Skip to content

Latest commit

 

History

History
683 lines (527 loc) · 14 KB

README.md

File metadata and controls

683 lines (527 loc) · 14 KB

Mazur 的 SQL 风格指南

您好!我是Matt Mazur ,是一名数据分析师,曾在几家初创公司工作过,帮助公司利用数据发展业务。本指南记录了我对格式化 SQL 的喜好,希望对其他人有一些用处。如果您或您的团队还没有 SQL 风格指南,那么它可以作为一个很好的起点,您可以根据自己的喜好来采用和更新它。

另外,我是一个坚定的信仰者Strong Opinions, Weakly Held。如果你不同意它, 给我私信, 我会喜欢并一起讨论的.

如果你喜欢这个课题,你应该也喜欢我的LookML Style Guide或者我的博客,我写了很多关于分析和数据分析的文章.

例子

这里是一些比较简单的查询,主要是为了展示下这个指南:

with hubspot_interest as (

    select
        email,
        timestamp_millis(property_beacon_interest) as expressed_interest_at
    from hubspot.contact
    where property_beacon_interest is not null

), 

support_interest as (

    select 
        conversation.email,
        conversation.created_at as expressed_interest_at
    from helpscout.conversation
    inner join helpscout.conversation_tag on conversation.id = conversation_tag.conversation_id
    where conversation_tag.tag = 'beacon-interest'

), 

combined_interest as (

    select * from hubspot_interest
    union all
    select * from support_interest

),

final as (

    select 
        email,
        min(expressed_interest_at) as expressed_interest_at
    from combined_interest
    group by email

)

select * from final

指南

使用小写字母 SQL

它就像大写 SQL 一样易读,而且你不必总是按住 Shift 键。

--
select * from users

-- 不好
SELECT * FROM users

-- 不好
Select * From users

单行查询 vs 多行查询

只需要查询单列,则推荐单行:

--
select * from users

--
select id from users

--
select count(*) from users

一旦你需要查询更多的列或比较复杂,分散在多行可以变得更容易阅读:

--
select
    id,
    email,
    created_at
from users

--
select *
from users
where email = '[email protected]'

--
select
    user_id,
    count(*) as total_charges
from charges
group by user_id

-- 不好
select id, email, created_at
from users

-- 不好
select id,
    email
from users

左对齐 SQL 关键字

有些 IDE 能够自动格式化 SQL,以便 SQL 关键字之后的空格垂直对齐。手动做这个格式化非常麻烦(在我看来这样也更难阅读),所以我建议所有的关键字都左对齐。

--
select id, email
from users
where email like '%@gmail.com'

-- 不好
select id, email
  from users
 where email like '%@gmail.com'

使用单引号

有些 SQL 分支(例如 BigQuery)支持使用双引号,但是对于大多数分支,双引号都使用在列名上,因此最好使用单引号。

--
select *
from users
where email = '[email protected]'

-- 不好
select *
from users
where email = "[email protected]"

使用 != 而不是 <>

很简单,因为 != 看起来像 “不等于”,更接近我们想要表达的意思。

--
select count(*) as paying_users_count
from users
where plan_name != 'free'

逗号应该在行尾

--
select
    id,
    email
from users

-- 不好
select
    id
    , email
from users

where 条件的缩进

当只有一个条件时,与 where 保持在同一行:

select email
from users
where id = 1234

当有多个条件时,每一个条件都比 where 缩进一层。将逻辑运算符放在前一个条件的末尾:

select id, email
from users
where 
    created_at >= '2019-03-01' and 
    vertical = 'work'

避免括号内的空格

--
select *
from users
where id in (1, 2)

-- 不好
select *
from users
where id in ( 1, 2 )

in 中比较长的列表,应该分在多个不同的缩进行

--
select *
from users
where email in (
    '[email protected]',
    '[email protected]',
    '[email protected]',
    '[email protected]'
)

表名应该是名称复数蛇形风格

--
select * from users
select * from visit_logs

-- 不好
select * from user
select * from visitLog

列名应该是蛇形风格

--
select
    id,
    email,
    timestamp_trunc(created_at, month) as signup_month
from users

-- 不好
select
    id,
    email,
    timestamp_trunc(created_at, month) as SignupMonth
from users

列名约定

  • Boolean 类型应该有 is_has_does_ 前缀。例如 is_customerhas_unsubscribed 等。
  • Date-only 类型应该有 _date 后缀。例如 report_date 等。
  • Date+time 类型应该有 _at 后缀。例如 created_atposted_at 等。

列命约定

将主键放到最前面,然后是外键,最后是其他列。如果有任何系统列(如 created_atupdated_atis_deleted 等等,把它们放到最后。

--
select
    id,
    name,
    created_at
from users

-- 不好
select
    created_at,
    name,
    id,
from users

显性 inner,在内部连接

最好是显性写出 inner join,而不是省略 inner

--
select
    users.email,
    sum(charges.amount) as total_revenue
from users
inner join charges on users.id = charges.user_id

-- 不好
select
    users.email,
    sum(charges.amount) as total_revenue
from users
join charges on users.id = charges.user_id

对于 join 条件,按引用的表顺序排序到 on 之后

通过这样做,可以更容易确定连接是否导致结果呈扇形分布:

--
select
    ...
from users
left join charges on users.id = charges.user_id
-- primary_key = foreign_key --> one-to-many --> fanout
  
select
    ...
from charges
left join users on charges.user_id = users.id
-- foreign_key = primary_key --> many-to-one --> no fanout

-- 不好
select
    ...
from users
left join charges on charges.user_id = users.id

单个连接条件应与 join 在同一行上

--
select
    users.email,
    sum(charges.amount) as total_revenue
from users
inner join charges on users.id = charges.user_id
group by email

-- 不好
select
    users.email,
    sum(charges.amount) as total_revenue
from users
inner join charges
on users.id = charges.user_id
group by email

当有多个连接条件时,请将每个条件放在它们自己的缩进行中:

--
select
    users.email,
    sum(charges.amount) as total_revenue
from users
inner join charges on 
    users.id = charges.user_id and
    refunded = false
group by email

大多数情况下尽量避免表名的别名

将表名 users 缩写为 u,将 charges 缩写为 c,这可能很诱人,但这最终会降低 SQL 的可读性

--
select
    users.email,
    sum(charges.amount) as total_revenue
from users
inner join charges on users.id = charges.user_id

-- 不好
select
    u.email,
    sum(c.amount) as total_revenue
from users u
inner join charges c on u.id = c.user_id

大多数情况下,最好是使用完整的表名。

有两个例外:

  • 如果需要在同一个查询中多次连接到一个表,并且需要区分这几个之间的不同,那么就需要别名。
  • 另外,如果表名很长或有歧义,可以使用别名(但仍然需要使用有意义的名称)。
-- 好:有意义的表别名
select
  companies.com_name,
  beacons.created_at
from stg_mysql_helpscout__helpscout_companies companies
inner join stg_mysql_helpscout__helpscout_beacons_v2 beacons on companies.com_id = beacons.com_id

-- 还行:没有表别名
select
  stg_mysql_helpscout__helpscout_companies.com_name,
  stg_mysql_helpscout__helpscout_beacons_v2.created_at
from stg_mysql_helpscout__helpscout_companies
inner join stg_mysql_helpscout__helpscout_beacons_v2 on stg_mysql_helpscout__helpscout_companies.com_id = stg_mysql_helpscout__helpscout_beacons_v2.com_id

-- 不好:不清晰的表别名
select
  c.com_name,
  b.created_at
from stg_mysql_helpscout__helpscout_companies c
inner join stg_mysql_helpscout__helpscout_beacons_v2 b on c.com_id = b.com_id

当存在 join 时,显性写出表名,否则省略表名

当没有涉及到 join 时,就不会对列来自哪个表产生歧义,因此可以省略表名:

--
select
    id,
    name
from companies

-- 不好
select
    companies.id,
    companies.name
from companies

当涉及到 join 时,最好是显式的,这样就可以清楚地知道列来源:

--
select
    users.email,
    sum(charges.amount) as total_revenue
from users
inner join charges on users.id = charges.user_id

-- 不好
select
    email,
    sum(amount) as total_revenue
from users
inner join charges on users.id = charges.user_id

总是重命名聚合和函数包装的参数

--
select count(*) as total_users
from users

-- 不好
select count(*)
from users

--
select timestamp_millis(property_beacon_interest) as expressed_interest_at
from hubspot.contact
where property_beacon_interest is not null

-- 不好
select timestamp_millis(property_beacon_interest)
from hubspot.contact
where property_beacon_interest is not null

明确布尔条件

--
select * from customers where is_cancelled = true
select * from customers where is_cancelled = false

-- 不好
select * from customers where is_cancelled
select * from customers where not is_cancelled

使用 as 作为列名别名

--
select
    id,
    email,
    timestamp_trunc(created_at, month) as signup_month
from users

-- 不好
select
    id,
    email,
    timestamp_trunc(created_at, month) signup_month
from users

使用列名或列号进行分组,但不要同时使用两种

我更喜欢按列名分组,但按数字分组也是极好的

--
select user_id, count(*) as total_charges
from charges
group by user_id

--
select user_id, count(*) as total_charges
from charges
group by 1

-- 不好
select
    timestamp_trunc(created_at, month) as signup_month,
    vertical,
    count(*) as users_count
from users
group by 1, vertical

按名称分组时,使用别名

--
select
  timestamp_trunc(com_created_at, year) as signup_year,
  count(*) as total_companies
from companies
group by signup_year

-- 不好
select
  timestamp_trunc(com_created_at, year) as signup_year,
  count(*) as total_companies
from companies
group by timestamp_trunc(com_created_at, year)

首先应该对列分组

--
select
  timestamp_trunc(com_created_at, year) as signup_year,
  count(*) as total_companies
from companies
group by signup_year

-- 不好
select
  count(*) as total_companies,
  timestamp_trunc(com_created_at, year) as signup_year
from mysql_helpscout.helpscout_companies
group by signup_year

调整 case/when 语句

每个 when 都应该独自一行( case 不包含别的信息),并且应该缩进比 case 深一层,then 可以和 when 保持在同一行,也可以换行。

--
select
    case
        when event_name = 'viewed_homepage' then 'Homepage'
        when event_name = 'viewed_editor' then 'Editor'
        else 'Other'
    end as page_name
from events

-- 也不错
select
    case
        when event_name = 'viewed_homepage'
            then 'Homepage'
        when event_name = 'viewed_editor'
            then 'Editor'
        else 'Other'            
    end as page_name
from events

-- 不好 
select
    case when event_name = 'viewed_homepage' then 'Homepage'
        when event_name = 'viewed_editor' then 'Editor'
        else 'Other'        
    end as page_name
from events

使用 CTE (公用表表达式),而不是子查询

避免使用子查询,CTE 将使查询更容易阅读和理解。

使用 CTE 时,用新行填充查询。

在使用任意的 CTE 时,始终使用 finalselect * from final 。通过这种方式,可以快速检查查询中使用的其他 CTE 输出,以便调试结果。

结尾的 CTE 括号应该使用与 with 和 CTE 名称相同的缩进。

--
with ordered_details as (

    select
        user_id,
        name,
        row_number() over (partition by user_id order by date_updated desc) as details_rank
    from billingdaddy.billing_stored_details

),

final as (

    select user_id, name
    from ordered_details
    where details_rank = 1

)

select * from final

-- 不好
select user_id, name
from (
    select
        user_id,
        name,
        row_number() over (partition by user_id order by date_updated desc) as details_rank
    from billingdaddy.billing_stored_details
) ranked
where details_rank = 1

使用有意义的 CTE 名称

--
with ordered_details as (

-- 不好
with d1 as (

窗口函数

你可以把它单独放在一行上,或者根据它的长度把它分成多行:

--
select
    user_id,
    name,
    row_number() over (partition by user_id order by date_updated desc) as details_rank
from billingdaddy.billing_stored_details

--
select
    user_id,
    name,
    row_number() over (
        partition by user_id
        order by date_updated desc
    ) as details_rank
from billingdaddy.billing_stored_details

Credits

这个风格指南的灵感部分来自于:

向 Peter Butler、Dan Wyman、Simon Ouderkirk、Alex Cano、Adam Stone、Brian Kim 和 Claire Carroll 致敬,感谢他们对本指南提供的反馈。