This is a simple MySQL Abstraction Layer compatible with PHP 7+ & PHP 8.0 that provides a simple and secure interaction with your database using mysqli_* functions at its core. This is perfect for small scale applications such as cron jobs, facebook canvas campaigns or micro frameworks or sites.
You can also use the 💍 "Simple Active Record"-class, it's based on this db class and add some OOP syntax. But please inform you about "Active Record" vs "Data Mapper" before you use it.
You can download it from here, or require it using composer.
{
"require": {
"voku/simple-mysqli": "8.*"
}
}
composer require voku/simple-mysqli
- Starting the driver
- Multiton && Singleton
- Doctrine/DBAL as parent driver
- Using the "DB"-Class
- Using the "Result"-Class
- Using the "Prepare"-Class
- Logging and Errors
- Changelog
use voku\db\DB;
require_once 'composer/autoload.php';
$db = DB::getInstance('yourDbHost', 'yourDbUser', 'yourDbPassword', 'yourDbName');
// example
// $db = DB::getInstance('localhost', 'root', '', 'test');
You can use DB::getInstance()
without any parameters and you will get your (as "singleton") first initialized connection. Or you can change the parameter and you will create an new "multiton"-instance which works like an singleton, but you need to use the same parameters again, otherwise (without the same parameter) you will get an new instance.
use voku\db\DB;
require_once 'composer/autoload.php';
$connectionParams = [
'dbname' => 'yourDbName',
'user' => 'yourDbUser',
'password' => 'yourDbPassword',
'host' => 'yourDbHost',
'driver' => 'mysqli', // 'pdo_mysql' || 'mysqli'
'charset' => 'utf8mb4',
];
$config = new \Doctrine\DBAL\Configuration();
$doctrineConnection = \Doctrine\DBAL\DriverManager::getConnection(
$connectionParams,
$config
);
$doctrineConnection->connect();
$db = DB::getInstanceDoctrineHelper($doctrineConnection);
There are numerous ways of using this library, here are some examples of the most common methods.
use voku\db\DB;
$db = DB::getInstance();
$result = $db->query("SELECT * FROM users");
$users = $result->fetchAll();
But you can also use a method for select-queries:
$db->select(string $table, array $where); // generate an SELECT query
Example: SELECT
$where = [
'page_type =' => 'article',
'page_type NOT LIKE' => '%öäü123',
'page_id >=' => 2,
];
$articles = $db->select('page', $where);
echo 'There are ' . count($articles) . ' article(s):' . PHP_EOL;
foreach ($articles as $article) {
echo 'Type: ' . $article['page_type'] . PHP_EOL;
echo 'ID: ' . $article['page_id'] . PHP_EOL;
}
Here is a list of connectors for the "WHERE"-array: 'NOT', 'IS', 'IS NOT', 'IN', 'NOT IN', 'BETWEEN', 'NOT BETWEEN', 'LIKE', 'NOT LIKE', '>', '<', '>=', '<=', '<>', '+', '-'
INFO: use an array as $value for "[NOT] IN" and "[NOT] BETWEEN"
INFO: use + / - in the value not in the key of the $data
Example: UPDATE with "page_template = page_template + 1"
$where = [
'page_type LIKE' => '%foo',
'page_type NOT LIKE' => 'bar',
];
$data = [
'page_template' => ['page_template +' => 1],
'page_type' => 'lall',
];
$resultSelect = $db->update('page', $data, $where);
Example: SELECT with "NOT IN"
$where = [
'page_type NOT IN' => [
'foo',
'bar'
],
'page_id >' => 2,
];
$resultSelect = $db->select('page', $where);
Example: SELECT with Cache
$resultSelect = $db->execSQL("SELECT * FROM users", true, 3600);
The result (via $result->fetchAllArray()) is only cached for 3600s when the query was a SELECT statement, otherwise you get the default result from the $db->query()
function.
to manipulate tables you have the most important methods wrapped, they all work the same way: parsing arrays of key/value pairs and forming a safe query
the methods are:
$db->insert( string $table, array $data ); // generate an INSERT query
$db->replace( string $table, array $data ); // generate an REPLACE query
$db->update( string $table, array $data, array $where ); // generate an UPDATE query
$db->delete( string $table, array $where ); // generate a DELETE query
All methods will return the resulting mysqli_insert_id()
or true/false depending on context.
The correct approach if to always check if they executed as success is always returned
Example: DELETE
$deleteArray = ['user_id' => 9];
$ok = $db->delete('users', $deleteArray);
if ($ok) {
echo "user deleted!";
} else {
echo "can't delete user!";
}
note: all parameter values are sanitized before execution, you don't have to escape values beforehand.
Example: INSERT
$insertArray = [
'name' => "John",
'email' => "[email protected]",
'group' => 1,
'active' => true,
];
$newUserId = $db->insert('users', $insertArray);
if ($newUserId) {
echo "new user inserted with the id $new_user_id";
}
Example: REPLACE
$replaceArray = [
'name' => 'lars',
'email' => '[email protected]',
'group' => 0
];
$tmpId = $db->replace('users', $replaceArray);
Binding parameters is a good way of preventing mysql injections as the parameters are sanitized before execution.
$sql = "SELECT * FROM users
WHERE id_user = :id_user
AND active = :active
LIMIT 1
";
$result = $db->query($sql, ['id_user' => 11, 'active' => 1]);
if ($result) {
$user = $result->fetchArray();
print_r($user);
} else {
echo "user not found";
}
Use begin()
, commit()
, and rollback()
to manage transactions:
$db->beginTransaction();
$db->query(
'UPDATE `users` SET `foo` = :foo WHERE id = :id',
['foo' => 100, 'id' => 1]
);
$db->query(
'UPDATE `users_noop` SET `foo` = :foo WHERE id = :id',
['foo' => 100, 'id' => 2]
);
$db->endTransaction();
Any SQL errors between begin()
and commit()
will yield a RuntimeException
.
You can also use the DB->transact()
method. The following is equivalent
to the above:
$db->transact(function($db) {
$db->query(
'UPDATE `users` SET `foo` = :foo WHERE id = :id',
['foo' => 100, 'id' => 1]
);
$db->query(
'UPDATE `users_noop` SET `foo` = :foo WHERE id = :id',
['foo' => 100, 'id' => 2]
);
});
After executing a SELECT
query you receive a Result
object that will help you manipulate the resultant data.
there are different ways of accessing this data, check the examples bellow:
$result = $db->query("SELECT * FROM users");
$allUsers = $result->fetchAll();
Fetching all data works as Result::RESULT_TYPE_* the fetchAll()
and fetch()
method will return the default based on the $_default_result_type
config.
Other methods are:
$row = $result->fetch(); // fetch an single result row as defined by the config (array, object or Arrayy)
$row = $result->fetchArray(); // fetch an single result row as array
$row = $result->fetchArrayy(); // fetch an single result row as Arrayy object
$row = $result->fetchObject(); // fetch an single result row as object
$row = $result->fetchYield(); // fetch an single result row as Generator
$data = $result->fetchAll(); // fetch all result data as defined by the config (array, object or Arrayy)
$data = $result->fetchAllArray(); // fetch all result data as array
$data = $result->fetchAllArrayy(); // fetch all result data as Array object
$data = $result->fetchAllObject(); // fetch all result data as object
$data = $result->fetchAllYield(); // fetch all result data as Generator
$data = $result->fetchColumn(string $column, bool $skipNullValues); // fetch a single column as string
$data = $result->fetchAllColumn(string $column, bool $skipNullValues); // fetch a single column as an 1-dimension array
$data = $result->fetchArrayPair(string $key, string $value); // fetch data as a key/value pair array
Returns rows of field information in a result set:
$fields = $result->fetchFields();
Pass true
as argument if you want each field information returned as an
associative array instead of an object. The default is to return each as an
object, exactly like the mysqli_fetch_fields
function.
Fetches a row or a single column within a row:
$data = $result->fetch($row_number, $column);
This method forms the basis of all fetch_ methods. All forms of fetch_ advances
the internal row pointer to the next row. null
will be returned when there are
no more rows to be fetched.
Returns all rows at once, transposed as an array of arrays:
$plan_details = $plans->fetchTranspose();
Transposing a result set of X rows each with Y columns will result in an array of Y rows each with X columns.
Pass a column name as argument to return each column as an associative array with keys taken from values of the provided column. If not provided, the keys will be numeric starting from zero.
e.g.:
$transposedExample = [
'title' => [
1 => 'Title #1',
2 => 'Title #2',
3 => 'Title #3',
],
);
Returns all rows at once as key-value pairs using the column in the first argument as the key:
$countries = $result->fetchPairs('id');
Pass a column name as the second argument to only return a single column as the value in each pair:
$countries = $result->fetchPairs('id', 'name');
/*
[
1 => 'Title #1',
2 => 'Title #2',
3 => 'Title #3',
]
*/
Returns all rows at once as a grouped array:
$students_grouped_by_gender = $result->fetchGroups('gender');
Pass a column name as the second argument to only return single columns as the values in each groups:
$student_names_grouped_by_gender = $result->fetchGroups('gender', 'name');
Returns the first row element from the result:
$first = $result->first();
Pass a column name as argument to return a single column from the first row:
$name = $result->first('name');
Returns the last row element from the result:
$last = $result->last();
Pass a column name as argument to return a single column from the last row:
$name = $result->last('name');
Returns a slice of rows from the result:
$slice = $result->slice(1, 10);
The above will return 10 rows skipping the first one. The first parameter is the
zero-based offset; the second parameter is the number of elements; the third
parameter is a boolean value to indicate whether to preserve the keys or not
(optional and defaults to false). This methods essentially behaves the same as
PHP's built-in array_slice()
function.
Sets a mapper callback function that's used inside the Result->fetchCallable()
method:
$result->map(function($row) {
return (object) $row;
});
$object = $result->fetchCallable(0);
The above example will map one row (0) from the result into a object. Set the mapper callback function to null to disable it.
$db->get() // alias for $db->fetch();
$db->getAll() // alias for $db->fetchAll();
$db->getObject() // alias for $db->fetchAllObject();
$db->getArray() // alias for $db->fetchAllArray();
$db->getArrayy() // alias for $db->fetchAllArrayy();
$db->getYield() // alias for $db->fetchAllYield();
$db->getColumn($key) // alias for $db->fetchColumn($key);
To iterate a result-set you can use any fetch() method listed above.
$result = $db->select('users');
// using while
while ($row = $result->fetch()) {
echo $row->name;
echo $row->email;
}
// using foreach (via "fetchAllObject()")
foreach($result->fetchAllObject() as $row) {
echo $row->name;
echo $row->email;
}
// using foreach (via "Result"-object)
foreach($result as $row) {
echo $row->name;
echo $row->email;
}
// using foreach (via "Generator"-object)
foreach($result->fetchAllYield() as $row) {
echo $row->name;
echo $row->email;
}
// INFO: "while + fetch()" and "fetchAllYield()" will use less memory that "foreach + "fetchAllObject()", because we will fetch each result entry seperatly
To execute multiple queries you can use the $db->multi_query()
method. You can use multiple queries separated by ";
".
Return-Types:
- "Result"-Array by "SELECT"-queries
- "bool" by only "INSERT"-queries
- "bool" by only (affected_rows) by "UPDATE / DELETE"-queries
- "bool" by only by e.g. "DROP"-queries
e.g.:
$sql = "
INSERT INTO foo
SET
page_template = 'lall1',
page_type = 'test1';
INSERT INTO lall
SET
page_template = 'lall2',
page_type = 'test2';
INSERT INTO bar
SET
page_template = 'lall3',
page_type = 'test3';
";
$result = $this->db->multi_query($sql); // true
$sql = "
SELECT * FROM foo;
SELECT * FROM lall;
SELECT * FROM bar;
";
$result = $this->db->multi_query($sql); // Result[]
foreach ($result as $resultForEach) {
$tmpArray = $resultForEach->fetchArray();
...
}
Prepare statements have the advantage that they are built together in the MySQL-Server, so the performance is better.
But the debugging is harder and logging is impossible (via PHP), so we added a wrapper for "bind_param" called "bind_param_debug". With this wrapper we pre-build the sql-query via php (only for debugging / logging). Now you can e.g. echo the query.
INFO: You can still use "bind_param" instead of "bind_param_debug", e.g. if you need better performance.
use voku\db\DB;
$db = DB::getInstance();
// -------------
// prepare the queries
$query = 'INSERT INTO users
SET
name = ?,
email = ?
';
$prepare = $db->prepare($query);
$name = '';
$email = '';
$prepare->bind_param_debug('ss', $name, $email);
// -------------
// execute query no. 1
// INFO: "$template" and "$type" are references, since we use "bind_param" or "bind_param_debug"
$name = 'name_1_中';
$email = '[email protected]';
$prepare->execute();
// DEBUG
echo $prepare->get_sql_with_bound_parameters();
// -------------
// execute query no. 2
// INFO: "$template" and "$type" are references, since we use "bind_param" or "bind_param_debug"
$name = 'Lars';
$email = '[email protected]';
$prepare->execute();
// DEBUG
echo $prepare->get_sql_with_bound_parameters();
use voku\db\DB;
$db = DB::getInstance();
// -------------
// insert some dummy-data, first
$data = [
'page_template' => 'tpl_test_new123123',
'page_type' => 'ö\'ä"ü',
];
// will return the auto-increment value of the new row
$resultInsert[1] = $db->insert($this->tableName, $data);
$resultInsert[2] = $db->insert($this->tableName, $data);
// -------------
// prepare the queries
$sql = 'SELECT * FROM ' . $this->tableName . '
WHERE page_id = ?
';
$prepare = $this->db->prepare($sql);
$page_id = 0;
$prepare->bind_param_debug('i', $page_id);
// -------------
// execute query no. 1
$page_id = $resultInsert[1];
$result = $prepare->execute();
$data = $result->fetchArray();
// $data['page_template'] === 'tpl_test_new123123'
// $data['page_id'] === $page_id
// -------------
// execute query no. 2
$page_id = $resultInsert[2];
$result = $prepare->execute();
$data = $result->fetchArray();
// $data['page_id'] === $page_id
// $data['page_template'] === 'tpl_test_new123123'
You can hook into the "DB"-Class, so you can use your personal "Logger"-Class. But you have to cover the methods:
$this->trace(string $text, string $name) { ... }
$this->debug(string $text, string $name) { ... }
$this->info(string $text, string $name) { ... }
$this->warn(string $text, string $name) { ... }
$this->error(string $text, string $name) { ... }
$this->fatal(string $text, string $name) { ... }
You can also disable the logging of every sql-query, with the "getInstance()"-parameter "logger_level" from "DB"-Class. If you set "logger_level" to something other than "TRACE" or "DEBUG", the "DB"-Class will log only errors anymore.
DB::getInstance(
getConfig('db', 'hostname'), // hostname
getConfig('db', 'username'), // username
getConfig('db', 'password'), // password
getConfig('db', 'database'), // database
getConfig('db', 'port'), // port
getConfig('db', 'charset'), // charset
true, // exit_on_error
true, // echo_on_error
'cms\Logger', // logger_class_name
getConfig('logger', 'level'), // logger_level | 'TRACE', 'DEBUG', 'INFO', 'WARN', 'ERROR', 'FATAL'
getConfig('session', 'db') // session_to_db
);
Showing the query log: The log comes with the SQL executed, the execution time and the result row count.
print_r($db->log());
To debug mysql errors, use $db->errors()
to fetch all errors (returns false if there are no errors) or $db->lastError()
for information about the last error.
if ($db->errors()) {
echo $db->lastError();
}
But the easiest way for debugging is to configure "DB"-Class via "DB::getInstance()" to show errors and exit on error (see the example above). Now you can see SQL-errors in your browser if you are working on "localhost" or you can implement your own "checkForDev()" via a simple function, you don't need to extend the "Debug"-Class. If you will receive error-messages via e-mail, you can implement your own "mailToAdmin()"-function instead of extending the "Debug"-Class.
See CHANGELOG.md.
For support and donations please visit Github | Issues | PayPal | Patreon.
For status updates and release announcements please visit Releases | Twitter | Patreon.
For professional support please contact me.
- Thanks to GitHub (Microsoft) for hosting the code and a good infrastructure including Issues-Managment, etc.
- Thanks to IntelliJ as they make the best IDEs for PHP and they gave me an open source license for PhpStorm!
- Thanks to Travis CI for being the most awesome, easiest continous integration tool out there!
- Thanks to StyleCI for the simple but powerfull code style check.
- Thanks to PHPStan && Psalm for relly great Static analysis tools and for discover bugs in the code!