket is not accessible, pass, maybe the bucket is existing but not accessible
|
|
elif e.response.get("Error", {}).get("Code") == "403":
|
|
pass
|
|
else:
|
|
# other error, raise exception
|
|
raise
|
|
|
|
def save(self, filename, data):
|
|
self.client.put_object(Bucket=self.bucket_name, Key=filename, Body=data)
|
|
|
|
def load_once(self, filename: str) -> bytes:
|
|
try:
|
|
data: bytes = self.client.get_object(Bucket=self.bucket_name, Key=filename)["Body"].read()
|
|
except ClientError as ex:
|
|
if ex.response.get("Error", {}).get("Code") == "NoSuchKey":
|
|
raise FileNotFoundError("File not found")
|
|
else:
|
|
raise
|
|
return data
|
|
|
|
def load_stream(self, filename: str) -> Generator:
|
|
try:
|
|
response = self.client.get_object(Bucket=self.bucket_name, Key=filename)
|
|
yield from response["Body"].iter_chunks()
|
|
except ClientError as ex:
|
|
if ex.response.get("Error", {}).get("Code") == "NoSuchKey":
|
|
raise FileNotFoundError("file not found")
|
|
elif "reached max retries" in str(ex):
|
|
raise ValueError("please do not request the same file too frequently")
|
|
else:
|
|
raise
|
|
|
|
def download(self, filename, target_filepath):
|
|
self.client.download_file(self.bucket_name, filename, target_filepath)
|
|
|
|
def exists(self, filename):
|
|
try:
|
|
self.client.head_object(Bucket=self.bucket_name, Key=filename)
|
|
return True
|
|
except:
|
|
return False
|
|
|
|
def delete(self, filename):
|
|
self.client.delete_object(Bucket=self.bucket_name, Key=filename)
|
|
|
|
def get_download_url(self, filename: str, expires_in: int = 3600) -> str:
|
|
url: str = self.client.generate_presigned_url(
|
|
ClientMethod="get_object",
|
|
Params={"Bucket": self.bucket_name, "Key": filename},
|
|
ExpiresIn=expires_in,
|
|
)
|
|
return url
|
|
|
|
def get_download_urls(self, filenames: list[str], expires_in: int = 3600) -> list[str]:
|
|
return [
|
|
self.client.generate_presigned_url(
|
|
ClientMethod="get_object",
|
|
Params={"Bucket": self.bucket_name, "Key": filename},
|
|
ExpiresIn=expires_in,
|
|
)
|
|
for filename in filenames
|
|
]
|
|
|
|
def get_upload_url(self, filename: str, expires_in: int = 3600) -> str:
|
|
url: str = self.client.generate_presigned_url(
|
|
ClientMethod="put_object",
|
|
Params={"Bucket": self.bucket_name, "Key": filename},
|
|
ExpiresIn=expires_in,
|
|
)
|
|
return url
|