postgresql序列号(SERIAL)类型包括smallserial(smallint,short),serial(int)和bigserial(bigint,long long int),不管是smallserial,serial还是bigserial,其范围都是(1,9223372036854775807),但是序列号类型其实不是真正的类型,当声明一个字段为序列号类型时其实是创建了一个序列,INSERT时如果没有给该字段赋值会默认获取对应序列的下一个值。
测试表1:
test=# create table tbl_serial(a serial,b varchar(2)); CREATE TABLE test=# \d tbl_serial Table "public.tbl_serial" Column | Type | Modifiers --------+----------------------+-------------------------------------------------------- a | integer | not null default nextval('tbl_serial_a_seq'::regclass) b | character varying(2) |
从结果中看,声明字段a为serial类型,会自动创建一个名为tbl_serial_a_seq的序列,INSERT时缺省为该序列的下一个序列值nextval。
自动创建的序列如下定义:
test=# \d tbl_serial_a_seq Sequence "public.tbl_serial_a_seq" Column | Type | Value ---------------+---------+--------------------- sequence_name | name | tbl_serial_a_seq last_value | bigint | 1 start_value | bigint | 1 increment_by | bigint | 1 max_value | bigint | 9223372036854775807 min_value | bigint | 1 cache_value | bigint | 1 log_cnt | bigint | 0 is_cycled | boolean | f is_called | boolean | f Owned by: public.tbl_serial.a
其实也可以先创建一个序列,然后将表的某字段默认值设为该序列的下一个序列值。
测试表2:
test=# create sequence sql_tbl_serial2_a increment by 1 minvalue 1 no maxvalue start with 1; CREATE SEQUENCE test=# create table tbl_serial2(a int not null default nextval('sql_tbl_serial2_a'),b varchar(2)); CREATE TABLE test=# \d tbl_serial2 Table "public.tbl_serial2" Column | Type | Modifiers --------+----------------------+--------------------------------------------------------- a | integer | not null default nextval('sql_tbl_serial2_a'::regclass) b | character varying(2) | test=# \d sql_tbl_serial2_a Sequence "public.sql_tbl_serial2_a" Column | Type | Value ---------------+---------+--------------------- sequence_name | name | sql_tbl_serial2_a last_value | bigint | 1 start_value | bigint | 1 increment_by | bigint | 1 max_value | bigint | 9223372036854775807 min_value | bigint | 1 cache_value | bigint | 1 log_cnt | bigint | 0 is_cycled | boolean | f is_called | boolean | f
此方法和第一种直接使用serial类型效果完全相同,但是这里可以自己定义序列名称。
创建序列的语法如下:
test=# \h create sequence Command: CREATE SEQUENCE Description: define a new sequence generator Syntax: CREATE [ TEMPORARY | TEMP ] SEQUENCE [ IF NOT EXISTS ] name [ INCREMENT [ BY ] increment ] [ MINVALUE minvalue | NO MINVALUE ] [ MAXVALUE maxvalue | NO MAXVALUE ] [ START [ WITH ] start ] [ CACHE cache ] [ [ NO ] CYCLE ] [ OWNED BY { table_name.column_name | NONE } ]
其实和上面使用\d一个序列时对应的,
INCREMENT BY : 每次序列增加(或减少)的步长
MINVALUE : 序列最小值,NO MINVALUE表示没有最小值
MAXVALUE : 序列最大值,NO MAXVALUE表示没有最大值
START WITH :以什么序列值开始
CYCLE : 序列是否循环使用
OWNED BY : 可以直接指定一个表的字段,也可以不指定。
序列涉及的函数有:
函 数 | 返 回 类 型 | 描 述 |
currval( regclass ) | bigint | 获取指定序列最近一次使用netxval后的数值,如果没有使用nextval而直接使用currval会出错。 |
lastval() | bigint | 返回最近一次用 nextval 获取的任意序列的数值 |
nextval( regclass ) | bigint | 递增序列并返回新值 |
setval( regclass,bigint ) | bigint | 设置序列的当前数值 |
setval( regclass,bigint ,boolean ) | bigint | 设置序列的当前数值以及 is_called 标志,如果为true则立即生效,如果为false,则调用一次nextval后才会生效。 |
示例1:获取序列tbl_serial_a_seq当前序列值
test=# select currval('tbl_serial_a_seq'); ERROR: currval of sequence "tbl_serial_a_seq" is not yet defined in this session test=# test=# select nextval('tbl_serial_a_seq'); nextval --------- 1 (1 row) test=# select currval('tbl_serial_a_seq'); currval --------- 1 (1 row)
示例2:获取最近一次使用nextval的返回值
test=# select nextval('sql_tbl_serial2_a'); nextval --------- 5 (1 row) test=# select lastval(); lastval --------- 5 (1 row) test=# select nextval('tbl_serial_a_seq'); nextval --------- 3 (1 row) test=# select lastval(); lastval --------- 3 (1 row)
示例3:将序列tbl_serial_a_seq当前值设为100
test=# select setval('tbl_serial_a_seq',100); setval -------- 100 (1 row) test=# select currval('tbl_serial_a_seq'); currval --------- 100 (1 row)
示例4:使用is_called标志设置序列值是马上生效(true)还是下次生效(false)
test=# select setval('tbl_serial_a_seq',200,true); setval -------- 200 (1 row) test=# select currval('tbl_serial_a_seq'); currval --------- 200 (1 row) test=# select setval('tbl_serial_a_seq',300,false); setval -------- 300 (1 row) test=# select currval('tbl_serial_a_seq'); currval --------- 200 (1 row) test=# select nextval('tbl_serial_a_seq'); nextval --------- 300 (1 row) test=# select currval('tbl_serial_a_seq'); currval --------- 300 (1 row)
修改序列和创建序列的语法基本相同,只是用ALTER替换了CREATE,请参考
test=# \h alter sequence Command: ALTER SEQUENCE Description: change the definition of a sequence generator Syntax: ALTER SEQUENCE [ IF EXISTS ] name [ INCREMENT [ BY ] increment ] [ MINVALUE minvalue | NO MINVALUE ] [ MAXVALUE maxvalue | NO MAXVALUE ] [ START [ WITH ] start ] [ RESTART [ [ WITH ] restart ] ] [ CACHE cache ] [ [ NO ] CYCLE ] [ OWNED BY { table_name.column_name | NONE } ]
如果一个序列是NO CYCLE,当序列值全部使用完会怎样呢?
test=# create sequence seq_test increment by 2 MINVALUE 1 MAXVALUE 5 START WITH 2 NO CYCLE; CREATE SEQUENCE test=# \d seq_test Sequence "public.seq_test" Column | Type | Value ---------------+---------+---------- sequence_name | name | seq_test last_value | bigint | 2 start_value | bigint | 2 increment_by | bigint | 2 max_value | bigint | 5 min_value | bigint | 1 cache_value | bigint | 1 log_cnt | bigint | 0 is_cycled | boolean | f is_called | boolean | f test=# select nextval('seq_test'); nextval --------- 2 (1 row) test=# select nextval('seq_test'); nextval --------- 4 (1 row) test=# select nextval('seq_test'); ERROR: nextval: reached maximum value of sequence "seq_test" (5)
如果序列值用完了当然是错误了!!!如果是CYCLE则会重新从START处开始再次循环。