如何用jq统计JSON数组元素数量（排除指定元素）

需求说明 #

用jq工具统计每个对象中snapshots数组的元素数量，需排除其中name字段值为"current"的元素，最终输出每个对象的name对应符合条件的数量。

给定JSON数据 #

{
  "name": "test1",
  "snapshots": [
    {
      "description": "You are here!",
      "name": "current",
      "digest": "ccc",
      "running": 1
    },
    {
      "description": "You are here!",
      "name": "snap1",
      "digest": "ddd",
      "running": 1
    }
  ]
}
{
  "name": "test2",
  "snapshots": [
    {
      "digest": "bbb",
      "description": "You are here!",
      "name": "current",
      "running": 1
    },
    {
      "description": "You are here!",
      "name": "snap1",
      "digest": "aaa",
      "running": 1
    },
    {
      "description": "You are here!",
      "name": "snap2",
      "digest": "abc",
      "running": 1
    }
  ]
}

期望输出 #

test1: 1
test2: 2

jq命令实现 #

执行以下jq命令即可完成需求：

jq -r '.name + ": " + ( [.snapshots[] | select(.name != "current")] | length | tostring )'

命令解释 #

• .name: 获取当前对象的name字段值
• [.snapshots[] | select(.name != "current")]: 过滤snapshots数组，仅保留name不等于"current"的元素并重组为新数组
• length: 计算过滤后数组的元素数量
• tostring: 将数字转为字符串，实现和name字段值的拼接
• -r: 输出原始字符串（而非JSON格式字符串），让结果匹配期望的输出样式

验证结果 #

将给定的JSON数据作为输入执行上述命令，会直接输出：

test1: 1
test2: 2

内容的提问来源于stack exchange，提问作者secdave