如何依据input.json的键顺序对output.json进行排序？

我用翻译管理工具处理多语言文件时遇到个问题：上传input.json后，译者完成翻译导出的output.json键和原文件完全一致，但键的顺序被打乱了。在Pull Request里对比新旧文件时，键顺序差异会产生大量无效diff，增加评审难度。我知道可以用jq按字母序排序两者，但能不能直接复用input.json的键顺序来重新排列output.json？

演示数据 #

input.json

{
  "BathroomType": "Type of bathroom",
  "BuildingType": "Type of building",
  "enums": {
    "OrientationGarden": {
      "EAST": "East",
      "SOUTH": "South",
      "NORTH": "North",
      "WEST": "West"
    },
    "TerrainRecentDestination": {
      "EXTRACTION": "extraction",
      "PARK": "park"
    }
  }
}

output.json

{
  "BathroomType": "Type of bathroom2",
  "enums": {
    "TerrainRecentDestination": {
      "EXTRACTION": "extraction",
      "PARK": "park"
    },
    "OrientationGarden": {
      "SOUTH": "South",
      "EAST": "East2",
      "NORTH": "North",
      "WEST": "West"
    }
  },
  "BuildingType": "Type of building"
}

解决方案 #

可以用jq提取input.json的键路径顺序，然后按这个顺序重构output.json，命令如下：

jq --argfile input input.json '($input | paths(scalars) | map(select(type=="string"))) as $paths | reduce $paths[] as $p (.; setpath($p; getpath($p)))' output.json > sorted_output.json

命令解释

1. ($input | paths(scalars) | map(select(type=="string")))：提取input.json中所有标量值的键路径（比如["BathroomType"]、["enums", "OrientationGarden", "EAST"]），确保只保留字符串类型的键，避免处理数组索引（如果有的话）。
2. reduce $paths[] as $p (.; setpath($p; getpath($p)))：遍历提取到的路径数组，按顺序从output.json中取出对应路径的值，重新构建新的JSON对象，完全复用input.json的键层级和顺序。

执行结果 #

生成的sorted_output.json会和input.json的键顺序完全一致，同时保留output.json的译文内容：

{
  "BathroomType": "Type of bathroom2",
  "BuildingType": "Type of building",
  "enums": {
    "OrientationGarden": {
      "EAST": "East2",
      "SOUTH": "South",
      "NORTH": "North",
      "WEST": "West"
    },
    "TerrainRecentDestination": {
      "EXTRACTION": "extraction",
      "PARK": "park"
    }
  }
}

内容的提问来源于stack exchange，提问作者Philimmo