arrays - jq - 为数组上的值添加前缀

Question

输入json:

[
{
  "login": "u1",
  "name": "u1",
  "role": "User",
  "groups": [
    {
      "id": "1234",
      "name": "G1"
    },
    {
      "id": "1235",
      "name": "G2"
    }
  ],
  "created": "2020-05-11 11:06"
},
{
  "login": "u2",
  "name": "u2",
  "role": "User",
  "groups": null,
  "created": "2020-05-11 11:06"
}
]

我使用以下过滤器来获取他们所属的用户和组:

$ jq -r '
  .[]
  | [.login,
     .name,
     ( if .groups == null 
       then "grp:-" 
       else (del(.groups[]
             | select(.name=="All Users"))
            | [.groups[].name] | join("|")) 
       end )]
  | @tsv' json

u1      u1      G1|G2
u2      u2      grp:-

如何为每个找到的组添加 grp: 前缀？

预期输出为:

u1      u1      grp:G1|grp:G2
u2      u2      grp:-

Answer 1

使用+连接字符串:

jq -r '.[]
       | [.login,.name,(
           if .groups == null
           then "grp:-"
           else (
               del(.groups[] | select(.name=="All Users"))
               | ["grp:" + .groups[].name]
#                 ~~~~~~~~
               | join("|")
           )
           end
      )]
      | @tsv' 

您甚至可以避免指定“grp:”两次，但您也需要从 then 分支返回一个数组:

jq -r '.[]
       | [.login, .name, (
           (
               if .groups == null
               then ["-"]
               else (
                   del(.groups[] | select(.name=="All Users"))
                   | [.groups[].name]
               )
               end
           )
           | map("grp:" + .)
           | join("|")
       )]
       | @tsv'

更新: 从峰值中获取灵感我想这会产生您想要的输出:

jq -r '.[]
        | .groups //= []
        | del(.groups[] | select(.name=="All Users"))
        | .groups[0] //= {name: "-"}
        | [.login, .name, ( ["grp:" + (.groups[].name)] | join("|") ) ]
        | @tsv'