diff --git a/agent/tools/retrieval.py b/agent/tools/retrieval.py index c3a01e517..fd9096cb6 100644 --- a/agent/tools/retrieval.py +++ b/agent/tools/retrieval.py @@ -137,7 +137,7 @@ class Retrieval(ToolBase, ABC): if not doc_ids: doc_ids = None elif self._param.meta_data_filter.get("method") == "manual": - filters=self._param.meta_data_filter["manual"] + filters = self._param.meta_data_filter["manual"] for flt in filters: pat = re.compile(self.variable_ref_patt) s = flt["value"] @@ -166,8 +166,8 @@ class Retrieval(ToolBase, ABC): out_parts.append(s[last:]) flt["value"] = "".join(out_parts) doc_ids.extend(meta_filter(metas, filters, self._param.meta_data_filter.get("logic", "and"))) - if not doc_ids: - doc_ids = None + if filters and not doc_ids: + doc_ids = ["-999"] if self._param.cross_languages: query = cross_languages(kbs[0].tenant_id, None, query, self._param.cross_languages) diff --git a/api/apps/chunk_app.py b/api/apps/chunk_app.py index e121bcba7..b43fb9af1 100644 --- a/api/apps/chunk_app.py +++ b/api/apps/chunk_app.py @@ -311,8 +311,8 @@ async def retrieval_test(): doc_ids = None elif meta_data_filter.get("method") == "manual": doc_ids.extend(meta_filter(metas, meta_data_filter["manual"], meta_data_filter.get("logic", "and"))) - if not doc_ids: - doc_ids = None + if meta_data_filter["manual"] and not doc_ids: + doc_ids = ["-999"] try: tenants = UserTenantService.query(user_id=current_user.id) diff --git a/api/apps/sdk/doc.py b/api/apps/sdk/doc.py index 84300ac3c..30fbd835e 100644 --- a/api/apps/sdk/doc.py +++ b/api/apps/sdk/doc.py @@ -1445,6 +1445,8 @@ async def retrieval_test(tenant_id): metadata_condition = req.get("metadata_condition", {}) or {} metas = DocumentService.get_meta_by_kbs(kb_ids) doc_ids = meta_filter(metas, convert_conditions(metadata_condition), metadata_condition.get("logic", "and")) + if metadata_condition and not doc_ids: + doc_ids = ["-999"] similarity_threshold = float(req.get("similarity_threshold", 0.2)) vector_similarity_weight = float(req.get("vector_similarity_weight", 0.3)) top = int(req.get("top_k", 1024)) diff --git a/api/apps/sdk/session.py b/api/apps/sdk/session.py index 533375622..074401ede 100644 --- a/api/apps/sdk/session.py +++ b/api/apps/sdk/session.py @@ -446,8 +446,8 @@ async def agent_completions(tenant_id, agent_id): if req.get("stream", True): - def generate(): - for answer in agent_completion(tenant_id=tenant_id, agent_id=agent_id, **req): + async def generate(): + async for answer in agent_completion(tenant_id=tenant_id, agent_id=agent_id, **req): if isinstance(answer, str): try: ans = json.loads(answer[5:]) # remove "data:" @@ -471,7 +471,7 @@ async def agent_completions(tenant_id, agent_id): full_content = "" reference = {} final_ans = "" - for answer in agent_completion(tenant_id=tenant_id, agent_id=agent_id, **req): + async for answer in agent_completion(tenant_id=tenant_id, agent_id=agent_id, **req): try: ans = json.loads(answer[5:]) @@ -873,7 +873,7 @@ async def agent_bot_completions(agent_id): resp.headers.add_header("Content-Type", "text/event-stream; charset=utf-8") return resp - for answer in agent_completion(objs[0].tenant_id, agent_id, **req): + async for answer in agent_completion(objs[0].tenant_id, agent_id, **req): return get_result(data=answer) @@ -981,8 +981,8 @@ async def retrieval_test_embedded(): doc_ids = None elif meta_data_filter.get("method") == "manual": doc_ids.extend(meta_filter(metas, meta_data_filter["manual"], meta_data_filter.get("logic", "and"))) - if not doc_ids: - doc_ids = None + if meta_data_filter["manual"] and not doc_ids: + doc_ids = ["-999"] try: tenants = UserTenantService.query(user_id=tenant_id) diff --git a/api/db/services/dialog_service.py b/api/db/services/dialog_service.py index db878574d..0a09ea532 100644 --- a/api/db/services/dialog_service.py +++ b/api/db/services/dialog_service.py @@ -415,9 +415,10 @@ def chat(dialog, messages, stream=True, **kwargs): if not attachments: attachments = None elif dialog.meta_data_filter.get("method") == "manual": - attachments.extend(meta_filter(metas, dialog.meta_data_filter["manual"], dialog.meta_data_filter.get("logic", "and"))) - if not attachments: - attachments = None + conds = dialog.meta_data_filter["manual"] + attachments.extend(meta_filter(metas, conds, dialog.meta_data_filter.get("logic", "and"))) + if conds and not attachments: + attachments = ["-999"] if prompt_config.get("keyword", False): questions[-1] += keyword_extraction(chat_mdl, questions[-1]) @@ -787,8 +788,8 @@ def ask(question, kb_ids, tenant_id, chat_llm_name=None, search_config={}): doc_ids = None elif meta_data_filter.get("method") == "manual": doc_ids.extend(meta_filter(metas, meta_data_filter["manual"], meta_data_filter.get("logic", "and"))) - if not doc_ids: - doc_ids = None + if meta_data_filter["manual"] and not doc_ids: + doc_ids = ["-999"] kbinfos = retriever.retrieval( question=question, @@ -862,8 +863,8 @@ def gen_mindmap(question, kb_ids, tenant_id, search_config={}): doc_ids = None elif meta_data_filter.get("method") == "manual": doc_ids.extend(meta_filter(metas, meta_data_filter["manual"], meta_data_filter.get("logic", "and"))) - if not doc_ids: - doc_ids = None + if meta_data_filter["manual"] and not doc_ids: + doc_ids = ["-999"] ranks = settings.retriever.retrieval( question=question,