MySQL/Tutorials

From Citizendium
< MySQL
Revision as of 07:43, 28 March 2008 by imported>Eric Evers (→‎Joins)
Jump to navigation Jump to search

MySQL tutorial

Intall

Create User

Create Database

Queries

Joins

Lets use two tables to join: table name and table phone.

Left outer Joins, Right outer Joins, Full outer Joins and Inner Joins in SQL

This is a table of names.

mysql> select * from name order by name_id; 
+---------+-------+-------+ 
| name_id | first | last  | 
+---------+-------+-------+
|       1 | joe   | smith |
|       2 | alice | jones |
|       3 | fred  | black |
|       4 | jane  | doe   |
+---------+-------+-------+

This is a table of phone numbers.

mysql> select * from  phone;
+-------+--------+---------+
| ph_id | number | name_id |
+-------+--------+---------+
|     1 |    100 |       1 |
|     2 |    101 |       1 |
|     3 |    102 |       2 |
|     4 |    103 |       3 |
|     5 |    104 |    NULL |
+-------+--------+---------+

A right outer join follows. It shows all the phone numbers.

mysql> select last, number from name right join phone 
on name.name_id = phone.name_id;
+-------+--------+
| last  | number |
+-------+--------+
| smith |    100 |
| smith |    101 |
| jones |    102 |
| black |    103 |
| NULL  |    104 |
+-------+--------+

A left outer join follows. It shows all the names.

mysql> select last, number from name left join phone 
on name.name_id = phone.name_id;
+-------+--------+
| last  | number |
+-------+--------+
| smith |    100 |
| smith |    101 |
| doe   |   NULL |
| jones |    102 |
| black |    103 |
+-------+--------+

An inner join follows. It shows only names that have numbers and only numbers that have names.

mysql> select last, number from name join phone 
on name.name_id = phone.name_id;
+-------+--------+
| last  | number |
+-------+--------+
| smith |    100 |
| smith |    101 |
| jones |    102 |
| black |    103 |
+-------+--------+

A full outer join follows. It shows all names and numbers.

mysql> select last, number from name left join phone 

on name.name_id = phone.name_id union select last, number from name right join phone on name.name_id=phone.name_id;

+-------+--------+
| last  | number |
+-------+--------+
| smith |    100 |
| smith |    101 |
| doe   |   NULL |
| jones |    102 |
| black |    103 |
| NULL  |    104 |
+-------+--------+

Here is some help for creating the source tables:

DROP TABLE IF EXISTS `name`; CREATE TABLE `name` (

 `name_id` int(10) default NULL,
 `first` varchar(20) default NULL,
 `last` varchar(20) default NULL

);

INSERT INTO `name` VALUES (1,'joe','smith'),(4,'jane','doe'),(2,'alice','jones'),(3,'fred','black');

DROP TABLE IF EXISTS `phone`; CREATE TABLE `phone` (

 `ph_id` int(10) default NULL,
 `number` int(10) default NULL,
 `name_id` int(10) default NULL

);

INSERT INTO `phone` VALUES (1,100,1),(2,101,1),(3,102,2),(4,103,3),(5,104,NULL);

Sub Quries

Aggragate Functions

Aggragate functions are functions that combine multiple rows into one output value. Nulls are ignored by aggragate functions unless all the input is NULL in which case the answer is NULL.

Table: Product
+----+---------------+-------+
| Id | Name          | Price |
+----+---------------+-------+
|  1 | bread         | 3.00  |
|  2 | peanut_butter | 3.25  |
|  3 | jam           | 2.75  |
+----+---------------+-------+

Sum

SELECT SUM(Price) as "Sum" FROM Product;
+------+
| sum  |
+------+
| 9.00 |  
+------+

Average

SELECT AVG(Price) as "Avg" FROM Product;
+------+
| Avg  |
+------+
| 3.00 |  
+------+

Min

SELECT MIN(Price) as "Min" FROM Product;
+------+
| Min  |
+------+
| 2.75 |  
+------+

Max

SELECT MAX(Price) as "Max" FROM Product;
+------+
| Max  |
+------+
| 3.25 |  
+------+

Std

Standard Deviation

SELECT STD(Price) AS "Std" FROM Product;
+------+
| Std  |
+------+
| 0.25 |  
+------+

String functions

Pattern Matching

Using Like
Using RegExp

Regular expressions in SQL

Lets create a simple table.

CREATE TABLE word (name varchar(20));
INSERT into word VALUES ("people"),("places"),("things"),("ppl");
SELECT * FROM word;
+--------+
| name   | 
+--------+
| people |
| places |
| things |
| ppl    |
+--------+

Look for two p's in a row.

SELECT * FROM word WHERE name RegExp "[p]{2}" 
+--------+
| name   | 
+--------+
| ppl    |
+--------+

A "." is any character. A "+" is one or more copies of a character. A "C{n}" looks for n copies of C. Look for two p's but not next to one another.

SELECT * FROM word WHERE name RegExp "p.+p" 
+--------+
| name   | 
+--------+
| people |
+--------+
Exercises

Give a sql regular expression query that will select:

1) only things  
2) only ppl and places
3) only people and places
4) only things and places