Implement should_accept_transaction(transaction, rules) for a fraud-rule engine.
Part 1 — String equality rules
Each transaction has string fields, such as merchant name and card brand.
Each ordered rule is either ACCEPT or BLOCK.
A rule has one equality condition, such as (merchant_name: "Not A Fraudster").
Apply rules in order; the first matching rule decides the result.
If no rule matches, accept the transaction by default.
Part 2 — Boolean, AND, and OR rules
Extend the same fraud-rule engine to support Boolean transaction fields.
A rule condition may be one Boolean variable, such as (:known_stolen_card:).
It may combine two Boolean variables with AND or OR, such as (:large_amount: AND :from_different_country:).
Keep Part 1 string-equality behavior working.
Rules are still evaluated in order, and no matching rule means accept.
这道题要求实现一个按顺序匹配的欺诈规则引擎：给定交易信息和一组规则，依次判断每条规则是否命中，命中后立即返回对应的 ACCEPT 或 BLOCK，若全部规则都不匹配，则默认接受交易。第一部分只涉及字符串字段的等值匹配，例如 merchant_name 与 card_brand；第二部分扩展到布尔字段，并支持单个布尔条件以及两个布尔条件的 AND / OR 组合。核心思路是把每条规则抽象成可求值的条件表达式，再按顺序线性扫描规则即可。
