[Bug]: Run the knowledge_graph parsing method, after parsing the document, an error occurs during the retrieval_test in search.py:323.

[fengyunweia]

Is there an existing issue for the same bug?

• I have checked the existing issues.

Branch name

main

Commit ID

#3306

Other environment information

No response

Actual behavior

Run the knowledge_graph parsing method, after parsing the document, an error occurs during the retrieval_test in search.py:323. The error message is: "sres.field[i][cfield].split(" ")", where cfield is set to "content_ltks", but the knowledge_graph parsing method has the following keywords without the "content_ltks" attribute: "knowledge_graph_kwd": "graph", "knowledge_graph_kwd": "mind_map", and "knowledge_graph_kwd": "text".

Expected behavior

No response

Steps to reproduce

1: Run the knowledge_graph parsing method
2:retrieval_test

Additional information

No response

[fengyunweia]

code:

chunks.append(
{
"content_with_weight": json.dumps(nx.node_link_data(graph), ensure_ascii=False, indent=2),
"knowledge_graph_kwd": "graph"
})
You can see that the "knowledge_graph_kwd" attribute is set to "graph". There are only two attributes in total.

[fengyunweia]

code:

chunks.append(
{
"content_with_weight": json.dumps(nx.node_link_data(graph), ensure_ascii=False, indent=2),
"knowledge_graph_kwd": "graph"
})
You can see that the "knowledge_graph_kwd" attribute is set to "graph". There are only two attributes in total.