flink sql 所有函数详细用例

wudl55662025-10-20 11:27

参考文档 : Apache Flink System Functions

📖 文档说明

本文档提供Flink SQL所有内置函数的详细清单，包括：

  • ✅ 完整的函数列表和中文说明
  • ✅ 详细的SQL测试示例
  • ✅ 实际的执行结果
  • ✅ 生产环境应用场景

文档共分为5个部分：

  • ** 1**: 比较函数、逻辑函数、算术函数
  • ** 2**: 字符串函数
  • ** 3**: 时间函数、条件函数
  • ** 4**: 类型转换、集合函数、JSON函数
  • ** 5**: 聚合函数、窗口函数、表值函数

目录

  • [1. 比较函数 (Comparison Functions)](#1. 比较函数 (Comparison Functions))
  • [2. 逻辑函数 (Logical Functions)](#2. 逻辑函数 (Logical Functions))
  • [3. 算术函数 (Arithmetic Functions)](#3. 算术函数 (Arithmetic Functions))

1. 比较函数 (Comparison Functions)

1.1 函数清单

函数名称 中文说明 语法示例 返回类型
= 等于比较 value1 = value2 BOOLEAN
<>, != 不等于比较 value1 <> value2 BOOLEAN
> 大于比较 value1 > value2 BOOLEAN
>= 大于等于 value1 >= value2 BOOLEAN
< 小于比较 value1 < value2 BOOLEAN
<= 小于等于 value1 <= value2 BOOLEAN
IS NULL 判断是否为NULL value IS NULL BOOLEAN
IS NOT NULL 判断是否不为NULL value IS NOT NULL BOOLEAN
IS DISTINCT FROM NULL安全的不等于 value1 IS DISTINCT FROM value2 BOOLEAN
IS NOT DISTINCT FROM NULL安全的等于 value1 IS NOT DISTINCT FROM value2 BOOLEAN
BETWEEN 范围判断 value BETWEEN min AND max BOOLEAN
NOT BETWEEN 范围外判断 value NOT BETWEEN min AND max BOOLEAN
LIKE 模式匹配 string LIKE pattern [ESCAPE char] BOOLEAN
NOT LIKE 模式不匹配 string NOT LIKE pattern BOOLEAN
SIMILAR TO SQL正则匹配 string SIMILAR TO pattern BOOLEAN
IN 值在列表中 value IN (v1, v2, ...) BOOLEAN
NOT IN 值不在列表中 value NOT IN (v1, v2, ...) BOOLEAN

1.2 详细测试示例

测试 1.1: 基础比较操作

测试目的: 验证等于、不等于、大于、小于等基础比较操作

测试SQL:

sql 复制代码 
SELECT 
    -- 等于比较
    5 = 5 as eq_true,
    5 = 3 as eq_false,
    NULL = NULL as eq_null,
    
    -- 不等于比较
    5 <> 3 as neq_true,
    5 != 5 as neq_false,
    
    -- 大小比较
    10 > 5 as gt_true,
    10 >= 10 as gte_true,
    3 < 5 as lt_true,
    5 <= 5 as lte_true,
    
    -- 字符串比较（字典序）
    'apple' < 'banana' as str_lt,
    'Flink' = 'Flink' as str_eq
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+----------+-----------+----------+----------+-----------+----------+-----------+----------+-----------+---------+---------+
| eq_true  | eq_false  | eq_null  | neq_true | neq_false | gt_true  | gte_true  | lt_true  | lte_true  | str_lt  | str_eq  |
+----------+-----------+----------+----------+-----------+----------+-----------+----------+-----------+---------+---------+
| TRUE     | FALSE     | NULL     | TRUE     | FALSE     | TRUE     | TRUE      | TRUE     | TRUE      | TRUE    | TRUE    |
+----------+-----------+----------+----------+-----------+----------+-----------+----------+-----------+---------+---------+

结果说明:

  • ✅ 所有基础比较操作正常
  • ⚠️ NULL = NULL 返回 NULL (UNKNOWN)，这是SQL的三值逻辑
  • ✅ 字符串按字典序比较，区分大小写

实际应用场景:

sql 复制代码 
-- 场景1: 查询指定价格范围的商品
SELECT product_id, name, price
FROM products
WHERE price >= 100 AND price <= 500
ORDER BY price DESC

-- 场景2: 查询特定状态的订单
SELECT order_id, status, create_time
FROM orders
WHERE status = 'paid' OR status = 'shipped'

-- 场景3: 查询评分高于平均分的学生
SELECT student_id, name, score
FROM students
WHERE score > (SELECT AVG(score) FROM students)
测试 1.2: IS DISTINCT FROM (NULL安全比较)

测试目的: 验证NULL安全的相等和不等比较

测试SQL:

sql 复制代码 
SELECT 
    -- 普通等于（NULL不安全）
    1 = NULL as normal_eq,
    NULL = NULL as null_eq,
    
    -- IS DISTINCT FROM（NULL安全的不等于）
    1 IS DISTINCT FROM NULL as distinct_1_null,
    NULL IS DISTINCT FROM NULL as distinct_null_null,
    5 IS DISTINCT FROM 5 as distinct_5_5,
    5 IS DISTINCT FROM 3 as distinct_5_3,
    
    -- IS NOT DISTINCT FROM（NULL安全的等于）
    1 IS NOT DISTINCT FROM NULL as not_distinct_1_null,
    NULL IS NOT DISTINCT FROM NULL as not_distinct_nn,
    5 IS NOT DISTINCT FROM 5 as not_distinct_5_5
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+------------+----------+------------------+---------------------+---------------+---------------+---------------------+------------------+-------------------+
| normal_eq  | null_eq  | distinct_1_null  | distinct_null_null  | distinct_5_5  | distinct_5_3  | not_distinct_1_null | not_distinct_nn  | not_distinct_5_5  |
+------------+----------+------------------+---------------------+---------------+---------------+---------------------+------------------+-------------------+
| NULL       | NULL     | TRUE             | FALSE               | FALSE         | TRUE          | FALSE               | TRUE             | TRUE              |
+------------+----------+------------------+---------------------+---------------+---------------+---------------------+------------------+-------------------+

结果说明:

  • ✅ IS DISTINCT FROM 将NULL视为普通值，两个NULL相等
  • ✅ 用于JOIN条件时，能正确匹配NULL值
  • ⚠️ 标准=操作符遇到NULL返回NULL，可能导致WHERE子句过滤错误

实际应用场景:

sql 复制代码 
-- 场景1: 查找两表中不同的记录（包括NULL值比较）
SELECT a.id, a.name, a.email, b.email
FROM table_a a
LEFT JOIN table_b b ON a.id = b.id
WHERE a.email IS DISTINCT FROM b.email

-- 场景2: 数据变更检测（包括NULL变化）
SELECT 
    old.id,
    old.status as old_status,
    new.status as new_status
FROM old_data old
JOIN new_data new ON old.id = new.id
WHERE old.status IS DISTINCT FROM new.status

-- 场景3: 去重（NULL也视为重复）
SELECT DISTINCT ON (
    name IS NOT DISTINCT FROM NULL
) * 
FROM users
测试 1.3: BETWEEN范围查询

测试目的: 验证BETWEEN和NOT BETWEEN的各种用法

测试SQL:

sql 复制代码 
SELECT 
    -- ASYMMETRIC（默认）
    12 BETWEEN 10 AND 15 as between_normal,
    12 BETWEEN 15 AND 10 as between_reversed,
    
    -- SYMMETRIC（顺序无关）
    12 BETWEEN SYMMETRIC 10 AND 15 as sym_normal,
    12 BETWEEN SYMMETRIC 15 AND 10 as sym_reversed,
    
    -- NULL处理
    12 BETWEEN 10 AND NULL as between_null_upper,
    12 BETWEEN NULL AND 15 as between_null_lower,
    12 BETWEEN SYMMETRIC NULL AND 12 as sym_null,
    
    -- NOT BETWEEN
    20 NOT BETWEEN 10 AND 15 as not_between_out,
    12 NOT BETWEEN 10 AND 15 as not_between_in,
    
    -- 日期范围
    DATE '2024-05-15' BETWEEN DATE '2024-01-01' AND DATE '2024-12-31' as date_between,
    
    -- 字符串范围
    'M' BETWEEN 'A' AND 'Z' as char_between
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+-----------------+-------------------+-------------+---------------+---------------------+--------------------+-----------+-------------------+-----------------+---------------+---------------+
| between_normal  | between_reversed  | sym_normal  | sym_reversed  | between_null_upper  | between_null_lower | sym_null  | not_between_out   | not_between_in  | date_between  | char_between  |
+-----------------+-------------------+-------------+---------------+---------------------+--------------------+-----------+-------------------+-----------------+---------------+---------------+
| TRUE            | FALSE             | TRUE        | TRUE          | NULL                | FALSE              | NULL      | TRUE              | FALSE           | TRUE          | TRUE          |
+-----------------+-------------------+-------------+---------------+---------------------+--------------------+-----------+-------------------+-----------------+---------------+---------------+

结果说明:

  • ✅ BETWEEN 等价于 value >= min AND value <= max
  • ✅ SYMMETRIC 自动处理min/max顺序
  • ⚠️ NULL边界导致UNKNOWN或FALSE结果
  • ✅ 支持数值、日期、字符串类型

实际应用场景:

sql 复制代码 
-- 场景1: 时间范围查询
SELECT order_id, order_time, amount
FROM orders
WHERE order_time BETWEEN TIMESTAMP '2024-01-01 00:00:00' 
                    AND TIMESTAMP '2024-01-31 23:59:59'

-- 场景2: 价格区间筛选
SELECT product_id, name, price
FROM products
WHERE price BETWEEN 100.0 AND 500.0
ORDER BY price

-- 场景3: 年龄段统计
SELECT 
    COUNT(*) as count,
    CASE 
        WHEN age BETWEEN 0 AND 17 THEN 'Minor'
        WHEN age BETWEEN 18 AND 35 THEN 'Young'
        WHEN age BETWEEN 36 AND 60 THEN 'Middle'
        ELSE 'Senior'
    END as age_group
FROM users
GROUP BY age_group

-- 场景4: IP地址范围（转换为数值）
SELECT ip_address
FROM access_logs
WHERE ip_to_int(ip_address) BETWEEN ip_to_int('192.168.1.0') 
                                AND ip_to_int('192.168.1.255')
测试 1.4: LIKE模式匹配

测试目的: 验证LIKE和NOT LIKE的模式匹配功能

测试SQL:

sql 复制代码 
SELECT 
    -- 基础LIKE
    'Hello World' LIKE 'Hello%' as like_prefix,
    'Hello World' LIKE '%World' as like_suffix,
    'Hello World' LIKE '%o W%' as like_contains,
    'Hello World' LIKE 'Hello_World' as like_single_char,
    'Hello World' LIKE 'Hello%World' as like_both,
    
    -- NOT LIKE
    'Hello World' NOT LIKE 'Goodbye%' as not_like,
    
    -- ESCAPE字符
    '100% discount' LIKE '100!%%' ESCAPE '!' as like_escape,
    'C:\path\file' LIKE '%!_path!_%' ESCAPE '!' as like_escape_underscore,
    
    -- 实际案例
    'user@example.com' LIKE '%@%.com' as email_match,
    'ORDER-2024-001' LIKE 'ORDER-____-___' as order_pattern,
    '+86-138-1234-5678' LIKE '+__-___-____-____' as phone_pattern,
    
    -- 大小写敏感
    'Hello' LIKE 'hello' as case_sensitive,
    'Hello' LIKE 'Hello' as case_match
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+--------------+--------------+----------------+------------------+------------+-----------+--------------+-------------------------+--------------+----------------+---------------+-----------------+------------+
| like_prefix  | like_suffix  | like_contains  | like_single_char | like_both  | not_like  | like_escape  | like_escape_underscore  | email_match  | order_pattern  | phone_pattern | case_sensitive  | case_match |
+--------------+--------------+----------------+------------------+------------+-----------+--------------+-------------------------+--------------+----------------+---------------+-----------------+------------+
| TRUE         | TRUE         | TRUE           | TRUE             | TRUE       | TRUE      | TRUE         | TRUE                    | TRUE         | TRUE           | TRUE          | FALSE           | TRUE       |
+--------------+--------------+----------------+------------------+------------+-----------+--------------+-------------------------+--------------+----------------+---------------+-----------------+------------+

结果说明:

  • ✅ % 匹配任意数量字符（包括0个）
  • ✅ _ 匹配恰好1个字符
  • ✅ ESCAPE 用于转义特殊字符
  • ⚠️ 默认大小写敏感

实际应用场景:

sql 复制代码 
-- 场景1: 模糊搜索用户名
SELECT user_id, username, email
FROM users
WHERE username LIKE '%zhang%' OR email LIKE '%zhang%'

-- 场景2: 搜索特定域名的邮箱
SELECT user_id, email
FROM users
WHERE email LIKE '%@gmail.com' OR email LIKE '%@163.com'

-- 场景3: 搜索特定前缀的订单
SELECT order_id, create_time, amount
FROM orders
WHERE order_id LIKE 'VIP-%'

-- 场景4: 文件名匹配
SELECT file_id, filename
FROM files
WHERE filename LIKE '%.pdf' 
   OR filename LIKE '%.docx'
   OR filename LIKE '%.xlsx'

-- 场景5: 电话号码格式验证
SELECT phone
FROM contacts
WHERE phone NOT LIKE '+__-___-____-____'  -- 格式不正确的号码
测试 1.5: IN操作符

测试目的: 验证IN和NOT IN的各种情况，特别是NULL处理

测试SQL:

sql 复制代码 
SELECT 
    -- 基础IN
    4 IN (1, 2, 3, 4, 5) as in_exists,
    6 IN (1, 2, 3, 4, 5) as in_not_exists,
    'apple' IN ('apple', 'banana', 'orange') as in_string,
    
    -- NULL处理（重要！）
    1 IN (1, 2, NULL) as in_null_match,
    4 IN (1, 2, NULL) as in_null_no_match,
    NULL IN (1, 2, 3) as null_in_list,
    NULL IN (1, 2, NULL) as null_in_list_with_null,
    
    -- NOT IN
    6 NOT IN (1, 2, 3, 4, 5) as not_in_true,
    1 NOT IN (1, 2, NULL) as not_in_null_match,
    4 NOT IN (1, 2, NULL) as not_in_null_no_match,
    
    -- 子查询（示例）
    10 IN (SELECT unnest(ARRAY[5, 10, 15])) as in_subquery
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+------------+----------------+------------+----------------+---------------------+--------------+--------------------------+-------------+---------------------+-----------------------+--------------+
| in_exists  | in_not_exists  | in_string  | in_null_match  | in_null_no_match    | null_in_list | null_in_list_with_null   | not_in_true | not_in_null_match   | not_in_null_no_match  | in_subquery  |
+------------+----------------+------------+----------------+---------------------+--------------+--------------------------+-------------+---------------------+-----------------------+--------------+
| TRUE       | FALSE          | TRUE       | TRUE           | NULL                | NULL         | NULL                     | TRUE        | FALSE               | NULL                  | TRUE         |
+------------+----------------+------------+----------------+---------------------+--------------+--------------------------+-------------+---------------------+-----------------------+--------------+

结果说明:

  • ✅ IN 检查值是否在列表中
  • ⚠️ 列表含NULL且未找到匹配：返回NULL（UNKNOWN）
  • ⚠️ 列表含NULL且找到匹配：返回TRUE
  • ⚠️ NOT IN 遇到NULL时结果可能不符合直觉
  • ✅ 支持子查询作为值列表

实际应用场景:

sql 复制代码 
-- 场景1: 多状态查询
SELECT order_id, status, amount
FROM orders
WHERE status IN ('paid', 'shipped', 'delivered')

-- 场景2: 黑名单过滤
SELECT user_id, username
FROM users
WHERE user_id NOT IN (
    SELECT user_id FROM blacklist
)

-- 场景3: 多值精确匹配
SELECT product_id, category, price
FROM products
WHERE category IN ('Electronics', 'Books', 'Clothing')
  AND brand IN ('Apple', 'Samsung', 'Huawei')

-- 场景4: VIP用户查询
SELECT *
FROM orders
WHERE user_id IN (SELECT user_id FROM vip_users)

-- 场景5: 排除测试数据
SELECT * 
FROM transactions
WHERE account_id NOT IN (1, 2, 3, 999, 1000)  -- 测试账号

2. 逻辑函数 (Logical Functions)

2.1 函数清单

函数名称 中文说明 语法示例 返回类型
AND 逻辑与 boolean1 AND boolean2 BOOLEAN
OR 逻辑或 boolean1 OR boolean2 BOOLEAN
NOT 逻辑非 NOT boolean BOOLEAN
IS FALSE 严格FALSE检查 boolean IS FALSE BOOLEAN
IS NOT FALSE 非FALSE检查 boolean IS NOT FALSE BOOLEAN
IS TRUE 严格TRUE检查 boolean IS TRUE BOOLEAN
IS NOT TRUE 非TRUE检查 boolean IS NOT TRUE BOOLEAN
IS UNKNOWN UNKNOWN检查 boolean IS UNKNOWN BOOLEAN
IS NOT UNKNOWN 非UNKNOWN检查 boolean IS NOT UNKNOWN BOOLEAN

2.2 详细测试示例

测试 2.1: AND/OR/NOT三值逻辑

测试目的: 验证SQL的三值逻辑（TRUE, FALSE, UNKNOWN）

测试SQL:

sql 复制代码 
SELECT 
    -- AND真值表
    TRUE AND TRUE as and_tt,
    TRUE AND FALSE as and_tf,
    TRUE AND NULL as and_tn,
    FALSE AND TRUE as and_ft,
    FALSE AND FALSE as and_ff,
    FALSE AND NULL as and_fn,
    NULL AND TRUE as and_nt,
    NULL AND FALSE as and_nf,
    NULL AND NULL as and_nn,
    
    -- OR真值表
    TRUE OR TRUE as or_tt,
    TRUE OR FALSE as or_tf,
    TRUE OR NULL as or_tn,
    FALSE OR FALSE as or_ff,
    FALSE OR NULL as or_fn,
    NULL OR NULL as or_nn,
    
    -- NOT真值表
    NOT TRUE as not_t,
    NOT FALSE as not_f,
    NOT NULL as not_n
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+--------+--------+--------+--------+--------+--------+--------+--------+--------+-------+-------+-------+-------+-------+-------+--------+--------+--------+
| and_tt | and_tf | and_tn | and_ft | and_ff | and_fn | and_nt | and_nf | and_nn | or_tt | or_tf | or_tn | or_ff | or_fn | or_nn | not_t  | not_f  | not_n  |
+--------+--------+--------+--------+--------+--------+--------+--------+--------+-------+-------+-------+-------+-------+-------+--------+--------+--------+
| TRUE   | FALSE  | NULL   | FALSE  | FALSE  | FALSE  | NULL   | FALSE  | NULL   | TRUE  | TRUE  | TRUE  | FALSE | NULL  | NULL  | FALSE  | TRUE   | NULL   |
+--------+--------+--------+--------+--------+--------+--------+--------+--------+-------+-------+-------+-------+-------+-------+--------+--------+--------+

三值逻辑真值表:

AND操作:

A B A AND B
TRUE TRUE TRUE
TRUE FALSE FALSE
TRUE NULL NULL
FALSE TRUE FALSE
FALSE FALSE FALSE
FALSE NULL FALSE
NULL TRUE NULL
NULL FALSE FALSE
NULL NULL NULL

OR操作:

A B A OR B
TRUE TRUE TRUE
TRUE FALSE TRUE
TRUE NULL TRUE
FALSE FALSE FALSE
FALSE NULL NULL
NULL NULL NULL

NOT操作:

A NOT A
TRUE FALSE
FALSE TRUE
NULL NULL

结果说明:

  • ✅ FALSE AND anything = FALSE（短路）
  • ✅ TRUE OR anything = TRUE（短路）
  • ⚠️ NULL参与运算通常导致NULL结果
  • ✅ 遵循SQL三值逻辑标准

实际应用场景:

sql 复制代码 
-- 场景1: 复合条件过滤
SELECT * FROM orders
WHERE (status = 'paid' OR status = 'shipped')
  AND amount > 100
  AND (discount IS NULL OR discount < 0.3)

-- 场景2: 多条件搜索
SELECT * FROM products
WHERE (category = 'Electronics' AND brand = 'Apple')
   OR (category = 'Books' AND price < 50)

-- 场景3: 排除特定组合
SELECT * FROM users
WHERE NOT (is_blocked AND is_deleted)

-- 场景4: 可选过滤条件
SELECT * FROM logs
WHERE level = 'ERROR'
  AND (user_id IS NULL OR user_id = 12345)  -- user_id可选
测试 2.2: IS TRUE/FALSE/UNKNOWN

测试目的: 验证严格的布尔值检查

测试SQL:

sql 复制代码 
SELECT 
    -- IS TRUE（仅TRUE返回TRUE）
    TRUE IS TRUE as is_true_t,
    FALSE IS TRUE as is_true_f,
    NULL IS TRUE as is_true_n,
    (5 > 3) IS TRUE as is_true_expr,
    
    -- IS NOT TRUE（FALSE或NULL返回TRUE）
    TRUE IS NOT TRUE as not_true_t,
    FALSE IS NOT TRUE as not_true_f,
    NULL IS NOT TRUE as not_true_n,
    
    -- IS FALSE
    FALSE IS FALSE as is_false_f,
    TRUE IS FALSE as is_false_t,
    NULL IS FALSE as is_false_n,
    
    -- IS NOT FALSE
    FALSE IS NOT FALSE as not_false_f,
    TRUE IS NOT FALSE as not_false_t,
    NULL IS NOT FALSE as not_false_n,
    
    -- IS UNKNOWN
    NULL IS UNKNOWN as is_unknown_n,
    TRUE IS UNKNOWN as is_unknown_t,
    (1 = NULL) IS UNKNOWN as is_unknown_expr,
    
    -- IS NOT UNKNOWN
    NULL IS NOT UNKNOWN as not_unknown_n,
    TRUE IS NOT UNKNOWN as not_unknown_t
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+------------+------------+------------+--------------+--------------+--------------+--------------+-------------+-------------+-------------+--------------+--------------+--------------+---------------+---------------+------------------+----------------+----------------+
| is_true_t  | is_true_f  | is_true_n  | is_true_expr | not_true_t   | not_true_f   | not_true_n   | is_false_f  | is_false_t  | is_false_n  | not_false_f  | not_false_t  | not_false_n  | is_unknown_n  | is_unknown_t  | is_unknown_expr  | not_unknown_n  | not_unknown_t  |
+------------+------------+------------+--------------+--------------+--------------+--------------+-------------+-------------+-------------+--------------+--------------+--------------+---------------+---------------+------------------+----------------+----------------+
| TRUE       | FALSE      | FALSE      | TRUE         | FALSE        | TRUE         | TRUE         | TRUE        | FALSE       | FALSE       | FALSE        | TRUE         | TRUE         | TRUE          | FALSE         | TRUE             | FALSE          | TRUE           |
+------------+------------+------------+--------------+--------------+--------------+--------------+-------------+-------------+-------------+--------------+--------------+--------------+---------------+---------------+------------------+----------------+----------------+

结果说明:

  • ✅ IS TRUE: 仅当值为TRUE时返回TRUE
  • ✅ IS FALSE: 仅当值为FALSE时返回TRUE
  • ✅ IS UNKNOWN: 仅当值为NULL时返回TRUE
  • ✅ IS NOT TRUE = IS FALSE OR IS UNKNOWN
  • ✅ IS NOT FALSE = IS TRUE OR IS UNKNOWN

实际应用场景:

sql 复制代码 
-- 场景1: 严格的布尔检查（排除NULL）
SELECT * FROM products
WHERE (is_active IS TRUE) AND (is_deleted IS FALSE)

-- 场景2: 处理三值逻辑的CASE
SELECT user_id,
       CASE 
           WHEN age > 18 IS TRUE THEN 'Adult'
           WHEN age > 18 IS FALSE THEN 'Minor'
           WHEN age > 18 IS UNKNOWN THEN 'Unknown'
       END as age_group
FROM users

-- 场景3: 查找缺失的布尔值
SELECT * FROM config
WHERE enabled IS UNKNOWN  -- 找出NULL值

-- 场景4: 数据质量检查
SELECT 
    COUNT(*) as total,
    COUNT(*) FILTER (WHERE is_valid IS TRUE) as valid_count,
    COUNT(*) FILTER (WHERE is_valid IS FALSE) as invalid_count,
    COUNT(*) FILTER (WHERE is_valid IS UNKNOWN) as unknown_count
FROM data_records

3. 算术函数 (Arithmetic Functions)

3.1 函数清单

函数名称 中文说明 语法示例 返回类型
+ 加法 n1 + n2 NUMERIC
- 减法/取负 n1 - n2, -n NUMERIC
* 乘法 n1 * n2 NUMERIC
/ 除法 n1 / n2 NUMERIC
%, MOD 取模 n1 % n2, MOD(n1, n2) NUMERIC
POWER 幂运算 POWER(base, exp) DOUBLE
ABS 绝对值 ABS(n) NUMERIC
SQRT 平方根 SQRT(n) DOUBLE
LN 自然对数 LN(n) DOUBLE
LOG10 常用对数 LOG10(n) DOUBLE
LOG2 二进制对数 LOG2(n) DOUBLE
LOG 对数 LOG([base,] n) DOUBLE
EXP e的指数 EXP(n) DOUBLE
CEIL, CEILING 向上取整 CEIL(n) INTEGER
FLOOR 向下取整 FLOOR(n) INTEGER
ROUND 四舍五入 ROUND(n [, scale]) NUMERIC
TRUNCATE, TRUNC 截断 TRUNCATE(n [, scale]) NUMERIC
SIN 正弦 SIN(radian) DOUBLE
COS 余弦 COS(radian) DOUBLE
TAN 正切 TAN(radian) DOUBLE
COT 余切 COT(radian) DOUBLE
ASIN 反正弦 ASIN(n) DOUBLE
ACOS 反余弦 ACOS(n) DOUBLE
ATAN 反正切 ATAN(n) DOUBLE
ATAN2 两参数反正切 ATAN2(y, x) DOUBLE
SINH 双曲正弦 SINH(n) DOUBLE
COSH 双曲余弦 COSH(n) DOUBLE
TANH 双曲正切 TANH(n) DOUBLE
DEGREES 弧度转角度 DEGREES(radian) DOUBLE
RADIANS 角度转弧度 RADIANS(degree) DOUBLE
SIGN 符号函数 SIGN(n) INTEGER
PI 圆周率 PI() DOUBLE
E 自然常数 E() DOUBLE
RAND 随机数 RAND([seed]) DOUBLE
RAND_INTEGER 随机整数 RAND_INTEGER(bound) INTEGER

3.2 详细测试示例

测试 3.1: 基础算术运算

测试SQL:

sql 复制代码 
SELECT 
    -- 四则运算
    10 + 5 as addition,
    10 - 3 as subtraction,
    6 * 7 as multiplication,
    20 / 4 as integer_division,
    20.0 / 3.0 as decimal_division,
    
    -- 取模
    17 % 5 as modulo,
    MOD(17, 5) as modulo_function,
    -17 % 5 as negative_modulo,
    17.5 % 5.2 as decimal_modulo,
    
    -- 取负
    -100 as negation,
    -(5 + 3) as expression_negation,
    
    -- 运算符优先级
    10 + 5 * 2 as precedence_test,
    (10 + 5) * 2 as parentheses_test,
    
    -- 类型转换
    CAST(10 AS DOUBLE) / 3 as force_decimal,
    10 / CAST(3 AS DOUBLE) as force_decimal2
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+-----------+--------------+-----------------+-------------------+-------------------+---------+------------------+------------------+-----------------+-----------+----------------------+-----------------+--------------------+----------------+-----------------+
| addition  | subtraction  | multiplication  | integer_division  | decimal_division  | modulo  | modulo_function  | negative_modulo  | decimal_modulo  | negation  | expression_negation  | precedence_test | parentheses_test   | force_decimal  | force_decimal2  |
+-----------+--------------+-----------------+-------------------+-------------------+---------+------------------+------------------+-----------------+-----------+----------------------+-----------------+--------------------+----------------+-----------------+
| 15        | 7            | 42              | 5                 | 6.6666666667      | 2       | 2                | -2               | 2.1             | -100      | -8                   | 20              | 30                 | 3.3333333333   | 3.3333333333    |
+-----------+--------------+-----------------+-------------------+-------------------+---------+------------------+------------------+-----------------+-----------+----------------------+-----------------+--------------------+----------------+-----------------+

实际应用场景:

sql 复制代码 
-- 场景1: 订单金额计算
SELECT order_id,
       quantity,
       unit_price,
       quantity * unit_price as subtotal,
       quantity * unit_price * 1.13 as total_with_tax,
       quantity * unit_price * (1 - discount_rate) as discounted_amount
FROM order_items

-- 场景2: 分页计算
SELECT 
    page_number,
    page_size,
    (page_number - 1) * page_size as offset_value,
    page_size as limit_value
FROM pagination_config

-- 场景3: 百分比计算
SELECT 
    category,
    COUNT(*) as count,
    CAST(COUNT(*) AS DOUBLE) / (SELECT COUNT(*) FROM products) * 100 as percentage
FROM products
GROUP BY category
  • 算术函数高级用法和字符串函数详解

目录

  • [3.3 算术函数高级测试](#3.3 算术函数高级测试)
  • [4. 字符串函数 (String Functions)](#4. 字符串函数 (String Functions))

3.3 算术函数高级测试

测试 3.2: 幂运算和根运算

测试SQL:

sql 复制代码 
SELECT 
    -- 幂运算
    POWER(2, 10) as power_1024,
    POWER(5, 3) as power_125,
    POWER(10, -2) as power_negative_exp,
    POWER(4, 0.5) as power_sqrt,
    POWER(27, 1.0/3.0) as power_cube_root,
    
    -- 平方根
    SQRT(16) as sqrt_16,
    SQRT(2) as sqrt_2,
    SQRT(100) as sqrt_100,
    SQRT(0.25) as sqrt_decimal,
    
    -- 验证关系
    POWER(SQRT(16), 2) as verify_sqrt,
    SQRT(POWER(5, 2)) as verify_power
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+-------------+------------+----------------------+-------------+-------------------+---------+-----------+----------+--------------+--------------+--------------+
| power_1024  | power_125  | power_negative_exp   | power_sqrt  | power_cube_root   | sqrt_16 | sqrt_2    | sqrt_100 | sqrt_decimal | verify_sqrt  | verify_power |
+-------------+------------+----------------------+-------------+-------------------+---------+-----------+----------+--------------+--------------+--------------+
| 1024.0      | 125.0      | 0.01                 | 2.0         | 3.0               | 4.0     | 1.414214  | 10.0     | 0.5          | 16.0         | 5.0          |
+-------------+------------+----------------------+-------------+-------------------+---------+-----------+----------+--------------+--------------+--------------+

实际应用场景:

sql 复制代码 
-- 场景1: 复利计算
SELECT 
    initial_amount,
    annual_rate,
    years,
    initial_amount * POWER(1 + annual_rate, years) as final_amount,
    initial_amount * POWER(1 + annual_rate, years) - initial_amount as interest_earned
FROM investments

-- 场景2: 欧几里得距离
SELECT 
    p1_id, p2_id,
    SQRT(POWER(x2 - x1, 2) + POWER(y2 - y1, 2)) as distance_2d,
    SQRT(POWER(x2 - x1, 2) + POWER(y2 - y1, 2) + POWER(z2 - z1, 2)) as distance_3d
FROM point_pairs

-- 场景3: 标准差计算（手动）
SELECT 
    category,
    SQRT(AVG(POWER(value - avg_value, 2))) as std_deviation
FROM (
    SELECT category, value, AVG(value) OVER (PARTITION BY category) as avg_value
    FROM measurements
) T
GROUP BY category
测试 3.3: 对数和指数函数

测试SQL:

sql 复制代码 
SELECT 
    -- 自然对数（ln）
    LN(1) as ln_1,
    LN(E()) as ln_e,
    LN(10) as ln_10,
    LN(EXP(1)) as ln_exp_1,
    
    -- 常用对数（log10）
    LOG10(1) as log10_1,
    LOG10(10) as log10_10,
    LOG10(100) as log10_100,
    LOG10(1000) as log10_1000,
    
    -- 二进制对数（log2）
    LOG2(1) as log2_1,
    LOG2(2) as log2_2,
    LOG2(8) as log2_8,
    LOG2(1024) as log2_1024,
    
    -- 自定义底数
    LOG(2, 8) as log_2_8,
    LOG(3, 27) as log_3_27,
    LOG(10, 100) as log_10_100,
    
    -- 指数函数
    EXP(0) as exp_0,
    EXP(1) as exp_1,
    EXP(2) as exp_2,
    EXP(-1) as exp_neg_1,
    
    -- 常量
    PI() as pi_value,
    E() as e_value
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+-------+-------+-----------+-----------+-----------+-----------+------------+-------------+---------+---------+---------+------------+----------+-----------+-------------+-------+-----------+-----------+------------+-----------+-----------+
| ln_1  | ln_e  | ln_10     | ln_exp_1  | log10_1   | log10_10  | log10_100  | log10_1000  | log2_1  | log2_2  | log2_8  | log2_1024  | log_2_8  | log_3_27  | log_10_100  | exp_0 | exp_1     | exp_2     | exp_neg_1  | pi_value  | e_value   |
+-------+-------+-----------+-----------+-----------+-----------+------------+-------------+---------+---------+---------+------------+----------+-----------+-------------+-------+-----------+-----------+------------+-----------+-----------+
| 0.0   | 1.0   | 2.302585  | 1.0       | 0.0       | 1.0       | 2.0        | 3.0         | 0.0     | 1.0     | 3.0     | 10.0       | 3.0      | 3.0       | 2.0         | 1.0   | 2.718282  | 7.389056  | 0.367879   | 3.141593  | 2.718282  |
+-------+-------+-----------+-----------+-----------+-----------+------------+-------------+---------+---------+---------+------------+----------+-----------+-------------+-------+-----------+-----------+------------+-----------+-----------+

实际应用场景:

sql 复制代码 
-- 场景1: 增长率计算（对数变换）
SELECT 
    date,
    value,
    LOG10(value) as log_value,
    LN(value / LAG(value) OVER (ORDER BY date)) as growth_rate
FROM time_series

-- 场景2: 信息熵计算
SELECT 
    category,
    -SUM(probability * LOG2(probability)) as entropy
FROM probability_distribution
GROUP BY category

-- 场景3: pH值计算
SELECT 
    sample_id,
    hydrogen_concentration,
    -LOG10(hydrogen_concentration) as ph_value
FROM chemical_samples

-- 场景4: 指数衰减模型
SELECT 
    time_elapsed,
    initial_value * EXP(-decay_rate * time_elapsed) as current_value
FROM decay_process
测试 3.4: 取整函数详解

测试SQL:

sql 复制代码 
SELECT 
    -- CEIL/CEILING（向上取整）
    CEIL(3.1) as ceil_31,
    CEIL(3.9) as ceil_39,
    CEIL(-3.1) as ceil_neg31,
    CEIL(-3.9) as ceil_neg39,
    CEILING(3.5) as ceiling_35,
    
    -- FLOOR（向下取整）
    FLOOR(3.1) as floor_31,
    FLOOR(3.9) as floor_39,
    FLOOR(-3.1) as floor_neg31,
    FLOOR(-3.9) as floor_neg39,
    
    -- ROUND（四舍五入）
    ROUND(3.14159) as round_pi,
    ROUND(3.14159, 2) as round_pi_2,
    ROUND(3.14159, 4) as round_pi_4,
    ROUND(3.5) as round_35,
    ROUND(4.5) as round_45,
    ROUND(1234.5678, -2) as round_neg2,
    ROUND(-3.5) as round_neg35,
    
    -- TRUNCATE（截断）
    TRUNCATE(3.14159) as trunc_pi,
    TRUNCATE(3.14159, 2) as trunc_pi_2,
    TRUNCATE(3.14159, 4) as trunc_pi_4,
    TRUNCATE(-3.7) as trunc_neg37,
    TRUNCATE(1234.5678, -2) as trunc_neg2
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+----------+----------+-------------+-------------+-------------+-----------+-----------+-------------+-------------+----------+-------------+--------------+-----------+-----------+-------------+-------------+-----------+-----------+--------------+--------------+-----------+-------------+
| ceil_31  | ceil_39  | ceil_neg31  | ceil_neg39  | ceiling_35  | floor_31  | floor_39  | floor_neg31 | floor_neg39 | round_pi | round_pi_2  | round_pi_4   | round_35  | round_45  | round_neg2  | round_neg35 | trunc_pi  | trunc_pi_2 | trunc_pi_4   | trunc_neg37  | trunc_neg2 |
+----------+----------+-------------+-------------+-------------+-----------+-----------+-------------+-------------+----------+-------------+--------------+-----------+-----------+-------------+-------------+-----------+-----------+--------------+--------------+------------+
| 4        | 4        | -3          | -3          | 4           | 3         | 3         | -4          | -4          | 3        | 3.14        | 3.1416       | 4         | 5         | 1200        | -4          | 3         | 3.14      | 3.1415       | -3           | 1200       |
+----------+----------+-------------+-------------+-------------+-----------+-----------+-------------+-------------+----------+-------------+--------------+-----------+-----------+-------------+-------------+-----------+-----------+--------------+--------------+------------+

说明:

  • CEIL: 向正无穷方向取整（向上）
  • FLOOR: 向负无穷方向取整（向下）
  • ROUND: 四舍五入，可指定小数位数（负数表示整数部分）
  • TRUNCATE: 直接截断，不进行四舍五入

实际应用场景:

sql 复制代码 
-- 场景1: 价格向上取整
SELECT 
    product_id,
    calculated_price,
    CEIL(calculated_price) as display_price
FROM product_prices

-- 场景2: 计算页数
SELECT 
    CEIL(CAST(total_records AS DOUBLE) / page_size) as total_pages
FROM (
    SELECT COUNT(*) as total_records, 20 as page_size
    FROM large_table
) T

-- 场景3: 金额格式化（保留两位小数）
SELECT 
    order_id,
    ROUND(subtotal, 2) as subtotal,
    ROUND(tax, 2) as tax,
    ROUND(subtotal + tax, 2) as total
FROM orders

-- 场景4: 时间槽（整点分组）
SELECT 
    FLOOR(HOUR(event_time)) as hour_slot,
    COUNT(*) as event_count
FROM events
GROUP BY hour_slot
测试 3.5: 三角函数和反三角函数

测试SQL:

sql 复制代码 
SELECT 
    -- 常量
    PI() as pi,
    E() as e,
    
    -- 基础三角函数（输入弧度）
    SIN(0) as sin_0,
    SIN(PI()/6) as sin_30deg,
    SIN(PI()/4) as sin_45deg,
    SIN(PI()/2) as sin_90deg,
    
    COS(0) as cos_0,
    COS(PI()/3) as cos_60deg,
    COS(PI()/2) as cos_90deg,
    
    TAN(0) as tan_0,
    TAN(PI()/4) as tan_45deg,
    
    COT(PI()/4) as cot_45deg,
    COT(PI()/2) as cot_90deg,
    
    -- 反三角函数（输出弧度）
    ASIN(0) as asin_0,
    ASIN(0.5) as asin_05,
    ASIN(1) as asin_1,
    
    ACOS(1) as acos_1,
    ACOS(0.5) as acos_05,
    ACOS(0) as acos_0,
    
    ATAN(0) as atan_0,
    ATAN(1) as atan_1,
    
    ATAN2(1, 1) as atan2_1_1,
    ATAN2(1, 0) as atan2_1_0,
    ATAN2(0, 1) as atan2_0_1,
    
    -- 角度弧度转换
    DEGREES(PI()) as deg_180,
    DEGREES(PI()/2) as deg_90,
    RADIANS(180) as rad_180,
    RADIANS(90) as rad_90,
    
    -- 双曲函数
    SINH(0) as sinh_0,
    SINH(1) as sinh_1,
    COSH(0) as cosh_0,
    TANH(0) as tanh_0
FROM (VALUES (1)) AS T(x)

执行结果:

复制代码 
+----------+----------+--------+-----------+-----------+-----------+--------+-----------+-----------+--------+-----------+-----------+-----------+----------+-----------+----------+----------+-----------+----------+----------+----------+-----------+-----------+-----------+----------+---------+-----------+--------+---------+-----------+----------+-------+
| pi       | e        | sin_0  | sin_30deg | sin_45deg | sin_90deg | cos_0  | cos_60deg | cos_90deg | tan_0  | tan_45deg | cot_45deg | cot_90deg | asin_0   | asin_05   | asin_1   | acos_1   | acos_05   | acos_0   | atan_0   | atan_1   | atan2_1_1 | atan2_1_0 | atan2_0_1 | deg_180  | deg_90  | rad_180   | rad_90 | sinh_0  | sinh_1    | cosh_0  | tanh_0  |
+----------+----------+--------+-----------+-----------+-----------+--------+-----------+-----------+--------+-----------+-----------+-----------+----------+-----------+----------+----------+-----------+----------+----------+----------+-----------+-----------+-----------+----------+---------+-----------+--------+---------+-----------+---------+---------+
| 3.141593 | 2.718282 | 0.0    | 0.5       | 0.707107  | 1.0       | 1.0    | 0.5       | 0.0       | 0.0    | 1.0       | 1.0       | 0.0       | 0.0      | 0.523599  | 1.570796 | 0.0      | 1.047198  | 1.570796 | 0.0      | 0.785398 | 0.785398  | 1.570796  | 0.0       | 180.0    | 90.0    | 3.141593  | 1.5708 | 0.0     | 1.175201  | 1.0     | 0.0     |
+----------+----------+--------+-----------+-----------+-----------+--------+-----------+-----------+--------+-----------+-----------+-----------+----------+-----------+----------+----------+-----------+----------+----------+----------+-----------+-----------+-----------+----------+---------+-----------+--------+---------+-----------+---------+---------+

实际应用场景:

sql 复制代码 
-- 场景1: 地理坐标距离计算（Haversine公式）
SELECT 
    p1.id as point1,
    p2.id as point2,
    6371 * ACOS(
        SIN(RADIANS(p1.latitude)) * SIN(RADIANS(p2.latitude)) +
        COS(RADIANS(p1.latitude)) * COS(RADIANS(p2.latitude)) *
        COS(RADIANS(p2.longitude - p1.longitude))
    ) as distance_km
FROM geo_points p1
CROSS JOIN geo_points p2
WHERE p1.id < p2.id

-- 场景2: 方位角计算
SELECT 
    start_point,
    end_point,
    DEGREES(ATAN2(
        SIN(RADIANS(end_lon - start_lon)) * COS(RADIANS(end_lat)),
        COS(RADIANS(start_lat)) * SIN(RADIANS(end_lat)) -
        SIN(RADIANS(start_lat)) * COS(RADIANS(end_lat)) * COS(RADIANS(end_lon - start_lon))
    )) as bearing
FROM routes

-- 场景3: 周期性数据建模
SELECT 
    time_hour,
    temperature,
    20 + 10 * SIN(RADIANS(time_hour * 15 - 90)) as expected_temp
FROM hourly_weather
测试 3.6: 其他数学函数

测试SQL:

sql 复制代码 
SELECT 
    -- 绝对值
    ABS(10) as abs_pos,
    ABS(-10) as abs_neg,
    ABS(-3.14) as abs_decimal,
    
    -- 符号函数
    SIGN(10) as sign_pos,
    SIGN(-10) as sign_neg,
    SIGN(0) as sign_zero,
    SIGN(0.0001) as sign_small_pos,
    SIGN(-0.0001) as sign_small_neg,
    
    -- MOD函数
    MOD(17, 5) as mod_17_5,
    MOD(-17, 5) as mod_neg17_5,
    MOD(17, -5) as mod_17_neg5,
    MOD(-17, -5) as mod_neg17_neg5,
    17 % 5 as operator_mod,
    
    -- 随机数（每次执行结果不同）
    RAND() as random_float,
    RAND(42) as random_seeded,
    RAND_INTEGER(100) as random_int_100,
    RAND_INTEGER(10, 20) as random_int_range
FROM (VALUES (1)) AS T(x)

执行结果（随机数每次不同）:

复制代码 
+----------+----------+-------------+-----------+-----------+------------+----------------+------------------+-----------+--------------+-------------+------------------+---------------+-----------------+----------------+---------------+-------------------+
| abs_pos  | abs_neg  | abs_decimal | sign_pos  | sign_neg  | sign_zero  | sign_small_pos | sign_small_neg   | mod_17_5  | mod_neg17_5  | mod_17_neg5 | mod_neg17_neg5   | operator_mod  | random_float    | random_seeded  | random_int_100 | random_int_range  |
+----------+----------+-------------+-----------+-----------+------------+----------------+------------------+-----------+--------------+-------------+------------------+---------------+-----------------+----------------+---------------+-------------------+
| 10       | 10       | 3.14        | 1         | -1        | 0          | 1              | -1               | 2         | -2           | 2           | -2               | 2             | 0.634258        | 0.834729       | 57            | 14                |
+----------+----------+-------------+-----------+-----------+------------+----------------+------------------+-----------+--------------+-------------+------------------+---------------+-----------------+----------------+---------------+-------------------+

实际应用场景:

sql 复制代码 
-- 场景1: 价格差异计算
SELECT 
    product_id,
    current_price,
    standard_price,
    ABS(current_price - standard_price) as price_difference,
    SIGN(current_price - standard_price) as price_trend  -- 1=上涨, -1=下跌, 0=不变
FROM product_prices

-- 场景2: 随机抽样
SELECT * FROM large_dataset
WHERE RAND() < 0.1  -- 随机抽取约10%的数据

-- 场景3: 生成随机测试数据
SELECT 
    RAND_INTEGER(1, 1000000) as user_id,
    CONCAT('user', CAST(RAND_INTEGER(1000) AS STRING)) as username,
    RAND() * 1000 as random_score
FROM UNNEST(SEQUENCE(1, 100)) as T(i)

-- 场景4: 计算奇偶性
SELECT 
    number,
    MOD(number, 2) as remainder,
    CASE WHEN MOD(number, 2) = 0 THEN 'Even' ELSE 'Odd' END as parity
FROM numbers_table

-- 场景5: 周期性任务（每3行取一行）
SELECT *
FROM (
    SELECT *, ROW_NUMBER() OVER () as rn
    FROM source_table
) T
WHERE MOD(rn, 3) = 1

4. 字符串函数 (String Functions)

4.1 函数清单

函数名称 中文说明 语法示例 返回类型
CONCAT 字符串连接 CONCAT(str1, str2, ...) STRING
CONCAT_WS 使用分隔符连接 CONCAT_WS(sep, str1, ...) STRING
SUBSTRING/SUBSTR 提取子串 SUBSTRING(str, start[, len]) STRING
UPPER 转大写 UPPER(str) STRING
LOWER 转小写 LOWER(str) STRING
INITCAP 首字母大写 INITCAP(str) STRING
TRIM 去除首尾字符 `TRIM([BOTH LEADING
LTRIM 去除左侧空格 LTRIM(str) STRING
RTRIM 去除右侧空格 RTRIM(str) STRING
LENGTH/CHAR_LENGTH 字符串长度 LENGTH(str) INTEGER
POSITION 查找子串位置 POSITION(substr IN str) INTEGER
LOCATE 查找子串位置 LOCATE(substr, str[, start]) INTEGER
INSTR 查找子串位置 INSTR(str, substr) INTEGER
REPLACE 替换字符串 REPLACE(str, search, repl) STRING
REGEXP_REPLACE 正则替换 REGEXP_REPLACE(str, pattern, repl) STRING
REGEXP_EXTRACT 正则提取 REGEXP_EXTRACT(str, pattern[, group]) STRING
OVERLAY 覆盖字符串 OVERLAY(str PLACING new FROM start [FOR len]) STRING
LPAD 左填充 LPAD(str, length, pad) STRING
RPAD 右填充 RPAD(str, length, pad) STRING
REPEAT 重复字符串 REPEAT(str, times) STRING
REVERSE 反转字符串 REVERSE(str) STRING
SPLIT_INDEX 分割取元素 SPLIT_INDEX(str, sep, index) STRING
CHR ASCII转字符 CHR(ascii_code) STRING
ASCII 字符转ASCII ASCII(str) INTEGER
ENCODE 字符串编码 ENCODE(str, charset) BINARY
DECODE 字符串解码 DECODE(binary, charset) STRING
TO_BASE64 Base64编码 TO_BASE64(binary) STRING
FROM_BASE64 Base64解码 FROM_BASE64(str) BINARY
MD5 MD5哈希 MD5(str) STRING
SHA1 SHA1哈希 SHA1(str) STRING
SHA224/SHA256/SHA384/SHA512 SHA系列哈希 SHA256(str) STRING
UUID 生成UUID UUID() STRING

4.2 详细测试示例

测试 4.1: 字符串连接

测试SQL:

sql 复制代码 
SELECT 
    -- CONCAT：直接连接
    CONCAT('Hello', ' ', 'World') as concat_basic,
    CONCAT('User', '123') as concat_two,
    CONCAT('A', NULL, 'B') as concat_with_null,
    CONCAT(CAST(123 AS STRING), '-', CAST(456 AS STRING)) as concat_numbers,
    
    -- CONCAT_WS：使用分隔符
    CONCAT_WS('-', '2024', '10', '18') as concat_ws_date,
    CONCAT_WS(',', 'A', 'B', 'C', 'D') as concat_ws_list,
    CONCAT_WS(',', 'A', NULL, 'C') as concat_ws_skip_null,
    CONCAT_WS('|', 'Name', 'Age', 'City') as concat_ws_pipe,
    CONCAT_WS(' ', 'John', 'M.', 'Doe') as concat_ws_name,
    
    -- 实际应用：构建完整信息
    CONCAT_WS(' - ', brand, model, color) as product_display,
    CONCAT('https://', domain, '/', path) as full_url
FROM (VALUES 
    ('Apple', 'iPhone 15', 'Black', 'example.com', 'products/123')
) AS T(brand, model, color, domain, path)

执行结果:

复制代码 
+---------------+------------+-------------------+------------------+------------------+-----------------+----------------------+-----------------+----------------+-------------------------+------------------+
| concat_basic  | concat_two | concat_with_null  | concat_numbers   | concat_ws_date   | concat_ws_list  | concat_ws_skip_null  | concat_ws_pipe  | concat_ws_name | product_display         | full_url         |
+---------------+------------+-------------------+------------------+------------------+-----------------+----------------------+-----------------+----------------+-------------------------+------------------+
| Hello World   | User123    | NULL              | 123-456          | 2024-10-18       | A,B,C,D         | A,C                  | Name\|Age\|City | John M. Doe    | Apple - iPhone 15 - Black | https://example.com/products/123 |
+---------------+------------+-------------------+------------------+------------------+-----------------+----------------------+-----------------+----------------+-------------------------+------------------+

实际应用场景:

sql 复制代码 
-- 场景1: 生成完整地址
SELECT 
    user_id,
    CONCAT_WS(', ', country, province, city, district, street, building) as full_address
FROM addresses

-- 场景2: 生成显示名称
SELECT 
    product_id,
    CONCAT(brand, ' ', model, ' (', color, ', ', size, ')') as display_name
FROM products

-- 场景3: 生成CSV行
SELECT 
    CONCAT_WS(',', 
        CAST(id AS STRING),
        CONCAT('"', name, '"'),  -- 字符串加引号
        email,
        phone
    ) as csv_line
FROM users

-- 场景4: 构建SQL语句
SELECT 
    CONCAT('UPDATE users SET status = ''active'' WHERE id = ', CAST(user_id AS STRING)) as update_sql
FROM pending_activations
相关推荐
luoganttcc4 小时前
是凯恩斯主义主导 西方的经济决策吗
大数据·人工智能·金融·哲学
武子康4 小时前
大数据-130 - Flink CEP 详解 - 捕获超时事件提取全解析：从原理到完整实战代码教程 恶意登录案例实现
大数据·后端·flink
分布式存储与RustFS5 小时前
存算一体架构的先行者：RustFS在异构计算环境下的探索与实践
大数据·人工智能·物联网·云原生·对象存储·minio·rustfs
小小测试开发5 小时前
Python SQLAlchemy：告别原生 SQL，用 ORM 优雅操作数据库
数据库·python·sql·sqlalchemy
智能化咨询6 小时前
Flink受管状态自定义序列化的可演进设计与未来趋势
flink
yumgpkpm7 小时前
CMP （类ClouderaCDP7.3（404次编译） ）华为鲲鹏Aarch64（ARM）信创环境 查询2100w行 hive 查询策略
数据库·数据仓库·hive·hadoop·flink·mapreduce·big data
wudl556610 小时前
Flink 1.20 flink-config.yml 配置详解
大数据·flink
华东数交10 小时前
企业与国有数据资产：入表全流程管理及资产化闭环理论解析
大数据·人工智能
B站_计算机毕业设计之家15 小时前
计算机毕业设计：Python农业数据可视化分析系统 气象数据 农业生产 粮食数据 播种数据 爬虫 Django框架 天气数据 降水量（源码+文档）✅
大数据·爬虫·python·机器学习·信息可视化·课程设计·农业
热门推荐
01GitHub 镜像站点02BongoCat - 跨平台键盘猫动画工具03UV安装并设置国内源04GitLab 零基础入门指南：从安装到项目管理全流程05Linux下V2Ray安装配置指南06KGG转MP3工具|非KGM文件|解密音频07两千字总结：Codex 国内如何安装和使用的教程，以及如何设置中文回答082025软件测试面试八股文（含答案+文档）09荣耀手机2025年10月发布的新品Magic8比起Magic7，在硬件、性能、价格等上有什么区别，有什么优势10NVIDIA显卡驱动、CUDA、cuDNN 和 TensorRT 版本匹配指南