mirror of
https://github.com/infiniflow/ragflow.git
synced 2026-08-04 06:40:29 +08:00
454dea686e
### Summary Agentic search <img width="1149" height="1575" alt="image" src="https://github.com/user-attachments/assets/bce9a3e7-0517-4fb2-80a2-5d2a81a4da78" /> --------- Co-authored-by: Yingfeng Zhang <yingfeng.zhang@gmail.com>
50 lines
1.5 KiB
Python
50 lines
1.5 KiB
Python
"""Low mode: direct single-pass search."""
|
|
|
|
import logging
|
|
|
|
from rag.advanced_rag.harness.tools.search import hybrid_search
|
|
|
|
_LOG = logging.getLogger(__name__)
|
|
|
|
|
|
async def direct_search(state: dict, tools) -> dict:
|
|
"""Single hybrid search → merge into kbinfos."""
|
|
question = state.get("question", "")
|
|
keywords = state.get("keywords", "")
|
|
_LOG.info("[direct_search] question=%s | keywords=%s", question, keywords)
|
|
|
|
result = await hybrid_search(tools, query=question, keywords=keywords)
|
|
_merge_kbinfos(tools, result)
|
|
|
|
if not _has_chunks(tools):
|
|
_LOG.info("[direct_search] no results found")
|
|
return {"empty_result": True, "kbinfos": tools.kbinfos}
|
|
|
|
return {"kbinfos": tools.kbinfos}
|
|
|
|
|
|
def _merge_kbinfos(tools, result: dict):
|
|
if not result or not result.get("chunks"):
|
|
return
|
|
seen = {_chunk_key(c) for c in tools.kbinfos.get("chunks", [])}
|
|
for c in result.get("chunks", []):
|
|
k = _chunk_key(c)
|
|
if k in seen:
|
|
continue
|
|
seen.add(k)
|
|
tools.kbinfos.setdefault("chunks", []).append(c)
|
|
dseen = {d.get("doc_id") for d in tools.kbinfos.get("doc_aggs", [])}
|
|
for d in result.get("doc_aggs", []):
|
|
if d.get("doc_id") in dseen:
|
|
continue
|
|
dseen.add(d.get("doc_id"))
|
|
tools.kbinfos.setdefault("doc_aggs", []).append(d)
|
|
|
|
|
|
def _chunk_key(ck: dict) -> str:
|
|
return ck.get("chunk_id") or ck.get("id") or str(id(ck))
|
|
|
|
|
|
def _has_chunks(tools) -> bool:
|
|
return bool(tools.kbinfos.get("chunks"))
|