feat: 上传文件后端提取文本内容(docx/txt)作为中性上下文注入agent,结合prompt由agent决定动作

wwwroot/api/cockpit_chat.dspy

@@ -1,9 +1,28 @@
 # cockpit_chat.dspy - SDLC Agent Loop v3.1 (streaming NDJSON)
 import aiohttp
 import os
+import re
 import shutil
+import zipfile
 from ahserver.filestorage import FileStorage
 
+
+def _extract_text(path, name):
+    """提取文件文本内容（docx/txt/md等），返回文本或空字符串。二进制/无法解析返回空。"""
+    ext = os.path.splitext(name)[1].lower()
+    try:
+        if ext in ('.txt', '.md', '.json', '.csv', '.py', '.log', '.yaml', '.yml', '.xml', '.html', '.ini'):
+            with open(path, 'r', encoding='utf-8', errors='ignore') as f:
+                return f.read()[:15000]
+        if ext == '.docx':
+            with zipfile.ZipFile(path) as z:
+                xml = z.read('word/document.xml').decode('utf-8', errors='ignore')
+                texts = re.findall(r'<w:t[^>]*>(.*?)</w:t>', xml)
+                return '\n'.join(texts)[:15000]
+    except Exception:
+        pass
+    return ''
+
 action = (params_kw or {}).get('action', 'send_message')
 # AgentIO sends JSON body without action field — detect from message_text presence
 if action != 'send_message':
@@ -567,8 +586,14 @@ if action == 'send_message':
                 except Exception as _e:
                     debug(f"UPLOAD FAIL {_fp}: {_e}")
             if uploaded_files:
-                _fl = "\n".join(f"- {n}（工作空间路径: {d}）" for n, d in uploaded_files)
+                _parts = []
-                msgs.append({"role": "system", "content": f"用户本次上传了文件，已保存到工作空间 uploads/ 目录，可用 read_file 工具读取文件内容。文件清单：\n{_fl}"})
+                for _n, _d in uploaded_files:
+                    _txt = _extract_text(_d, _n)
+                    if _txt:
+                        _parts.append(f"【文件 {_n} 内容】\n{_txt}")
+                    else:
+                        _parts.append(f"【文件 {_n}】无法直接读取文本（可能是二进制），相对路径 uploads/{_n}，可用 run_command 处理")
+                msgs.append({"role": "system", "content": "用户本次上传了文件，内容如下：\n\n" + "\n\n".join(_parts)})
             msgs.append({"role":"user","content":message_text})
             if uid:
                 await sor.C('pipeline_conversations',{'id':getID(),'role':'user','content':message_text,'msg_type':'text','org_id':org_id,'created_by':uid,'iteration_id':ctx['pid'] or ''})