mirror of
https://github.com/langgenius/dify.git
synced 2026-11-13 06:41:07 +08:00
88 lines
3.4 KiB
Python
88 lines
3.4 KiB
Python
from collections.abc import Generator
|
|
from typing import override
|
|
|
|
import tos
|
|
|
|
from configs import dify_config
|
|
from extensions.storage.base_storage import BaseStorage
|
|
|
|
|
|
class VolcengineTosStorage(BaseStorage):
|
|
"""Implementation for Volcengine TOS storage."""
|
|
|
|
def __init__(self):
|
|
super().__init__()
|
|
if not dify_config.VOLCENGINE_TOS_ACCESS_KEY:
|
|
raise ValueError("VOLCENGINE_TOS_ACCESS_KEY is not set")
|
|
if not dify_config.VOLCENGINE_TOS_SECRET_KEY:
|
|
raise ValueError("VOLCENGINE_TOS_SECRET_KEY is not set")
|
|
if not dify_config.VOLCENGINE_TOS_ENDPOINT:
|
|
raise ValueError("VOLCENGINE_TOS_ENDPOINT is not set")
|
|
if not dify_config.VOLCENGINE_TOS_REGION:
|
|
raise ValueError("VOLCENGINE_TOS_REGION is not set")
|
|
self.bucket_name = dify_config.VOLCENGINE_TOS_BUCKET_NAME
|
|
self.client = tos.TosClientV2(
|
|
ak=dify_config.VOLCENGINE_TOS_ACCESS_KEY,
|
|
sk=dify_config.VOLCENGINE_TOS_SECRET_KEY,
|
|
endpoint=dify_config.VOLCENGINE_TOS_ENDPOINT,
|
|
region=dify_config.VOLCENGINE_TOS_REGION,
|
|
)
|
|
|
|
@override
|
|
def save(self, filename, data):
|
|
if not self.bucket_name:
|
|
raise ValueError("VOLCENGINE_TOS_BUCKET_NAME is not set")
|
|
self.client.put_object(bucket=self.bucket_name, key=filename, content=data)
|
|
|
|
@override
|
|
def load_once(self, filename: str) -> bytes:
|
|
if not self.bucket_name:
|
|
raise FileNotFoundError("VOLCENGINE_TOS_BUCKET_NAME is not set")
|
|
data = self.client.get_object(bucket=self.bucket_name, key=filename).read()
|
|
if not isinstance(data, bytes):
|
|
raise TypeError(f"Expected bytes, got {type(data).__name__}")
|
|
return data
|
|
|
|
@override
|
|
def load_stream(self, filename: str) -> Generator:
|
|
if not self.bucket_name:
|
|
raise FileNotFoundError("VOLCENGINE_TOS_BUCKET_NAME is not set")
|
|
response = self.client.get_object(bucket=self.bucket_name, key=filename)
|
|
while chunk := response.read(4096):
|
|
yield chunk
|
|
|
|
@override
|
|
def download(self, filename, target_filepath):
|
|
if not self.bucket_name:
|
|
raise ValueError("VOLCENGINE_TOS_BUCKET_NAME is not set")
|
|
self.client.get_object_to_file(bucket=self.bucket_name, key=filename, file_path=target_filepath)
|
|
|
|
@override
|
|
def exists(self, filename):
|
|
if not self.bucket_name:
|
|
return False
|
|
res = self.client.head_object(bucket=self.bucket_name, key=filename)
|
|
return res.status_code == 200
|
|
|
|
@override
|
|
def delete(self, filename: str):
|
|
if not self.bucket_name:
|
|
return
|
|
self.client.delete_object(bucket=self.bucket_name, key=filename)
|
|
|
|
@override
|
|
def scan(self, path: str, files: bool = True, directories: bool = False) -> list[str]:
|
|
if not self.bucket_name:
|
|
raise ValueError("VOLCENGINE_TOS_BUCKET_NAME is not set")
|
|
keys: list[str] = []
|
|
marker = ""
|
|
while True:
|
|
page = self.client.list_objects(bucket=self.bucket_name, prefix=path, marker=marker, max_keys=1000)
|
|
keys.extend(obj.key for obj in page.contents)
|
|
if not page.is_truncated:
|
|
break
|
|
if not page.next_marker or page.next_marker == marker:
|
|
raise RuntimeError("TOS object listing did not advance its pagination marker")
|
|
marker = page.next_marker
|
|
return self._filter_scan_keys(keys, files, directories)
|