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

review_agent/application_form_fill/services/field_extract.py

@@ -17,11 +17,11 @@ from review_agent.regulatory_review.services.text_extract import extract_text
 
 FIELD_ALIASES = {
     "product_name": ["产品名称"],
-    "applicant_name": ["注册人名称", "生产企业名称", "企业名称", "生产企业"],
-    "applicant_address": ["注册人住所", "生产企业住所", "企业住所", "住所"],
+    "applicant_name": ["注册人名称", "申请人名称", "生产企业名称"],
+    "applicant_address": ["注册人住所", "申请人住所", "生产企业住所"],
     "manufacturer_address": ["生产地址", "生产企业地址", "生产场所"],
-    "agent_name": ["代理人名称", "生产企业名称", "企业名称", "生产企业", "注册人名称"],
-    "agent_address": ["代理人住所", "生产企业住所", "企业住所", "住所", "注册人住所"],
+    "agent_name": ["代理人名称", "生产企业名称", "注册人名称", "申请人名称"],
+    "agent_address": ["代理人住所", "生产企业住所", "注册人住所", "申请人住所"],
     "package_specification": ["包装规格", "规格"],
     "main_components": ["主要组成成分", "主要组成", "组成成分"],
     "intended_use": ["预期用途"],
@@ -35,6 +35,14 @@ STATIC_STOP_LABELS = [
     "注",
     "保证书",
     "应附资料",
+    "优先通道申请",
+    "分类编码",
+    "医疗器械唯一标识",
+    "注册产品目前是否",
+    "临床评价路径",
+    "临床试验",
+    "其他需要说明的问题",
+    "国家药监局器审中心医疗器械",
 ]
 
 

review_agent/application_form_fill/services/summary.py

@@ -22,10 +22,11 @@ def build_assistant_summary(batch: ApplicationFormFillBatch, exports: list[Expor
         lines.extend(["", "| 冲突字段 | 采用值 | 冲突来源 | 处理 |", "| --- | --- | --- | --- |"])
         for item in conflicts:
             conflict_sources = "；".join(
-                f"{value.get('source_file', '')}：{value.get('value', '')}" for value in item.get("conflict_values", [])
+                f"{_compact_table_text(value.get('source_file', ''))}：{_compact_table_text(value.get('value', ''))}"
+                for value in item.get("conflict_values", [])
             )
             lines.append(
-                f"| {item.get('field_label', item.get('field_key', ''))} | {item.get('selected_value', '')} | {conflict_sources or '-'} | {item.get('handling', '')} |"
+                f"| {_compact_table_text(item.get('field_label', item.get('field_key', '')))} | {_compact_table_text(item.get('selected_value', ''))} | {_compact_table_text(conflict_sources or '-')} | {_compact_table_text(item.get('handling', ''))} |"
             )
 
     if trace_exports:
@@ -33,3 +34,10 @@ def build_assistant_summary(batch: ApplicationFormFillBatch, exports: list[Expor
         for export in trace_exports:
             lines.append(f"[下载{export.file_name}](/api/review-agent/file-summary/exports/{export.pk}/download/)")
     return "\n".join(lines).strip()
+
+
+def _compact_table_text(value: object, *, limit: int = 80) -> str:
+    text = " ".join(str(value or "").replace("|", " ").split())
+    if len(text) <= limit:
+        return text
+    return f"{text[:limit]}..."

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