将 jq output 格式化为表格

Question

我想从下面的 JSON 代码中获取一些数据：

我可以使用下面的命令获得 output，但现在我想以这样的方式格式化它，使我的 output 看起来像预期的 output。

命令：

cat dump | jq -r '["name","IP","NAT","location","method"], 
                  (.objects[] | [.name, ."ipv4-address", ."nat-settings"."ipv4-address", ."nat-settings"."install-on", ."nat-settings".method]) 
                              | @csv' 
                              | sed -e 's/"//g'

使用@csv 后，我得到了以下 output：

name,IP,NAT,location,method
H_103.109.135.25,103.109.135.25,1.1.1.1,All,static
H_103.109.135.250,103.109.135.250,,,

每当我使用@tsv 时，我都会收到“jq: error: tsv is not a valid format”

任何人都可以建议我如何实现低于 output 的目标：

预计 Output：

原始 JSON 代码：

{
  "from" : 1,
  "to" : 2,
  "total" : 2,
  "objects" : [ {
    "uid" : "73b7036d-e8ec-47b7-99b5-19ca89eb5fd0",
    "name" : "H_103.109.135.25",
    "type" : "host",
    "domain" : {
      "uid" : "41e821a0-3720-11e3-aa6e-0800200c9fde",
      "name" : "SMC User",
      "domain-type" : "domain"
    },
    "ipv4-address" : "103.109.135.25",
    "interfaces" : [ ],
    "nat-settings" : {
      "auto-rule" : true,
      "ipv4-address" : "1.1.1.1",
      "ipv6-address" : "",
      "install-on" : "All",
      "method" : "static"
    },
    "comments" : "",
    "color" : "black",
    "icon" : "Objects/host",
    "tags" : [ ],
    "meta-info" : {
      "lock" : "unlocked",
      "validation-state" : "ok",
      "last-modify-time" : {
        "posix" : 1674820459413,
        "iso-8601" : "2023-01-27T17:24+0530"
      },
      "last-modifier" : "admin",
      "creation-time" : {
        "posix" : 1674818326777,
        "iso-8601" : "2023-01-27T16:48+0530"
      },
      "creator" : "admin"
    },
    "read-only" : false,
    "available-actions" : {
      "edit" : "true",
      "delete" : "true",
      "clone" : "true"
    }
  }, {
    "uid" : "7300c38a-a496-497a-b9e3-5701fa081393",
    "name" : "H_103.109.135.250",
    "type" : "host",
    "domain" : {
      "uid" : "41e821a0-3720-11e3-aa6e-0800200c9fde",
      "name" : "SMC User",
      "domain-type" : "domain"
    },
    "ipv4-address" : "103.109.135.250",
    "interfaces" : [ ],
    "nat-settings" : {
      "auto-rule" : false
    },
    "comments" : "",
    "color" : "black",
    "icon" : "Objects/host",
    "tags" : [ ],
    "meta-info" : {
      "lock" : "unlocked",
      "validation-state" : "ok",
      "last-modify-time" : {
        "posix" : 1674818341888,
        "iso-8601" : "2023-01-27T16:49+0530"
      },
      "last-modifier" : "admin",
      "creation-time" : {
        "posix" : 1674818341888,
        "iso-8601" : "2023-01-27T16:49+0530"
      },
      "creator" : "admin"
    },
    "read-only" : false,
    "available-actions" : {
      "edit" : "true",
      "delete" : "true",
      "clone" : "true"
    }
  } ]
}

笔记：

仅使用 jq 将 output 打印在表格中并不是强制性的。 “awk”或“sed”也可以。

我从以下原始 json 数据中提取了所需的数据：

提取的数据：

{
    "name": "H_103.109.135.25",
    "IP": "103.109.135.25",
    "NAT": "1.1.1.1",
    "location": "All",
    "method": "static"
  },
  {
    "name": "H_103.109.135.250",
    "IP": "103.109.135.250",
    "NAT": "NA",
    "location": "NA",
    "method": "NA"
  }

我现在只需要将这些数据格式化成如下表或类似的表：

| name              | IP              | NAT     | location   | method   |
|-------------------|-----------------|---------|------------|----------|
| H_103.109.135.25  | 103.109.135.25  | 1.1.1.1 | All        | static   |
| H_103.109.135.250 | 103.109.135.250 | NA      | NA         | NA       |

Answer 1

有jbtl可以产生你正在寻找的东西。 例如，如果您在output.jq中有这个：

.objects
| map(
    { name, IP: ."ipv4-address" } +
    (."nat-settings" | {
      NAT: (."ipv4-address" // "NA"), 
      location: (."install-on" // "NA"), 
      method: (.method // "NA")
    })
  )

然后通过此过滤器传递数据并使用-m选项将其通过管道传输到jtbl ，如下所示：

cat dump | jq -f output.jq | jtbl -m

给这个

| name              | IP              | NAT     | location   | method   |
|-------------------|-----------------|---------|------------|----------|
| H_103.109.135.25  | 103.109.135.25  | 1.1.1.1 | All        | static   |
| H_103.109.135.250 | 103.109.135.250 | NA      | NA         | NA       |

Answer 2

miller可以方便地进行漂亮的打印 output。

echo 'name,IP,NAT,location,method
H_103.109.135.25,103.109.135.25,1.1.1.1,All,static
H_103.109.135.250,103.109.135.250,,,' \
| mlr --c2p --barred put 'for (i,v in $*) {if (v == "") {$[i] = "NA"}}'

--c2p是--icsv --opprint的快捷方式，它读取 CSV 输入并输出漂亮的表格形式。

+-------------------+-----------------+---------+----------+--------+
| name              | IP              | NAT     | location | method |
+-------------------+-----------------+---------+----------+--------+
| H_103.109.135.25  | 103.109.135.25  | 1.1.1.1 | All      | static |
| H_103.109.135.250 | 103.109.135.250 | NA      | NA       | NA     |
+-------------------+-----------------+---------+----------+--------+

miller put动词采用类似 awk 的脚本。

参见https://miller.readthedocs.io/en/latest/

---

更实用的风格：

mlr --c2p --barred put '$* = apply($*, func(k,v) {return {k: v == "" ? "NA" : v}})'

---

我建议删除引号并在 jq 中添加“NA”，然后将 pipe 和 output 添加到column中

jq -r '
  [
    ["name","IP","NAT","location","method"],
    ( .objects[]
    | {"nat-settings": {"ipv4-address": "NA", "install-on": "NA", method: "NA"}} * .
    | [.name, ."ipv4-address"] + (."nat-settings" | [."ipv4-address", ."install-on", .method])
    )
  ][] | join(",")
' dump | column -s, -t

假设“nat-settings”object缺少“ipv4-address”等键。

Answer 3

我建议使用 jq 的@tsv和非常标准的工具column ，例如如下：

< dump jq -r '
  ["name","IP","NAT","location","method"], 
  (.objects[] | [.name, ."ipv4-address", ."nat-settings"."ipv4-address", ."nat-settings"."install-on", ."nat-settings".method]) 
  | @tsv' | column -t