REGEXP_SUBSTR(description,'ird|irs'): Treat a pattern as a group : REGEXP_SUBSTR « Regular Expressions « Oracle PL / SQL






REGEXP_SUBSTR(description,'ird|irs'): Treat a pattern as a group



SQL> create table TestTable(
  2    ID                    VARCHAR2(4 BYTE)         NOT NULL,
  3    Description           VARCHAR2(30 BYTE)
  4  )
  5  /

Table created.

SQL>
SQL>
SQL> insert into TestTable (ID, Description) values('1','1234 5th Street');

1 row created.

SQL> insert into TestTable (ID, Description) values('2','1 Culloden Street');

1 row created.

SQL> insert into TestTable (ID, Description) values('3','1234 Road');

1 row created.

SQL> insert into TestTable (ID, Description) values('4','33 Thrid Road');

1 row created.

SQL> insert into TestTable (ID, Description) values('5','One than another');

1 row created.

SQL> insert into TestTable (ID, Description) values('6','2003 Movie');

1 row created.

SQL> insert into TestTable (ID, Description) values('7','Start With Letters');

1 row created.

SQL>
SQL> select * from TestTable;

ID   DESCRIPTION
---- ------------------------------
1    1234 5th Street
2    1 Culloden Street
3    1234 Road
4    33 Thrid Road
5    One than another
6    2003 Movie
7    Start With Letters

7 rows selected.

SQL>
SQL> -- Grouping
SQL>
SQL> -- Treat a pattern as a group.
SQL>
SQL> -- Find all occurrences of the letter sequence 'irs' or 'ird'
SQL>
SQL> SELECT description, REGEXP_SUBSTR(description,'ird|irs')
  2  FROM testTable;

DESCRIPTION                    REGEXP_SUBSTR(DESCRIPTION,'IRD
------------------------------ ------------------------------
1234 5th Street
1 Culloden Street
1234 Road
33 Thrid Road
One than another
2003 Movie
Start With Letters

7 rows selected.

SQL>
SQL> drop table TestTable;

Table dropped.

SQL>
SQL>
SQL>
           
       








Related examples in the same category

1.REGEXP_SUBSTR('Yababa dababa do','(ab)\1')
2.REGEXP_SUBSTR('Mississippi', 'si', 1, 2, 'i')