[Nomad PHP EU] Introduction to Databases

This talk will cover the basics necessary to help you decide what data to store, where, and how.

We will cover PDO — PHP’s Data Object extension, which allows you to talk to a variety of databases, including MySQL.

You will learn how to CRUD — Create, Retrieve, Update and Delete data, database schema, and when to use indexes.

Additionally, we will look at other data storage systems such as memcache, and nosql databases.

Transcript

1. Databases Getting Started with Data
2. •Community Engineer at Engine Yard •Author of Zend PHP 5 Certification Study Guide, Sitepoints PHP Anthology: 101 Essential Tips, Tricks & Hacks & PHP Master: Write Cutting Edge Code •A contributor to Zend Framework 1 & 2, phpdoc, and PHP internals • @dshafik Davey Shafik
3. What is a Database?
4. What is a Database? A database is an organized collection of data. The data is typically organized to model relevant aspects of reality, in a way that supports processes requiring this information. “ ” Source: Wikipedia (Emphasis Mine)
5. Types of Database
6. SQL (Relational)
7. SQL (Relational) • Highly Structured Data • Using Tables, Columns and Rows • One or more relationships exist between datas • Constraints • Primary Keys (a unique row identifier) • Unique Keys (one or more columns that must have unique values, either individually, or as a group) • Foreign Keys (a column value that must be derived from a column value in another table) • Indexes • A lookup for one, or multiple columns aggregate data !7
8. NoSQL
   (Document/Key-Value/Graph)
9. NoSQL (Document/Key-Value/Graph) • Sometimes called “Not Only SQL” because some NoSQL DBs have a SQL-like query language • Not always non-relational • Always unstructured • Intended to provide higher scalability and higher availability • Looser consistency models !9
10. Non-Relational Document Stores / Key-Value Stores
11. NoSQL (Document/Key-Value/Graph) • NoSQL is non-relational • Document Stores • Centers around the concept of a document, and it’s related meta-data • Collections of documents • Hierarchies of documents • Examples: Couchbase Server, CouchDB, MongoDB, Amazon SimpleDB, Oracle NoSQL DB • Key-Value Stores • Data stored and accessible directly by a unique key • Examples: Memcache, MongoDB, Couchbase Server, Cassandra, Riak, Amazon DynamoDB, Redis, Oracle NoSQL DB !11
12. Relational Graph Databases
13. NoSQL (Document/Key-Value/Graph) • NoSQL is relational (say what?!) • Graph Databases • All data is related to N other data • Relationships are in the data, not indexes • Examples: OQGraph for MySQL • Example Implementation: Facebook’s Graph API !13
14. A Note about MySQL
15. A Note on MySQL • MySQL supports multiple drivers (called engines) for it’s tables. • These engine provide different features. • The two most common are InnoDB (default since MySQL 5.5) and MyISAM (previously the default). • InnoDB has far more features, and is recommended for almost all situations • We will assume InnoDB for all MySQL examples !15
16. Relational Concepts
17. Relational Concepts • Schema • Tables • Indexes • Relationships • Stored Procedures • Triggers !17
18. Data Types
19. Name What Notes int exact whole numbers Signed or unsigned decimal exact decimal numbers (fixed length) Signed or unsigned float approximate decimal number (variable length) Signed or unsigned char strings (fixed length) Max size: 255 bytes varchar strings (variable length) Max size: 255 bytes text strings (variable length) Max size: 255 bytes - 4GB blob binary strings (variable length) Max size: 255 Bytes - 4GB date dates (no time) Any date is valid datetime dates (with time) Any date/time is valid timestamp timestamp UNIX timestamp, must be > 1/1/1970 NULL Null values Does not equal anything, even NULL
20. Create a Users Table
21. Users Table • Unique Identifier • Username • Password • Email Address • Name or First Name/Last Name !21 Consider: • Column Names • Column Types • Column Lengths
22. Users Table !22 Users id int username varchar(20) password varchar(60) email varchar(150) first_name varchar(45) last_name varchar(55)
23. Users Table !23 Users id int username varchar(20) password varchar(60) email varchar(150) first_name varchar(45) last_name varchar(55) CREATE TABLE ( , , , , , );
24. Users Table (Schema) CREATE TABLE users ( id INT, username VARCHAR(20), password VARCHAR(60), email VARCHAR(150), first_name VARCHAR(45), last_name VARCHAR(55) );
25. SQL Structured Query Language
26. SQL • INSERT — Create Data • UPDATE — Update Existing Data • SELECT — Fetch Data • DELETE — Delete Data !26 Four Main Queries
27. CRUD Also Known As:
28. CRUD !28 Create INSERT R etrieve SELECT Update UPDATE Delete DELETE
29. Conditions
30. Conditions • Used with: • SELECT • UDPATE • DELETE • JOINs • Preceded by the WHERE, ON, or USING keyword !30
31. Operators !31 Operator = Equality <>, != Inequality < Less Than <= Less Than or Equal To > Greater Than >= Greater Than or Equal To IS NULL NULL Equality IS NOT NULL NULL Inequality AND Boolean AND OR Boolean OR BETWEEN Range Equality
32. INSERT
33. INSERT INSERT INTO table name ( list, of, columns ) VALUES ( "list", "of", "values" ); !33
34. INSERT INSERT INTO users ( id, username, password, email, first_name, last_name ) VALUES ( 1, "dshafik", "$2y$10$Ol/KS4/Bhs5ENUh7OpIDL.Gs1SIWDG.rPaBkPAjjQ2UTITI60YDmG ", "davey@engineyard.com", "Davey", "Shafik" ); !34
35. UPDATE
36. UPDATE UPDATE table name SET column = "some", name = "value" WHERE some condition;
37. WARNING: Don’t forget your conditions! Otherwise you update every row in the table!
38. UPDATE UPDATE users SET username = "davey", email = "davey@engineyard.com" WHERE id = 1;
39. SELECT
40. SELECT SELECT list, of, columns FROM table WHERE column = "some" AND name = "value" OR other_column = "other value" ORDER BY some ASC, columns DESC LIMIT start, end;
41. SELECT SELECT    *   FROM    users   WHERE    username  =  "davey"    AND  password  =  "$2y$10$Ol..."   LIMIT  1;
42. SELECT SELECT first_name, last_name, email FROM users ORDER BY first_name, last_name LIMIT 0, 10;
43. SELECT SELECT first_name, last_name, email FROM users ORDER BY first_name, last_name LIMIT 10, 20;
44. DELETE
45. DELETE DELETE FROM table WHERE column = "some" AND name = "value" OR other_column = "other value" ORDER BY some ASC, columns DESC LIMIT number;
46. DELETE DELETE FROM users;
48. DELETE DELETE FROM users WHERE id = 1;
49. Constraints
50. Constraints: Users Table • IDs should be unique • Usernames should be unique • Passwords should not be unique • Email Address should be unique • First Name should not be unique • Last Name should not be unique ! • All column should not be NULL !50
51. Constraints: Users Table !51 Users Constraints id int unique, not null username varchar(20) unique, not null password varchar(60) not null email varchar(150) unique, not null first_name varchar(45) not null last_name varchar(55) not null
52. Constraints: Users Table Schema CREATE TABLE users ( id INT NOT NULL, username VARCHAR(20) NOT NULL, password VARCHAR(60) NOT NULL, email VARCHAR(150) NOT NULL, first_name VARCHAR(45) NOT NULL, last_name VARCHAR(55) NOT NULL, UNIQUE INDEX id_UNIQUE (id), UNIQUE INDEX username_UNIQUE (username), UNIQUE INDEX email_UNIQUE (email) );
53. Features
54. Name What Notes Auto Increment (auto_increment) Automatically inserts the (last row)+1 value when inserting • Column must be set as PRIMARY KEY • One Per Table Signed/Unsigned Sets Numeric columns to signed (may be positive or negative) or unsigned (must be positive) Unsigned numbers start at 0 and allow for much larger numbers. Zero Fill (zerofill) Left Pads numeric values to the column size Only applied on retrieval (i.e. it’s not stored this way)
55. Features: Users Table •ID should be auto increment •ID should be the Primary Key •ID should be unsigned !55
56. Features: Users Table Schema CREATE TABLE users ( id INT UNSIGNED NOT NULL AUTO_INCREMENT, username VARCHAR(20) NOT NULL, password VARCHAR(60) NOT NULL, email VARCHAR(150) NOT NULL, first_name VARCHAR(45) NOT NULL, last_name VARCHAR(55) NOT NULL, UNIQUE INDEX username_UNIQUE (username), UNIQUE INDEX email_UNIQUE (email), PRIMARY KEY (id) );
57. Indexes
58. Indexes !58 Name Constraints Notes Index None May have NULL values Unique Unique May have NULL values Primary Key Unique Must NOT have NULL values. May auto_increment. There can only be one. Foreign Key Must match data in linked table May have NULL values
59. INDEX, UNIQUE,
    & PRIMARY KEY
60. INDEX, UNIQUE, & PRIMARY KEY • Can be added during table creation: • CREATE TABLE foo (
 column_name TYPE,
 INDEX name (column, list),
 UNIQUE name (column, list),
 PRIMARY KEY (column)
 ); • Index/Unique can have multiple columns • Can be added after table creation: • CREATE INDEX name ON table_name (column, list);
 ALTER TABLE table name ADD INDEX (column, list);
 CREATE UNIQUE INDEX name ON table (column, list);
 ALTER TABLE table name ADD UNIQUE (column, list);
 ALTER TABLE table name ADD PRIMARY KEY (column); • Must be added with caution! !60
61. Foreign Keys
62. Foreign Keys • Used to create inter-table relationships • Value must be in the foreign table or NULL • Can update when the foreign table updates • Can delete when the foreign table deletes • Can be set to NULL when the foreign table deletes !62
63. Profiles & Foreign Keys
64. Profiles Table • Unique Identifier • Short Introductory Summary • Full Biography • Location !64 Consider: • Must link to the Users table • One Profile per User
65. Profiles Table Profiles Constraints id int UNSIGNED primary key, auto_increment, not null users_id int UNSIGNED unique, foreign key -> users.id, not null summary varchar(200) none bio TEXT none location varchar(100) none
66. Profiles Table Schema CREATE TABLE profiles ( id INT UNSIGNED NOT NULL AUTO_INCREMENT, users_id INT UNSIGNED NOT NULL, summary VARCHAR(200) NOT NULL, bio TEXT NULL, location VARCHAR(150) NULL, PRIMARY KEY (id), UNIQUE INDEX users_id_UNIQUE (users_id), CONSTRAINT FOREIGN KEY (users_id) REFERENCES users (id) ON DELETE CASCADE );
67. INSERT INSERT INTO profiles ( users_id, summary, bio, location ) VALUES ( 1, "Community Engineer at Engine Yard", NULL, "Florida, USA" );
68. Indexing Your Data
70. Indexing Your Data • Indexes make writes slower, and reads (much faster). The more indexes, the slower your writes. • The creation of indexes should be determined by the SELECT queries being run upon the data. Nothing else. • For example, if you run a query that SELECTs using two WHERE criteria with an AND condition, that is probably a good combination index. • MySQL can only use one index [per table] at a time and will (generally) pick the best option based on the query • Indexes cannot be used with LIKE if starting with a wildcard (e.g. %foo%) !70
71. Indexes
72. Indexes • Add Indexes to the users table • Remember we already have a PRIMARY KEY and UNIQUE indexes • Example queries we will perform against it: !72 SELECT  *  FROM    users   WHERE    username  =  "davey"
  AND  password  =  "$2y$10$Ol..."; SELECT  *  FROM  users     WHERE  email  =  "davey@engineyard.com"; SELECT  *  FROM  users  
    WHERE  first_name  LIKE  "%Dave%";
73. Indexes !73 Users Key Type id PRIMARY KEY username UNIQUE username, password UNIQUE email UNIQUE
74. JOINs Connecting Tables
75. JOINs • Used to JOIN multiple tables • INNER JOIN • LEFT or RIGHT OUTER JOIN !75
76. INNER JOIN Get the intersection of two tables Users Profiles Users with Profiles
77. OUTER JOIN For when one side or the other doesn’t match
78. LEFT OUTER JOIN For when one side or the other doesn’t match Users Users with Profiles Profiles
79. RIGHT OUTER JOIN Users Posts with Users Posts
80. SELECT... INNER JOIN SELECT * FROM users INNER JOIN profiles ON ( profiles.user_id = users.id ) WHERE 
 profiles.location LIKE '%Dallas%' ORDER BY 
 users.first_name, users.last_name;
81. SELECT... LEFT OUTER JOIN SELECT * FROM users LEFT OUTER JOIN profiles ON ( profiles.user_id = users.id ) WHERE users.id = 1;
82. SELECT... RIGHT OUTER JOIN SELECT * FROM users
 LEFT OUTER JOIN profiles ON ( profiles.user_id = users.id ) RIGHT OUTER JOIN posts ON ( posts.user_id = users.id ) WHERE posts.content LIKE '%PHP%';
83. Databases and PHP
84. Connecting to Databases Using PDO
85. Connecting to Databases • PDO • MySQL • PostgreSQL • MSSQL • Oracle • SQLite • ODBC and DB2 • Firebird • DSN — Data Source Name • Driver Name • Hostname & Port • or • Unix Socket • Username • Password • Database Name • Charset • Connecting • new PDO() • Disconnecting • PDO::__destruct() !85
86. Connecting to MySQL <?php $pdo = new PDO( "mysql:dbname=db;host=localhost", "user", "pass" ); ?>
87. Querying Data
88. Executing Queries <?php $pdo = new PDO(...); $query = $pdo->prepare( "SELECT * FROM user WHERE id = :id" ); $conditions = array( ':id' => 1 ); $result = $query->execute($conditions); ?>
89. Handling Results
90. Handling Results <?php $result = $query->execute($conditions); if ($result) {
    echo  "Results  Found:  "  .$query-­‐>rowCount(); while ($row = $query->fetch()) { echo "<a href='/edit/" .$row['id']. "'>" .$row['first_name']. ' ' . $row['last_name'] .'</a>'; } } ?>
91. Handling Results as Objects <?php $result = $query->execute($conditions); if ($result) { echo  "Results  Found:  "  .$query-­‐>rowCount(); while ($row = $query->fetchObject()) { echo "<a href='/edit/" .$row->id. "'>" .$row->first_name. ' ' . $row->last_name .'</a>'; } } ?>
92. Handling Results as Custom Objects class User { function getName() { return $this->first_name . ' ' . $this->last_name; } } if ($result) { echo  "Results  Found:  "  .$query-­‐>rowCount(); while ($row = $query->fetchObject("User")) { echo "<a href='/edit/" .$row->id. "'>" .$row->getName(). '</a>'; } }
93. Thank You Feedback: https://joind.in/9909 Twitter: @dshafik Email: davey@engineyard.com ! Slides: http://daveyshafik.com/slides