fix(application-form-fill): 过滤申请表噪声冲突内容

tests/test_application_form_fill_field_extract.py

@@ -124,6 +124,26 @@ def test_rule_stops_product_name_before_application_form_instructions():
     assert "填表说明" not in values["product_name"]
 
 
+def test_rule_ignores_generic_enterprise_name_from_application_form():
+    texts = {
+        "CH1.4 申请表.docx": "\n".join(
+            [
+                "注册人制度\t是  企业名称：否",
+                "优先通道申请   应急通道    同品种首个产品首次申报",
+                "临床试验",
+                "临床试验机构名称： 中国医学科学院北京协和医院、晋中市第一人民医院",
+                "应附资料",
+            ]
+        )
+    }
+
+    result = extract_by_rules(texts, _registration_specs())
+
+    values = {field["key"]: field["value"] for field in result["fields"]}
+    assert "applicant_name" not in values
+    assert "agent_name" not in values
+
+
 def test_llm_extract_parses_structured_json(monkeypatch):
     monkeypatch.setattr(
         "review_agent.application_form_fill.services.field_extract.generate_completion",

tests/test_application_form_fill_summary.py

@@ -0,0 +1,39 @@
+import pytest
+
+from review_agent.application_form_fill.services.summary import build_assistant_summary
+from review_agent.models import ApplicationFormFillBatch, Conversation, FileSummaryBatch
+
+
+pytestmark = pytest.mark.django_db
+
+
+def test_assistant_summary_compacts_long_conflict_values(django_user_model):
+    user = django_user_model.objects.create_user(username="owner", password="pass")
+    conversation = Conversation.objects.create(user=user, title="会话")
+    summary = FileSummaryBatch.objects.create(conversation=conversation, user=user, batch_no="FS-SUMMARY")
+    batch = ApplicationFormFillBatch.objects.create(
+        conversation=conversation,
+        user=user,
+        source_summary_batch=summary,
+        batch_no="AFF-SUMMARY",
+        conflict_summary=[
+            {
+                "field_key": "applicant_name",
+                "field_label": "注册人名称",
+                "selected_value": "卡尤迪生物科技宜兴有限公司",
+                "conflict_values": [
+                    {
+                        "source_file": "CH1.4 申请表.docx",
+                        "value": "否\n临床试验\n临床试验机构名称： 中国医学科学院北京协和医院、晋中市第一人民医院、北京市疾病预防控制中心 临床数据库.zip\n应附资料",
+                    }
+                ],
+                "handling": "说明书优先，模板内黄底红字高亮",
+            }
+        ],
+    )
+
+    content = build_assistant_summary(batch, [])
+
+    assert "临床试验机构名称" in content
+    assert len([line for line in content.splitlines() if "临床试验机构名称" in line][0]) < 220
+    assert "\n临床试验\n" not in content