From 1df3fc416ade2bd39b165f189b8c1e95e727e712 Mon Sep 17 00:00:00 2001
From: Timebomb2018 <18868801967@163.com>
Date: Thu, 26 Mar 2026 16:19:40 +0800
Subject: [PATCH] feat(workflow): Document extraction node

---
 api/app/core/workflow/nodes/document_extractor/config.py | 7 +------
 api/app/core/workflow/nodes/document_extractor/node.py   | 4 +++-
 2 files changed, 4 insertions(+), 7 deletions(-)

diff --git a/api/app/core/workflow/nodes/document_extractor/config.py b/api/app/core/workflow/nodes/document_extractor/config.py
index dd946422..69f7f76d 100644
--- a/api/app/core/workflow/nodes/document_extractor/config.py
+++ b/api/app/core/workflow/nodes/document_extractor/config.py
@@ -7,17 +7,12 @@ class DocExtractorNodeConfig(BaseNodeConfig):
         ...,
         description="File variable selector, e.g. {{ sys.files }} or {{ node_id.file }}"
     )
-    output_format: str = Field(
-        default="text",
-        description="Output format: 'text' | 'markdown'"
-    )
 
     class Config:
         json_schema_extra = {
             "examples": [
                 {
-                    "file_selector": "{{ sys.files }}",
-                    "output_format": "text"
+                    "file_selector": "{{ sys.files }}"
                 }
             ]
         }
diff --git a/api/app/core/workflow/nodes/document_extractor/node.py b/api/app/core/workflow/nodes/document_extractor/node.py
index 050f693f..40641f3c 100644
--- a/api/app/core/workflow/nodes/document_extractor/node.py
+++ b/api/app/core/workflow/nodes/document_extractor/node.py
@@ -29,12 +29,14 @@ def _normalise_files(val: Any) -> list[FileObject]:
     if isinstance(val, dict) and val.get("is_file"):
         return [FileObject(**val)]
     if isinstance(val, list):
-        result = []
+        result: list[FileObject] = []
         for item in val:
             if isinstance(item, FileObject):
                 result.append(item)
             elif isinstance(item, dict) and item.get("is_file"):
                 result.append(FileObject(**item))
+            else:
+                logger.warning("Ignoring non-file entry in file list for document extractor: %r", item)
         return result
     return []