格式化 json 数据

有时候，我们在命令行请求 Restful 风格的接口，返回的 json 响应是连在一起，如何格式化 json 数据呢？一般的做法是，先复制 json 数据，然后粘贴到类似 www.bejson.com 在线网站，格式化数据。可以在命令行中，直接格式化 json 数据嘛？当然可以了，使用 python json 包的工具就可以，例如。

➜  ~ curl --silent 'https://www.jianshu.com/users/recommended' | python3 -m json.tool
{
    "users": [
        {
            "id": 3292545,
            "slug": "c4165d16d0ad",
            "nickname": "\u5317\u7f8e\u4e4b\u5317",
            "avatar_source": "http://upload.jianshu.io/users/upload_avatars/3292545/78f2855a-80fd-419a-9fb2-1aeed39690f0.png",
            "total_likes_count": 28933,
            "total_wordage": 407114,
            "is_following_user": false
        },
        ...
        {
            "id": 1835526,
            "slug": "55b597320c4e",
            "nickname": "\u884c\u8ddd\u7248\u541b",
            "avatar_source": "http://upload.jianshu.io/users/upload_avatars/1835526/83d24e1a-0a0f-43f6-8a1d-289be6101e73.png",
            "total_likes_count": 53898,
            "total_wordage": 495624,
            "is_following_user": false
        }
    ],
    "total_count": 21115
}

jq 命令

使用 jq 也可以美化 json 数据， e.g.

➜  ~ curl --silent 'https://www.jianshu.com/users/recommended' | jq '.'
{
  "users": [
    {
      "id": 3292545,
      "slug": "c4165d16d0ad",
      "nickname": "北美之北",
      "avatar_source": "http://upload.jianshu.io/users/upload_avatars/3292545/78f2855a-80fd-419a-9fb2-1aeed39690f0.png",
      "total_likes_count": 28933,
      "total_wordage": 407114,
      "is_following_user": false
    },
    ...
    {
      "id": 1835526,
      "slug": "55b597320c4e",
      "nickname": "行距版君",
      "avatar_source": "http://upload.jianshu.io/users/upload_avatars/1835526/83d24e1a-0a0f-43f6-8a1d-289be6101e73.png",
      "total_likes_count": 53898,
      "total_wordage": 495624,
      "is_following_user": false
    }
  ],
  "total_count": 21146
}

如何在命令行中，提取 json 数据的字段值呢？使用 jq 命令，可以帮助你。

例如请求简书的推荐用户接口 GET https://www.jianshu.com/users/recommended，过滤出字段 total_count。

e.g.

➜  ~ curl --silent 'https://www.jianshu.com/users/recommended' | jq '.total_count'
21129

jq 命令不是系统自带的，使用前，需要安装该命令。从jq 官网下载该命令，然后放到系统 PATH 中。

# Linux
sudo wget https://github.com/stedolan/jq/releases/download/jq-1.6/jq-linux64 -O /usr/local/bin/jq
sudo chmod +x /usr/local/bin/jq

# Mac OS X
sudo wget https://github.com/stedolan/jq/releases/download/jq-1.6/jq-osx-amd64 -O /usr/local/bin/jq
sudo chmod +x /usr/local/bin/jq

# Windows 下
# 从下载 https://github.com/stedolan/jq/releases/download/jq-1.6/jq-win64.exe
# 然后放到系统 PATH 下.

还有一个网站 jqplay 用于在线使用 jq 命令，可以登上去，试下。

基础使用和使用例子如下，点击官网文档查看更多使用。

1. 过滤 key 值，    `.foo, .foo.bar, .foo?`
➜  ~ echo '{"foo": 42, "bar": "less interesting data"}' | jq '.foo'
42

2. 数组操作 `.[], .[]?, .[2], .[10:15]`
➜  ~ echo '[{"name":"JSON", "good":true}, {"name":"XML", "good":false}]' | jq '.[1]'
{
  "name": "XML",
  "good": false
}

3. 构造数组/对象 `[], {}`
➜  ~ echo '{"user":"stedolan","titles":["JQ Primer", "More JQ"]}' | jq '{user, title: .titles[]}'
{
  "user": "stedolan",
  "title": "JQ Primer"
}
{
  "user": "stedolan",
  "title": "More JQ"
}

4.  获取 json 值的长度 `length`
➜  ~ echo '[[1,2], "string", {"a":2}, null]' | jq '.[] | length'
2
6
1
0

5. 过滤 json 对象的 keys `keys`
➜  ~ echo '{"abc": 1, "abcd": 2, "Foo": 3}' | jq 'keys'
[
  "Foo",
  "abc",
  "abcd"
]

6. 使用 `,` 过滤多个值
➜  ~ echo '{ "foo": 42, "bar": "something else", "baz": true}' | jq '.foo, .bar'
42
"something else"

7. 使用管道符号 `|` 多次过滤 
➜  ~ echo '[{"name":"JSON", "good":true}, {"name":"XML", "good":false}]' | jq '.[] | .name'
"JSON"
"XML"

8. 如果正确，则选择值 `select(foo)`
➜  ~ echo '[1,5,3,0,7]' | jq 'map(select(. >= 2))'
[
  5,
  3,
  7
]

9. 对每个输入进行操作 `map(foo)`
➜  ~ echo '[1,2,3]' | jq 'map(.+1)'
[
  2,
  3,
  4
]

10. if-then-else-end 条件
➜  ~ echo '2' | jq 'if . == 0 then "zero" elif . == 1 then "one" else "many" end'
"many"

11. 字符插入 `\(foo)`
➜  ~ echo '42' | jq '"The input was \(.), which is one less than \(.+1)"'
"The input was 42, which is one less than 43"