geekan
AttributeError: 'list' object has no attribute 'keys'
Bug description Traceback (most recent call last): File "/root/workspace/jjyue/MetaGPT-main/metagpt/utils/common.py", line 640, in wrapper return await func(self, *args, **kwargs) File "/root/workspace/jjyue/MetaGPT-main/metagpt/roles/role.py", line 555, in run rsp = await self.react() File "/root/workspace/jjyue/MetaGPT-main/metagpt/roles/researcher.py", line 130, in react msg = await super().react() File "/root/workspace/jjyue/MetaGPT-main/metagpt/roles/role.py", line 524, in react rsp = await self._act_by_order() File "/root/workspace/jjyue/MetaGPT-main/metagpt/roles/role.py", line 478, in _act_by_order rsp = await self._act() File "/root/workspace/jjyue/MetaGPT-main/metagpt/roles/researcher.py", line 74, in _act links = await todo.run(topic, 4, 4) File "/root/workspace/jjyue/MetaGPT-main/metagpt/actions/research.py", line 123, in run results = await asyncio.gather( File "/root/workspace/jjyue/MetaGPT-main/metagpt/tools/search_engine.py", line 164, in run raise e File "/root/workspace/jjyue/MetaGPT-main/metagpt/tools/search_engine.py", line 159, in run return await self.run_func(query, max_results=max_results, as_string=as_string) File "/root/workspace/jjyue/MetaGPT-main/metagpt/tools/search_engine_serpapi.py", line 55, in run return self._process_response(result, as_string=as_string) File "/root/workspace/jjyue/MetaGPT-main/metagpt/tools/search_engine_serpapi.py", line 103, in _process_response if "answer_box" in res.keys() and "answer" in res["answer_box"].keys(): AttributeError: 'list' object has no attribute 'keys'
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/root/workspace/jjyue/MetaGPT-main/entry.py", line 23, in
Bug solved method
Environment information
- LLM type and model name:GLM-4
- System version:centos 7.6
- Python version:3.10.12
- MetaGPT version or branch:0.8.1
- packages version:
- installation method: pip install from source
Screenshots or logs
It seems res["answer_box"] is a list not a dict.
Is it reproduced? could you post the value of res which in this func
yes it is.
res["answer_box"] is a list, but the content is a dict,see below.
[
{
"position":1,
"title":"桥梁施工方法",
"link":"http://nourl.ubs.baidu.com/278",
"type":"ai_content",
"snippet":"桥梁施工方法多样，主要包括以下几种：\n钻孔灌注桩施工。包括测量放样、设置护简、钻孔、清孔、放钢筋笼和浇筑混凝土。\n扩大基础施工。包括基坑开挖、基坑排水、基底处理以及砌筑或浇筑基础结构物。\n承台墩身施工。涉及安装模板、钢筋绑扎和分层浇筑混凝土。\n梁板预制施工。包括地基处理、浇筑台座、安装板梁、绑扎钢筋、浇筑混凝土和养护。\n先简支后连续施工。包括设置临时支座、安装梁板、安装底模、浇筑现浇段、张拉预应力连续束和压浆。\n桥面工程施工。包括安装泄水管、浇筑护栏混凝土、铺设钢筋网和浇筑桥面混凝土。\n悬臂施工法A。适用于较大的桥梁，通过逐渐延伸的臂架进行施工。\n顶推施工法A。在桥梁的一端施力，使其沿轨道逐渐移动直至完成。\n移动模架逐孔施工法A。使用移动模架逐个完成每孔梁的施工。\n支架现浇法A。通过搭建脚手架或其他支撑结构，在现场直接浇筑混凝土。\n预制安装法A。在工厂预制梁体，然后运输至现场进行安装。\n每种方法都有其适用场景和优势，选择哪种方法取决于具体的桥梁设计、地理位置、环境条件等因素。",
"snippet_links":[],
"references":[]
}
]
The type of the field may have changed, or there is a special situation. FYI: @shenchucheng
@Yue-Rain Could you please provide the params you use? The default values are:
{
"engine": "google",
"google_domain": "google.com",
"gl": "us",
"hl": "en"
}
Due to the lack of updates or replies by the user for a long time, we will close it. Please reopen it if necessary.
