baodan/api/insurance/poster/compliance.py

"""海报文案合规规则与字符定位。"""
import hashlib
import json


RULES = (
    {
        "ruleId": "RETURN_GUARANTEE",
        "terms": ("保证", "稳赚", "保底"),
        "severity": "block",
        "message": "不得使用确定性收益承诺",
        "suggestion": "改为“利益以正式计划书为准”",
        "replacement": "相关利益",
    },
    {
        "ruleId": "RISK_FREE_CLAIM",
        "terms": ("无风险", "零风险"),
        "severity": "block",
        "message": "不得宣称保险或投资安排不存在风险",
        "suggestion": "删除绝对化表述，并补充必要的风险提示",
        "replacement": "存在一定风险",
    },
    {
        "ruleId": "EXAGGERATED_RETURN",
        "terms": ("最高收益", "回报率"),
        "severity": "block",
        "message": "收益表述必须有正式计划书依据且不得夸大",
        "suggestion": "引用已确认的计划书数据并注明非保证利益",
        "replacement": "演示利益",
    },
    {
        "ruleId": "UNSUPPORTED_SUPERLATIVE",
        "terms": ("行业领先", "最佳"),
        "severity": "warn",
        "message": "比较性或最高级表述需要可核验依据",
        "suggestion": "改为客观描述产品特点，或补充权威依据",
        "replacement": "具有特色",
    },
)

FIELDS = ("headline", "body", "call_to_action")


def copy_revision(copy_content: dict) -> str:
    """生成与字段顺序无关的文案版本摘要。"""
    normalized = {
        field: str((copy_content or {}).get(field) or "")
        for field in FIELDS
    }
    payload = json.dumps(normalized, ensure_ascii=False, sort_keys=True)
    return hashlib.sha256(payload.encode("utf-8")).hexdigest()


def check_copy_compliance(copy_content: dict) -> dict:
    """返回合规状态、文案版本和可定位的问题列表。"""
    issues = []
    for field in FIELDS:
        text = str((copy_content or {}).get(field) or "")
        for rule in RULES:
            for term in rule["terms"]:
                start = 0
                while True:
                    index = text.find(term, start)
                    if index < 0:
                        break
                    issues.append({
                        "ruleId": rule["ruleId"],
                        "severity": rule["severity"],
                        "field": field,
                        "start": index,
                        "end": index + len(term),
                        "text": term,
                        "message": rule["message"],
                        "suggestion": rule["suggestion"],
                        "replacement": rule["replacement"],
                    })
                    start = index + len(term)

    severity_order = {"pass": 0, "warn": 1, "block": 2}
    status = "pass"
    for issue in issues:
        if severity_order[issue["severity"]] > severity_order[status]:
            status = issue["severity"]
    return {
        "status": status,
        "revision": copy_revision(copy_content),
        "issues": issues,
    }