如何用jq遍历字符串正则匹配结果并生成关联控制数组？

需求说明 #

从JSON的description字段中提取"Related controls:"后的控制字符串，转移到新的relatedControls数组中，并移除原description里的对应内容。

输入JSON #

{"description": "Fail-safe procedures include, for example, alerting operator personnel and providing specific instructions on subsequent steps to take (e.g., do nothing, re-establish system settings, shut down processes, restart the system, or contact designated organizational personnel). Related controls: CA-2, CA-7, CM-3, CM-5, CM-8, MA-2, IR-4, RA-5, SA-10, SA-1x, SI-1x"}

期望输出JSON #

{"description": "Fail-safe procedures include, for example, alerting operator personnel and providing specific instructions on subsequent steps to take (e.g., do nothing, re-establish system settings, shut down processes, restart the system, or contact designated organizational personnel).",
"relatedControls": ["CA-2", "CA-7", "CM-3", "CM-5", "CM-8", "MA-2", "IR-4", "RA-5", "SA-10", "SA-1x", "SI-1x"]}

原命令问题分析 #

原命令的核心问题是正则表达式匹配逻辑不严谨：
原正则[A-Z]{2}-\\d[0-9x]?仅匹配"单个数字+可选单个数字/x"的组合，虽能覆盖当前示例，但对于更长数字后缀的控制符（如AB-123）会无法完整匹配，且存在拆分匹配结果的风险。

修正方案 #

方案1：精准捕获拆分（推荐） #

通过capture直接定位并拆分目标内容，避免全局扫描可能带来的误匹配：

jq '
.description |= capture("(?<desc>.*?) Related controls: (?<controls>.*)") |
.description = .desc |
.relatedControls = (.controls | split(", ") | map(trim))
'

命令解释：

1. capture将description拆分为两部分：desc（控制符前的纯描述文本）和controls（控制符列表字符串）；
2. 替换原description字段为纯描述内容；
3. 将控制符字符串按, 分割为数组，并用map(trim)清理可能存在的首尾空格。

方案2：修正正则后的全局扫描 #

如果偏好使用scan，可调整正则表达式以覆盖所有目标格式：

jq '
{
  description: .description | sub(" Related controls:.*"; ""),
  relatedControls: .description | scan("[A-Z]{2}-\\d+x?")
}
'

正则说明：

[A-Z]{2}-\\d+x?表示：两位大写字母 + 连字符 + 一个或多个数字 + 可选的x结尾，能准确匹配所有目标控制符。

验证命令 #

echo '{"description": "Fail-safe procedures include, for example, alerting operator personnel and providing specific instructions on subsequent steps to take (e.g., do nothing, re-establish system settings, shut down processes, restart the system, or contact designated organizational personnel). Related controls: CA-2, CA-7, CM-3, CM-5, CM-8, MA-2, IR-4, RA-5, SA-10, SA-1x, SI-1x"}' | jq '
.description |= capture("(?<desc>.*?) Related controls: (?<controls>.*)") |
.description = .desc |
.relatedControls = (.controls | split(", ") | map(trim))
'

内容的提问来源于stack exchange，提问作者mikewilliamson