fix(regulatory): 优先从附件字段识别适用条件

review_agent/regulatory_review/services/info_extract.py

@@ -1,6 +1,11 @@
 from __future__ import annotations
 
+from pathlib import Path
+
+from django.conf import settings
+
 from review_agent.models import FileSummaryBatch
+from review_agent.regulatory_review.services.text_extract import extract_text
 
 
 OPTION_FIELDS = {
@@ -14,9 +19,14 @@ def detect_regulatory_condition_candidates(summary_batch: FileSummaryBatch) -> d
     """Infers review-scope conditions from the summary batch and file names."""
 
     corpus_parts = [summary_batch.product_name or ""]
+    field_candidates: dict[str, str] = {}
     for item in summary_batch.items.order_by("file_index"):
         corpus_parts.extend([item.directory_level, item.file_name, item.relative_path])
+        extracted = _extract_item_fields(item)
+        field_candidates.update({key: value for key, value in extracted.items() if value and key not in field_candidates})
+        corpus_parts.extend(extracted.values())
     corpus = "\n".join(part for part in corpus_parts if part)
+    product_name = field_candidates.get("产品名称") or _safe_summary_product_name(summary_batch.product_name)
 
     return {
         "product_category": {
@@ -40,21 +50,42 @@ def detect_regulatory_condition_candidates(summary_batch: FileSummaryBatch) -> d
         "product_name": {
             "label": "产品名称",
             "input_type": "text",
-            "suggested": summary_batch.product_name or "",
+            "suggested": product_name,
         },
         "model_spec": {
             "label": "型号规格",
             "input_type": "text",
-            "suggested": "",
+            "suggested": field_candidates.get("型号规格", ""),
         },
         "intended_use": {
             "label": "预期用途",
             "input_type": "text",
-            "suggested": "",
+            "suggested": field_candidates.get("预期用途", ""),
         },
     }
 
 
+def _extract_item_fields(item) -> dict[str, str]:
+    path = Path(item.storage_path)
+    if not path.is_absolute():
+        path = Path(settings.MEDIA_ROOT) / item.storage_path
+    if not path.exists():
+        return {}
+    result = extract_text(path)
+    if result.status != "success" or not result.field_candidates:
+        return {}
+    return result.field_candidates
+
+
+def _safe_summary_product_name(product_name: str) -> str:
+    value = (product_name or "").strip()
+    if not value:
+        return ""
+    if any(keyword in value for keyword in ["第1章", "第2章", "监管信息", "综述资料", "非临床资料", "章节目录"]):
+        return ""
+    return value
+
+
 def _detect_product_category(corpus: str) -> str:
     if any(keyword in corpus for keyword in ["体外诊断", "检测试剂", "试剂盒", "IVD"]):
         return "体外诊断试剂"