实例表：
postgres=# \d book
Table "public.book"
Column | Type | Modifiers
--------+-------------------+---------------------------------------------------
id | integer | not null default nextval('book_id_seq'::regclass)
name | character varying |
tag | character varying |
postgres=# select * from book; id | name | tag ----+------+---------- 1 | java | aa,bb,cc 2 | C++ | dd,ee (2 rows)需要的效果：
name | unnest ------+-------- java | aa java | bb java | cc C++ | dd C++ | ee (5 rows)
方法1： unnest(数组，分隔符)
postgres=# select name,unnest(string_to_array(tag,',')) from book; name | unnest ------+-------- java | aa java | bb java | cc C++ | dd C++ | ee (5 rows)
postgres=# select string_to_array(tag,') from book;
string_to_array
-----------------
{aa,cc}
{dd,ee}
(2 rows)
方法2：regexp_split_to_table(字符串，分隔符)
postgres=# select name,regexp_split_to_table(tag,') from book; name | regexp_split_to_table ------+----------------------- java | aa java | bb java | cc C++ | dd C++ | ee (5 rows)
另外，字符串变数组的方法：
postgres=# select name,string_to_array(tag,') from book;
name | string_to_array
------+-----------------
java | {aa,cc}
C++ | {dd,ee}
(2 rows)
postgres=# select name,regexp_split_to_array(tag,') from book;
name | regexp_split_to_array
------+-----------------------
java | {aa,cc}
C++ | {dd,ee}
(2 rows)