YueGuobin eb15c7138b
Fix list_node_files PermissionError on os.scandir
Wrap os.scandir() in try-except to return empty list instead of
crashing when a node directory is not readable.
2026-06-10 00:15:26 +08:00

587 lines
20 KiB
Python

#
# Copyright (C) 2015 GNS3 Technologies Inc.
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
import os
import shutil
import magic
import asyncio
import hashlib
import datetime
from uuid import UUID, uuid4
from fastapi import HTTPException, status
from gns3server.compute.compute_error import ComputeError, ComputeNotFoundError, ComputeForbiddenError
from .port_manager import PortManager
from .notification_manager import NotificationManager
from ..config import Config
from ..utils.asyncio import wait_run_in_executor
from ..utils.path import check_path_allowed, get_default_project_directory
import logging
log = logging.getLogger(__name__)
class Project:
"""
A project contains a list of nodes.
In theory nodes are isolated project/project.
:param project_id: force project identifier (None by default auto generate an UUID)
:param path: path of the project. (None use the standard directory)
"""
def __init__(self, name=None, project_id=None, path=None, variables=None):
self._name = name
if project_id:
try:
UUID(project_id, version=4)
except ValueError:
raise ComputeError(f"{project_id} is not a valid UUID")
else:
project_id = str(uuid4())
self._id = project_id
self._deleted = False
self._nodes = set()
self._used_tcp_ports = set()
self._used_udp_ports = set()
self._variables = variables
if path is None:
location = get_default_project_directory()
path = os.path.join(location, self._id)
try:
os.makedirs(path, exist_ok=True)
except OSError as e:
raise ComputeError(f"Could not create project directory: {e}")
self.path = path
try:
if os.path.exists(self.tmp_working_directory()):
shutil.rmtree(self.tmp_working_directory())
except OSError as e:
raise ComputeError(f"Could not clean project directory: {e}")
log.info(f"Project {self._id} with path '{self._path}' created")
def asdict(self):
return {
"name": self._name,
"project_id": self._id,
"variables": self._variables
}
@property
def id(self):
return self._id
@property
def path(self):
return self._path
@path.setter
def path(self, path):
if hasattr(self, "_path"):
if path != self._path:
raise ComputeForbiddenError("Changing the project directory path is not allowed")
check_path_allowed(path)
self._path = path
@property
def name(self):
return self._name
@name.setter
def name(self, name):
if "/" in name or "\\" in name:
raise ComputeForbiddenError("Project names cannot contain path separators")
self._name = name
@property
def nodes(self):
return self._nodes
@property
def variables(self):
return self._variables
@variables.setter
def variables(self, variables):
self._variables = variables
def record_tcp_port(self, port):
"""
Associate a reserved TCP port number with this project.
:param port: TCP port number
"""
if port not in self._used_tcp_ports:
self._used_tcp_ports.add(port)
def record_udp_port(self, port):
"""
Associate a reserved UDP port number with this project.
:param port: UDP port number
"""
if port not in self._used_udp_ports:
self._used_udp_ports.add(port)
def remove_tcp_port(self, port):
"""
Removes an associated TCP port number from this project.
:param port: TCP port number
"""
if port in self._used_tcp_ports:
self._used_tcp_ports.remove(port)
def remove_udp_port(self, port):
"""
Removes an associated UDP port number from this project.
:param port: UDP port number
"""
if port in self._used_udp_ports:
self._used_udp_ports.remove(port)
def module_working_directory(self, module_name):
"""
Returns a working directory for the module
The directory is created if the directory doesn't exist.
:param module_name: name for the module
:returns: working directory
"""
workdir = self.module_working_path(module_name)
if not self._deleted:
try:
os.makedirs(workdir, exist_ok=True)
except OSError as e:
raise ComputeError(f"Could not create module working directory: {e}")
return workdir
def module_working_path(self, module_name):
"""
Returns the working directory for the module. If you want
to be sure to have the directory on disk take a look on:
module_working_directory
"""
return os.path.join(self._path, "project-files", module_name)
def node_working_directory(self, node):
"""
Returns a working directory for a specific node.
If the directory doesn't exist, the directory is created.
:param node: Node instance
:returns: Node working directory
"""
workdir = self.node_working_path(node)
if not self._deleted:
try:
os.makedirs(workdir, exist_ok=True)
except OSError as e:
raise ComputeError(f"Could not create the node working directory: {e}")
return workdir
def node_working_path(self, node):
"""
Returns a node working path for node. It doesn't create structure if not present on system.
:param node: Node instance
:return: Node working path
"""
return os.path.join(self._path, "project-files", node.manager.module_name.lower(), node.id)
def tmp_working_directory(self):
"""
A temporary directory. Will be clean at project open and close
"""
return os.path.join(self._path, "tmp")
def capture_working_directory(self):
"""
Returns a working directory where to store packet capture files.
:returns: path to the directory
"""
workdir = os.path.join(self._path, "project-files", "captures")
if not self._deleted:
try:
os.makedirs(workdir, exist_ok=True)
except OSError as e:
raise ComputeError(f"Could not create the capture working directory: {e}")
return workdir
def add_node(self, node):
"""
Adds a node to the project.
In theory this should be called by the node manager.
:param node: Node instance
"""
self._nodes.add(node)
def get_node(self, node_id):
"""
Returns a Node instance.
:param node_id: Node identifier
:returns: Node instance
"""
try:
UUID(node_id, version=4)
except ValueError:
raise ComputeError(f"Node ID {node_id} is not a valid UUID")
for node in self._nodes:
if node.id == node_id:
return node
raise ComputeNotFoundError(f"Node ID {node_id} doesn't exist")
async def remove_node(self, node):
"""
Removes a node from the project.
In theory this should be called by the node manager.
:param node: Node instance
"""
if node in self._nodes:
await node.delete()
self._nodes.remove(node)
async def update(self, variables=None, **kwargs):
original_variables = self.variables
self.variables = variables
# we need to update docker nodes when variables changes
if original_variables != variables:
# Parallelize node updates for better performance
tasks = []
for node in self.nodes:
if hasattr(node, "update"):
tasks.append(node.update())
if tasks:
await asyncio.gather(*tasks, return_exceptions=True)
async def close(self):
"""
Closes the project, but keep project data on disk
"""
project_nodes_id = {n.id for n in self.nodes}
for module in self.compute():
module_nodes_id = {n.id for n in module.instance().nodes}
# We close the project only for the modules using it
if len(module_nodes_id & project_nodes_id):
await module.instance().project_closing(self)
await self._close_and_clean(False)
for module in self.compute():
module_nodes_id = {n.id for n in module.instance().nodes}
# We close the project only for the modules using it
if len(module_nodes_id & project_nodes_id):
await module.instance().project_closed(self)
try:
if os.path.exists(self.tmp_working_directory()):
shutil.rmtree(self.tmp_working_directory())
except OSError:
pass
async def _close_and_clean(self, cleanup):
"""
Closes the project, and cleanup the disk if cleanup is True
:param cleanup: Whether to delete the project directory
"""
tasks = []
for node in self._nodes:
tasks.append(asyncio.ensure_future(node.manager.close_node(node.id)))
if tasks:
done, _ = await asyncio.wait(tasks)
for future in done:
try:
future.result()
except (Exception, GeneratorExit) as e:
log.error(f"Could not close node: {e}", exc_info=1)
if cleanup and os.path.exists(self.path):
self._deleted = True
try:
await wait_run_in_executor(shutil.rmtree, self.path)
log.info(f"Project {self._id} with path '{self._path}' deleted")
except OSError as e:
raise ComputeError(f"Could not delete the project directory: {e}")
else:
log.info(f"Project {self._id} with path '{self._path}' closed")
if self._used_tcp_ports:
log.warning(f"Project {self.id} has TCP ports still in use: {self._used_tcp_ports}")
if self._used_udp_ports:
log.warning(f"Project {self.id} has UDP ports still in use: {self._used_udp_ports}")
# clean the remaining ports that have not been cleaned by their respective node.
port_manager = PortManager.instance()
for port in self._used_tcp_ports.copy():
port_manager.release_tcp_port(port, self)
for port in self._used_udp_ports.copy():
port_manager.release_udp_port(port, self)
async def delete(self):
"""
Removes project from disk
"""
for module in self.compute():
await module.instance().project_closing(self)
await self._close_and_clean(True)
for module in self.compute():
await module.instance().project_closed(self)
def compute(self):
"""
Returns all loaded modules from compute.
"""
# We import it at the last time to avoid circular dependencies
from ..compute import MODULES
return MODULES
def emit(self, action, event):
"""
Send an event to all the client listening for notifications
:param action: Action name
:param event: Event to send
"""
NotificationManager.instance().emit(action, event, project_id=self.id)
async def list_files(self):
"""
:returns: Array of files in project without temporary files. The files are dictionary {"path": "test.bin", "md5sum": "aaaaa"}
"""
files = []
for dirpath, dirnames, filenames in os.walk(self.path, followlinks=False):
for filename in filenames:
if not filename.endswith(".ghost"):
path = os.path.relpath(dirpath, self.path)
path = os.path.join(path, filename)
path = os.path.normpath(path)
file_info = {"path": path}
try:
file_info["md5sum"] = await wait_run_in_executor(
self._hash_file, os.path.join(dirpath, filename)
)
except OSError:
continue
files.append(file_info)
return files
async def list_node_files(self, node_path: str, subpath: str = "", recursive: bool = False):
"""
List files in a specific node directory with detailed metadata.
:param node_path: Relative path to node directory (e.g., "project-files/qemu/node-id")
:param subpath: Optional subdirectory path. Defaults to root of node directory.
:param recursive: If True, recursively list all files (use with caution on large directories).
:returns: Array of files in the node directory with metadata
"""
node_full_path = os.path.normpath(os.path.join(self.path, node_path))
subpath = subpath.lstrip("/")
if subpath:
target_path = os.path.normpath(os.path.join(node_full_path, subpath))
else:
target_path = node_full_path
# Security check: ensure the path is within the node directory
if not os.path.commonpath([target_path, node_full_path]) == node_full_path:
raise HTTPException(status_code=status.HTTP_403_FORBIDDEN,
detail="Path is outside the node directory")
if not os.path.exists(target_path):
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND,
detail="Path not found")
if recursive:
return await self._list_node_files_recursive(node_full_path, target_path,
node_full_path if subpath else None)
# Non-recursive: list only the current directory level
files = []
try:
scandir_iter = os.scandir(target_path)
except PermissionError:
return files
except OSError as e:
log.error(f"Error listing node directory '{target_path}': {e}")
return files
for entry in scandir_iter:
name = entry.name
rel_path = name if not subpath else os.path.join(subpath, name)
try:
stat_info = await wait_run_in_executor(lambda e=entry: e.stat())
is_dir = await wait_run_in_executor(lambda e=entry: e.is_dir())
if is_dir:
try:
created_at = datetime.datetime.fromtimestamp(stat_info.st_ctime).isoformat()
modified_at = datetime.datetime.fromtimestamp(stat_info.st_mtime).isoformat()
except (OSError, OverflowError, ValueError):
created_at = modified_at = ""
files.append({
"path": rel_path,
"size": stat_info.st_size,
"created_at": created_at,
"modified_at": modified_at,
"file_type": "directory"
})
else:
if name.endswith(".ghost"):
continue
try:
file_type = await wait_run_in_executor(
lambda e=entry: magic.from_file(e.path, mime=False)
)
except Exception as e:
log.warning(f"Error getting file type for '{rel_path}': {e}")
file_type = ""
try:
created_at = datetime.datetime.fromtimestamp(stat_info.st_ctime).isoformat()
modified_at = datetime.datetime.fromtimestamp(stat_info.st_mtime).isoformat()
except (OSError, OverflowError, ValueError):
created_at = modified_at = ""
files.append({
"path": rel_path,
"size": stat_info.st_size,
"created_at": created_at,
"modified_at": modified_at,
"file_type": file_type
})
except OSError:
continue
return files
async def _list_node_files_recursive(self, node_full_path, start_path, base_path=None):
"""
Recursively list all files and directories under start_path.
"""
if base_path is None:
base_path = node_full_path
files = []
for dirpath, dirnames, filenames in os.walk(start_path, followlinks=False):
for dirname in dirnames:
dir_full_path = os.path.join(dirpath, dirname)
try:
stat_info = await wait_run_in_executor(os.stat, dir_full_path)
try:
created_at = datetime.datetime.fromtimestamp(stat_info.st_ctime).isoformat()
modified_at = datetime.datetime.fromtimestamp(stat_info.st_mtime).isoformat()
except (OSError, OverflowError, ValueError):
created_at = modified_at = ""
files.append({
"path": os.path.relpath(dir_full_path, base_path),
"size": stat_info.st_size,
"created_at": created_at,
"modified_at": modified_at,
"file_type": "directory"
})
except OSError:
continue
for filename in filenames:
if filename.endswith(".ghost"):
continue
file_path = os.path.join(dirpath, filename)
rel_path = os.path.relpath(file_path, base_path)
try:
stat_info = await wait_run_in_executor(os.stat, file_path)
try:
file_type = await wait_run_in_executor(
lambda fp=file_path: magic.from_file(fp, mime=False)
)
except Exception as e:
log.warning(f"Error getting file type for '{rel_path}': {e}")
file_type = ""
try:
created_at = datetime.datetime.fromtimestamp(stat_info.st_ctime).isoformat()
modified_at = datetime.datetime.fromtimestamp(stat_info.st_mtime).isoformat()
except (OSError, OverflowError, ValueError) as e:
log.warning(f"Invalid timestamp for '{rel_path}': {e}")
created_at = modified_at = ""
files.append({
"path": rel_path,
"size": stat_info.st_size,
"created_at": created_at,
"modified_at": modified_at,
"file_type": file_type
})
except OSError as e:
log.warning(f"Error getting metadata for file '{rel_path}': {e}")
continue
return files
def _hash_file(self, path):
"""
Compute and md5 hash for file
:returns: hexadecimal md5
"""
m = hashlib.md5()
with open(path, "rb") as f:
while True:
buf = f.read(128)
if not buf:
break
m.update(buf)
return m.hexdigest()