feat(regulatory): 增加法规核查基础服务

review_agent/regulatory_review/services/consistency_check.py

@@ -0,0 +1,41 @@
+from __future__ import annotations
+
+import re
+from collections import defaultdict
+
+from review_agent.regulatory_review.schemas import Finding
+
+
+FIELDS = {
+    "产品名称": r"产品名称[:：]\s*([^\n\r]+)",
+    "型号规格": r"型号规格[:：]\s*([^\n\r]+)",
+    "预期用途": r"预期用途[:：]\s*([^\n\r]+)",
+}
+
+
+def run_consistency_check(document_texts: dict[str, str]) -> list[Finding]:
+    findings: list[Finding] = []
+    for label, pattern in FIELDS.items():
+        values: dict[str, list[str]] = defaultdict(list)
+        for file_name, text in document_texts.items():
+            match = re.search(pattern, text)
+            if match:
+                values[_normalize(match.group(1))].append(file_name)
+        if len(values) <= 1:
+            continue
+        findings.append(
+            Finding(
+                rule_code=f"consistency:{label}",
+                category="consistency",
+                severity="high",
+                title=f"{label}在不同文件中不一致",
+                detail=f"发现 {len(values)} 个不同的{label}取值。",
+                suggestion=f"请统一各注册资料中的{label}。",
+                evidence={"field": label, "values": dict(values)},
+            )
+        )
+    return findings
+
+
+def _normalize(value: str) -> str:
+    return " ".join(value.strip().split())