jq命令的基本介绍

# 常用命令-jq

jq (opens new window) 是个非常强大的处理 JSON 数据的命令行工具，本文主要讲的是常用的 <jq filter> ，大部分操作可以通过 jq play (opens new window) 直接在线测试。

# 一、基本操作

jq 的命令行使用方法如下:

$ jq -h
# jq - commandline JSON processor [version 1.5]
# Usage: jq [options] <jq filter> [file...]

1
2
3

# 1 数据输出

# 1.1 输出原始的 JSON 数据

默认不指定 filter 就会原样输出，也可以使用 . 过滤器

echo '{"url": "[mozillazg.com](http://mozillazg.com/)"}' |jq .
# {
#   "url": "[mozillazg.com](http://mozillazg.com/)"
# }

1
2
3
4

# 1.2 紧凑（compact）输出

echo '{
  "b": "hello",
  "d": "hello",
  "a": "hello",
  "c": "hello"
}'  | jq --compact-output
# {"b":"hello","d":"hello","a":"hello","c":"hello"}

1
2
3
4
5
6
7

echo '{"k1":"v1","k2":"v2"}' | jq   '[.k1, .k3 // ""]' 
# [
#   "v1",
#   ""
# ]
echo '{"k1":"v1","k2":"v2"}' | jq  -c '[.k1, .k3 // ""]'
# ["v1",""]

1
2
3
4
5
6
7

# 1.3 按照 Key 排序删除

echo '{
  "b": "hello",
  "d": "hello",
  "a": "hello",
  "c": "hello"
}' | jq --sort-keys
# {
#   "a": "hello",
#   "b": "hello",
#   "c": "hello",
#   "d": "hello"
# }

1
2
3
4
5
6
7
8
9
10
11
12

# 2 object 操作

# 2.1 获取某个 key 的值

- `.key, .foo.bar, .["key"]`

echo '{"url": "[mozillazg.com](http://mozillazg.com/)"}' |jq .url
# "[mozillazg.com](http://mozillazg.com/)"
echo '{"url": "[mozillazg.com](http://mozillazg.com/)"}' | jq '.["url"]'
# "[mozillazg.com](http://mozillazg.com/)"

1
2
3
4
5
6

可以在后面加个问号表示当输入不是 object 时不会报错 .key? :

echo '1' |jq '.a'
# jq: error (at <stdin>:1): Cannot index number with string "a"
echo '1' |jq '.a?'

1
2
3
4

? 规则适合所有正确的 filter，在 filter 最后加上 ? 可以忽略错误信息

# 2.2 取某几个 key 的值

echo '{"foo": 42, "bar": "something else", "baz": true}' | jq '.foo, .bar'
# 42
# "something else"

echo '{"foo": 42, "bar": "something else", "baz": true}' | jq '[.foo, .bar]'
# [
#   42,
#   "something else"
# ]

echo '{"foo": 42, "bar": "something else", "baz": true}' | jq -c '[.foo, .bar]'
# [42,"something else"]

echo '{"foo": 42, "bar": "something else", "baz": true}' | jq -c '{foo: .foo, bar:.bar}'
# {"foo":42,"bar":"something else"}

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

# 2.3 所有的 key 组成的数组

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' |jq keys
# [
#   "name",
#   "url"
# ]

1
2
3
4
5

# 2.4 所有的 value

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' |jq .[]
# "[mozillazg.com](http://mozillazg.com/)"
# "mozillazg"

1
2
3

# 2.5 所有的 value 组成的数组

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' |jq [.[]]
# [
#   "[mozillazg.com](http://mozillazg.com/)",
#   "mozillazg"
# ]

1
2
3
4
5

# 2.6 添加某个字段

# 新增：
echo '{"hello":"world"}' | jq '. +={"debug":1}'
# {
#   "hello": "world",
#   "debug": 1
# }

# 覆盖
echo '{"hello":"world", "debug":"a"}' | jq '. +={"debug":1}'
# {
#   "hello": "world",
#   "debug": 1
# }

# 引用变量
$ debug=100;echo '{"hello":"world", "debug":"a"}' | jq '. +={"debug":'"$debug"'}'
# {
#   "hello": "world",
#   "debug": 100
# }

$ debug=100;echo '{"hello":"world", "debug":"a"}' | jq '. +={"debug":"'"$debug"'"}'
# {
#   "hello": "world",
#   "debug": "100"
# }

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26

# Add field
echo '{"hello": "world"}' | jq --arg foo bar '. + {foo: $foo}'
# {
#   "hello": "world",
#   "foo": "bar"
# }

# Override field value
echo '{"hello": "world"}' | jq --arg foo bar '. + {hello: $foo}'
# {
#   "hello": "bar"
# }

# Concat and add
echo '{"hello": "world"}' | jq --arg foo bar '. + {hello: ("not" + $foo)}'
# {
#   "hello": "world",
#   "foo": "notbar"
# }

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

# 2.7 删除某个字段

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' | jq 'del(.name)'
# {
#   "url": "[mozillazg.com](http://mozillazg.com/)"
# }

1
2
3
4

# 2.8 某字段不存在则使用默认值

echo '{"k1":"v1","k2":"v2"}' | jq '.k1, .k3 // ""'
# "v1"
# ""

1
2
3

# 2.9 判断某字段是否存在

echo '{"k1":"v1","k2":"v2"}' | jq 'has("k1"),has("k3")'
# true
# false

echo '{"user":{"profile":{"age":20,"gender":"M"}}}' | jq '.user.profile | has("age")'
# true

echo '{"user":{"profile":{"age":20,"gender":"M"}}}' | jq '.user.profile | has("address")'
# false

1
2
3
4
5
6
7
8
9

# 2.10 jq to awk

echo '{"k1":"v1","k2":"v2"}' | jq '[.k1, .k2] | map(tostring) | join(",")'
# "v1,v2"
$  echo '{"k1":"v1","k2":"v2"}' | jq '[.k1, .k2] | map(tostring) | join(",")' | sed 's/"//g'
# v1,v2
echo '{"k1":"v1","k2":"v2"}' | jq '[.k1, .k2] | map(tostring) | join(" ")' | sed 's/"//g'
# v1 v2

1
2
3
4
5
6

# 2.11 删除某元素

echo '{"k1":"v1","k2":"v2"}' | jq 'del(.k1)'
# {"k2":"v2"}
echo '[{"name": "tom", "id":"1"}, {"name": "mozillazg","id":"2"}]'  | jq -c 'del(.[].name)'
# [{"id":"1"},{"id":"2"}]
echo '[{"name": "tom", "id":"1"}, {"name": "mozillazg","id":"2"}]'  | jq -c 'del(.[0].name)'
# [{"id":"1"},{"name":"mozillazg","id":"2"}]

1
2
3
4
5
6

# 3 数组操作

# 3.1 取出所有元素

echo '[{"name": "tom"}, {"name": "mozillazg"}]' |jq .[]
# {
#   "name": "tom"
# }
# {
#   "name": "mozillazg"
# }

1
2
3
4
5
6
7

# 3.2 取出所有元素的某个key

echo '[{"k1":1,"k2":2,"k3":3},{"k1":11,"k2":12,"k3":13},{"k1":21,"k2":22,"k3":23}]' | jq '.[]' | jq -c '{k1:.k1}'
# {"k1":1}
# {"k1":11}
# {"k1":21}

1
2
3
4

# 3.3 切分(slice)

.[1], .[0:2]

echo '[{"name": "tom"}, {"name": "mozillazg"}, {"name": "jim"}]' |jq  .[1]
# {
#   "name": "mozillazg"
# }
echo '[{"name": "tom"}, {"name": "mozillazg"}, {"name": "jim"}]' |jq  .[1].name
# "mozillazg"
echo '[{"name": "tom"}, {"name": "mozillazg"}, {"name": "jim"}]' |jq .[0:2]
# [
#   {
#     "name": "tom"
#   },
#   {
#     "name": "mozillazg"
#   }
# ]

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

# 3.4 操作 object 组成的数组

echo '[{"name": "foo"},{"name": "bar"},{"name": "foobar"}]' |jq .[].name
# "foo"
# "bar"
# "foobar"

1
2
3
4

echo '[{"name": "foo"},{"name": "bar"},{"name": "foobar"}]' |jq '.[]|.name'
# "foo"
# "bar"
# "foobar"

1
2
3
4

如果要将结果重新组成数组，可以这样:

echo '[{"name": "foo"},{"name": "bar"},{"name": "foobar"}]' |jq [.[].name]
# [
#   "foo",
#   "bar",
#   "foobar"
# ]

1
2
3
4
5
6

也可以使用下面会提到的 map:

echo '[{"name": "foo"},{"name": "bar"},{"name": "foobar"}]' |jq 'map(.name)'
[
  "foo",
  "bar",
  "foobar"
]

1
2
3
4
5
6

# 3.5 获取数组的元素个数

echo  '{"a":[1,2]}' | jq  '.a|length'
# 2

1
2

# 4 字符串操作

# 4.1 获取字符串的元素个数

echo  '{"a":"hello"}' | jq  '.a|length'
# 5

1
2

# 4.2 使用多个 filter

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' |jq '.url, .name'
# "[mozillazg.com](http://mozillazg.com/)"
# "mozillazg"

1
2
3

# 二、高级操作

# 1 管道

可以通过 | 实现类似管道的功能:

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' |jq '.|.url'
# "[mozillazg.com](http://mozillazg.com/)"

1
2

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "tests": [{"foobar": "v1"}, {"foobar": "v2"}]}' |jq '.tests |.[] |.foobar'
# "v1"
# "v2"

1
2
3

也可以直接用 shell 的 | 实现:

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' |jq '.' | jq '.url'
# "[mozillazg.com](http://mozillazg.com/)"

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "tests": [{"foobar": "v1"}, {"foobar": "v2"}]}' |jq '.tests' | jq '.[]' | jq '.foobar'
# "v1"
# "v2"

1
2
3
4
5
6

# 2 获取内容的长度

length 可以获取字符串或数组的长度:

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' |jq '.url|length'
# 13

echo '["[mozillazg.com](http://mozillazg.com/)", "mozillazg"]' |jq '.|length'
# 2

1
2
3
4
5

# 3 map

map(foo) 可以实现对数组的每一项进行操作，然后合并结果的功能:

echo '["[mozillazg.com](http://mozillazg.com/)", "mozillazg"]' | jq 'map(length)'
# [
#   38,
#   9
# ]

echo '["[mozillazg.com](http://mozillazg.com/)", "mozillazg"]' | jq 'map(length)|add'
# 47

1
2
3
4
5
6
7
8

# 4 select

select(foo) 可以实现对输入项进行判断，只返回符合条件的项:

echo '["[mozillazg.com](http://mozillazg.com/)", "mozillazg"]' | jq 'map(select(.|length > 9))'
# [
#   "[mozillazg.com](http://mozillazg.com/)"
# ]

1
2
3
4

# 5 字符串插值，拼接

可以使用 \\(foo) 实现字符串插值功能:

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' |jq '"hi \\(.name)"'
# "hi mozillazg"

1
2

注意要用双引号包围起来，表示是一个字符串，使用 + 实现字符串拼接:

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' |jq '"hi " + .name'
# "hi mozillazg"

1
2

# 6 输出字符串原始值

使用 \-r 选项输出字符串原始值而不是 json 序列化后的值:

echo '{"value":"{\"url\":\"[mozillazg.com](http://mozillazg.com/)\"}"}' |jq .value
# "{\\"url\\": \\"[mozillazg.com](http://mozillazg.com/)\\"}"

echo '{"value":"{\"url\":\"[mozillazg.com](http://mozillazg.com/)\"}"}' |jq -r .value
# {"url": "[mozillazg.com](http://mozillazg.com/)"}

1
2
3
4
5

# 7 if/elif/else

可以使用 if .. then .. elif .. then .. else .. end 实现条件判断:

echo '[0, 1, 2, 3]' | jq 'map(if . == 0 then "zero" elif . == 1 then "one" elif . == 2 then "two" else "many" end)'
# [
#   "zero",
#   "one",
#   "two",
#   "many"
# ]

1
2
3
4
5
6
7

# 8 构造 object 或数组

可以通过 {} 和 [] 构造新的 object 或数组。

echo '["[mozillazg.com](http://mozillazg.com/)", "mozillazg"]' |jq '{name: .[1]}'
# {
#   "name": "mozillazg"
# }

1
2
3
4

echo '{"url": "[mozillazg.com](http://mozillazg.com/)", "name": "mozillazg"}' |jq '[.name, .url]'
# [
#   "mozillazg",
#   "[mozillazg.com](http://mozillazg.com/)"
# ]

1
2
3
4
5

echo '{"name": "mozillazg", "ages": [1, 2]}' | jq '{name, age: .ages[]}'
# {
#   "name": "mozillazg",
#   "age": 1
# }
# {
#   "name": "mozillazg",
#   "age": 2
# }

1
2
3
4
5
6
7
8
9

# 9 数组 join

echo '["[mozillazg.com](http://mozillazg.com/)", "mozillazg"]' | jq '.|join(" | ")'
# "[mozillazg.com](http://mozillazg.com/) | mozillazg"

1
2

# 10 字符串 split

echo '"[mozillazg.com](http://mozillazg.com/) | mozillazg"' |jq 'split(" | ")'
# [
#   "[mozillazg.com](http://mozillazg.com/)",
#   "mozillazg"
# ]

1
2
3
4
5

# 11 to_entries

{
    "services": {
        "10000021": {
            "mobile": {
                "business": "10000021",
                "floor_price": {
                    "price": 1
                }
            }
        },
        "10000190": {
            "mobile": {
                "business": "10000190",
                "floor_price": {
                    "price": 1
                }
            }
        }
    }
}

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

从上述json中提取 price

$ cat file | jq  '.services | to_entries[].value|to_entries[].value' | jq -c '{business:.business, price:.floor_price.price}'
# {"business":"10000021","price":1}
# {"business":"10000190","price":1}

cat file | jq  '.services|to_entries[]' | jq -c '{key:.key, price:.value|to_entries[].value.floor_price.price }'
# {"key":"10000021","price":1}
# {"key":"10000190","price":1}

1
2
3
4
5
6
7

# 12 算数运算

echo '{"k":"k", "v":10}' | jq  '{k:.k, v:(.v*10)}'
# {
#   "k": "k",
#   "v": 100
# }

1
2
3
4
5

# 13 jq 引用 shell 变量

$ export k="v"
echo '{"a":1}' | jq --arg k $k -c '{a:.a,k:$k}'
# {"a":1,"k":"v"}

1
2
3

# 参考资料

jq: https://stedolan.github.io/jq/
jq play: https://jqplay.org/ *【jq 中文手册(v1.5) _ jq 中文手册】 http://alingse.github.io/jq-manual-cn/manual/v1.5/
JSON解析工具-jq： https://hellogitlab.com/OS/Centos/json_tool_jq.html

编辑

上次更新: 2024/10/23, 19:38:27

← iNode使用率过高的问题 MySQL 基本介绍→