输出包含排序内容的整个JSON文档

是否可以对json文档的特定数组进行排序，并输出整个文档，其中包含排序后的内容，否则将不被触及

例如，如果我有这样的东西：

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Sid": "BLAHBLAHBLAH",
      "Effect": "Allow",
      "Principal": {
        "AWS": [
          "arn:aws:iam::346654243625:root",
          "arn:aws:iam::984895836564:root",
          "arn:aws:iam::636877599363:root",
          "arn:aws:iam::577836285792:root",
          "arn:aws:iam::836666644595:root",
          "arn:aws:iam::652379234777:root",
          "arn:aws:iam::339659563489:root",
          "arn:aws:iam::293576423935:root",
          "arn:aws:iam::682354689262:root",
          "arn:aws:iam::349855857558:root",
          "arn:aws:iam::398259495958:root",
          "arn:aws:iam::735277384553:root"
        ]
      },
      "Action": "s3:*",
      "Resource": [
        "arn:aws:s3:::blah-blah-blah",
        "arn:aws:s3:::blah-blah-blah/*"
      ]
    }
  ]
}

我想拿回这个：

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Sid": "BLAHBLAHBLAH",
      "Effect": "Allow",
      "Principal": {
        "AWS": [
          "arn:aws:iam::293576423935:root",
          "arn:aws:iam::339659563489:root",
          "arn:aws:iam::346654243625:root",
          "arn:aws:iam::349855857558:root",
          "arn:aws:iam::398259495958:root",
          "arn:aws:iam::577836285792:root",
          "arn:aws:iam::636877599363:root",
          "arn:aws:iam::652379234777:root",
          "arn:aws:iam::682354689262:root",
          "arn:aws:iam::735277384553:root",
          "arn:aws:iam::836666644595:root",
          "arn:aws:iam::984895836564:root"
        ]
      },
      "Action": "s3:*",
      "Resource": [
        "arn:aws:s3:::blah-blah-blah",
        "arn:aws:s3:::blah-blah-blah/*"
      ]
    }
  ]
}

我看到过使用

jq

和

JMESPath

进行排序的示例，但只看到它们输出被排序的部分，而没有输出带有排序部分的整个文档。正在寻求有关这一问题的指导，并将对任何有效的解决方案感到满意

---

这就是我用来排序数组

|jq'.Statement[].Principal.AWS | sort'

的方法，它是有效的，但我想回到整个文档的上下文中，它的格式是一样的。

对于jq处理器来说，这是一项很好的工作：

jq '.Statement[0].Principal.AWS |= sort' file

• .Statement[0].Principal.AWS

  -所选项目

• |=

  -更新语句/操作（更新所选项目）

---

输出：

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Sid": "BLAHBLAHBLAH",
      "Effect": "Allow",
      "Principal": {
        "AWS": [
          "arn:aws:iam::293576423935:root",
          "arn:aws:iam::339659563489:root",
          "arn:aws:iam::346654243625:root",
          "arn:aws:iam::349855857558:root",
          "arn:aws:iam::398259495958:root",
          "arn:aws:iam::577836285792:root",
          "arn:aws:iam::636877599363:root",
          "arn:aws:iam::652379234777:root",
          "arn:aws:iam::682354689262:root",
          "arn:aws:iam::735277384553:root",
          "arn:aws:iam::836666644595:root",
          "arn:aws:iam::984895836564:root"
        ]
      },
      "Action": "s3:*",
      "Resource": [
        "arn:aws:s3:::blah-blah-blah",
        "arn:aws:s3:::blah-blah-blah/*"
      ]
    }
  ]
}

---

要对多个阵列进行排序，请使用：

jq '.Statement[].Principal.AWS |= sort' file

---

好极了这确实有效，你愿意把它分解并解释它是如何实现的吗？@Trytry再次，当然，请参阅我的解释两条评论：1。您不必将其限制在第一条语句中，您可以将其保留为

.statement[].Principal…

，2。当不需要排序时，初始的

可以是

排序

.Statement[].Principal.AWS |=sort

@trytryreach:这里的关键是，因为这里应用的过滤器是一个赋值，所以它不会更改上下文。因此，我们不是将范围缩小到每个语句的单个

AWS

属性，然后进行排序，而是选择单个AWS属性并通过排序进行更新。@JeffMercado…很有意义，我总是喜欢额外的解释，喜欢能够删掉任何不必要的内容。确认和高兴：）谢谢