fix(task-deeplink): 共享任务对话的 md/产物按对话创建者解析路径,修复切换账号后找不到文件

xdfx/rwfx/3qfx 任务深链对话按 metadata.taskId 全局共享、不区分用户,但生成的
md/artifacts 与上传文件物理存在 users/{user_id}/threads/{thread_id}/ 下,user_id
之前取当前登录账号。导致 A 账号生成的文件,B 账号打开同一共享对话时后端去
users/B/ 找 → 404「找不到」。

新增 deerflow/runtime/thread_paths.py(resolve_path_user_id /
aresolve_path_user_id):带 taskId 的共享对话统一解析到对话创建者
(threads_meta.user_id),个人对话仍走当前登录(行为不变)。在所有线程文件路径关口替换:
写入源 ThreadDataMiddleware、present_files、uploads(manager+middleware+router)、
产物读写(path_utils → artifacts/skills router)、嵌入式 client.get_artifact。
正向结果进程内缓存;无 DB/内存态/无行均安全回退到当前登录。

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
AdministratorandClaude Opus 4.8 committed 2026-06-19 14:35:22 +08:00
1 parent 13ef1942f2
commit 23e6685977
11 files changed
+269 -99

No files matched your search

@@ -1,10 +1,10 @@
"""Middleware to inject uploaded files information into agent context."""
import logging
import shutil
from pathlib import Path
from typing import NotRequired, override
from urllib.parse import unquote
import logging
import shutil
from pathlib import Path
from typing import NotRequired, override
from urllib.parse import unquote
from langchain.agents import AgentState
from langchain.agents.middleware import AgentMiddleware
@@ -12,10 +12,10 @@ from langchain_core.messages import HumanMessage
from langgraph.runtime import Runtime
from deerflow.config.paths import Paths, get_paths
from deerflow.runtime.user_context import get_effective_user_id
from deerflow.runtime.thread_paths import resolve_path_user_id
from deerflow.utils.file_conversion import extract_outline
logger = logging.getLogger(__name__)
logger = logging.getLogger(__name__)
_OUTLINE_PREVIEW_LINES = 5
@@ -24,8 +24,8 @@ _OUTLINE_PREVIEW_LINES = 5
# configured vision model), never by reading their bytes. Listing them with the
# generic read_file/grep guidance made the agent try to parse image binaries
# with bash/python, which fails and wastes turns.
_IMAGE_EXTENSIONS = {".jpg", ".jpeg", ".png", ".webp", ".gif", ".bmp"}
_ARTIFACT_URL_MARKER = "/artifacts/"
_IMAGE_EXTENSIONS = {".jpg", ".jpeg", ".png", ".webp", ".gif", ".bmp"}
_ARTIFACT_URL_MARKER = "/artifacts/"
def _extract_outline_for_file(file_path: Path) -> tuple[list[dict], list[str]]:
@@ -72,7 +72,7 @@ class UploadsMiddlewareState(AgentState):
uploaded_files: NotRequired[list[dict] | None]
class UploadsMiddleware(AgentMiddleware[UploadsMiddlewareState]):
class UploadsMiddleware(AgentMiddleware[UploadsMiddlewareState]):
"""Middleware to inject uploaded files information into the agent context.
Reads file metadata from the current message's additional_kwargs.files
@@ -125,7 +125,7 @@ class UploadsMiddleware(AgentMiddleware[UploadsMiddlewareState]):
lines.append(" Use `grep` to search for keywords (e.g. `grep(pattern='keyword', path='/mnt/user-data/uploads/')`).")
lines.append("")
def _create_files_message(self, new_files: list[dict], historical_files: list[dict]) -> str:
def _create_files_message(self, new_files: list[dict], historical_files: list[dict]) -> str:
"""Create a formatted message listing uploaded files.
Args:
@@ -163,61 +163,61 @@ class UploadsMiddleware(AgentMiddleware[UploadsMiddlewareState]):
lines.append("- Only fall back to web search if the file content is clearly insufficient to answer the question.")
lines.append("</uploaded_files>")
return "\n".join(lines)
def _import_file_from_artifact_url(
self,
file_info: dict,
uploads_dir: Path | None,
) -> bool:
"""Copy a previously uploaded file into the current thread when possible.
iframe-style callers may pass file metadata that points at an upload
from another DeerFlow thread. The current run thread still needs a real
file under its own ``/mnt/user-data/uploads`` directory, otherwise this
middleware will filter the entry out during existence checks.
"""
filename = file_info.get("filename") or ""
artifact_url = str(file_info.get("artifact_url") or "").strip()
if not filename or uploads_dir is None or not artifact_url:
return False
if not artifact_url.startswith("/api/threads/") or _ARTIFACT_URL_MARKER not in artifact_url:
return False
prefix, encoded_virtual_path = artifact_url.split(_ARTIFACT_URL_MARKER, 1)
source_thread_id = prefix.removeprefix("/api/threads/").strip("/")
if not source_thread_id or not encoded_virtual_path:
return False
virtual_path = "/" + unquote(encoded_virtual_path.lstrip("/"))
if Path(virtual_path).name != filename:
return False
try:
source_path = self._paths.resolve_virtual_path(
source_thread_id,
virtual_path,
user_id=get_effective_user_id(),
)
except ValueError:
logger.debug("Failed to resolve uploaded artifact path: %s", artifact_url, exc_info=True)
return False
if not source_path.is_file():
return False
uploads_dir.mkdir(parents=True, exist_ok=True)
target_path = uploads_dir / filename
shutil.copy2(source_path, target_path)
source_markdown = source_path.with_suffix(".md")
if source_markdown.is_file() and source_markdown != source_path:
shutil.copy2(source_markdown, target_path.with_suffix(".md"))
return True
def _files_from_kwargs(self, message: HumanMessage, uploads_dir: Path | None = None) -> list[dict] | None:
return "\n".join(lines)
def _import_file_from_artifact_url(
self,
file_info: dict,
uploads_dir: Path | None,
) -> bool:
"""Copy a previously uploaded file into the current thread when possible.
iframe-style callers may pass file metadata that points at an upload
from another DeerFlow thread. The current run thread still needs a real
file under its own ``/mnt/user-data/uploads`` directory, otherwise this
middleware will filter the entry out during existence checks.
"""
filename = file_info.get("filename") or ""
artifact_url = str(file_info.get("artifact_url") or "").strip()
if not filename or uploads_dir is None or not artifact_url:
return False
if not artifact_url.startswith("/api/threads/") or _ARTIFACT_URL_MARKER not in artifact_url:
return False
prefix, encoded_virtual_path = artifact_url.split(_ARTIFACT_URL_MARKER, 1)
source_thread_id = prefix.removeprefix("/api/threads/").strip("/")
if not source_thread_id or not encoded_virtual_path:
return False
virtual_path = "/" + unquote(encoded_virtual_path.lstrip("/"))
if Path(virtual_path).name != filename:
return False
try:
source_path = self._paths.resolve_virtual_path(
source_thread_id,
virtual_path,
user_id=resolve_path_user_id(source_thread_id),
)
except ValueError:
logger.debug("Failed to resolve uploaded artifact path: %s", artifact_url, exc_info=True)
return False
if not source_path.is_file():
return False
uploads_dir.mkdir(parents=True, exist_ok=True)
target_path = uploads_dir / filename
shutil.copy2(source_path, target_path)
source_markdown = source_path.with_suffix(".md")
if source_markdown.is_file() and source_markdown != source_path:
shutil.copy2(source_markdown, target_path.with_suffix(".md"))
return True
def _files_from_kwargs(self, message: HumanMessage, uploads_dir: Path | None = None) -> list[dict] | None:
"""Extract file info from message additional_kwargs.files.
The frontend sends uploaded file metadata in additional_kwargs.files
@@ -238,17 +238,17 @@ class UploadsMiddleware(AgentMiddleware[UploadsMiddlewareState]):
files = []
for f in kwargs_files:
if not isinstance(f, dict):
continue
filename = f.get("filename") or ""
if not filename or Path(filename).name != filename:
continue
if uploads_dir is not None and not (uploads_dir / filename).is_file():
if not self._import_file_from_artifact_url(f, uploads_dir):
continue
files.append(
{
"filename": filename,
if not isinstance(f, dict):
continue
filename = f.get("filename") or ""
if not filename or Path(filename).name != filename:
continue
if uploads_dir is not None and not (uploads_dir / filename).is_file():
if not self._import_file_from_artifact_url(f, uploads_dir):
continue
files.append(
{
"filename": filename,
"size": int(f.get("size") or 0),
"path": f"/mnt/user-data/uploads/{filename}",
"extension": Path(filename).suffix,
@@ -294,7 +294,7 @@ class UploadsMiddleware(AgentMiddleware[UploadsMiddlewareState]):
thread_id = get_config().get("configurable", {}).get("thread_id")
except RuntimeError:
pass # get_config() raises outside a runnable context (e.g. unit tests)
uploads_dir = self._paths.sandbox_uploads_dir(thread_id, user_id=get_effective_user_id()) if thread_id else None
uploads_dir = self._paths.sandbox_uploads_dir(thread_id, user_id=resolve_path_user_id(thread_id)) if thread_id else None
# Get newly uploaded files from the current message's additional_kwargs.files
new_files = self._files_from_kwargs(last_message, uploads_dir) or []