ascetic-soft / rowcast
Lightweight DataMapper and QueryBuilder over PDO with DTO hydration and type casting
Requires
- php: >=8.4
- ext-pdo: *
Requires (Dev)
- friendsofphp/php-cs-fixer: ^3.0
- phpstan/phpstan: ^2.0
- phpunit/phpunit: ^11.0
README
Lightweight DataMapper over PDO for PHP 8.4+.
Rowcast maps database rows to DTOs and back using reflection, supports explicit/auto mapping, type conversion, and includes a fluent query builder with dialect-aware UPSERT.
Documentation: English | Русский
Requirements
- PHP >= 8.4
ext-pdo
Installation
composer require ascetic-soft/rowcast
Quick Start
use AsceticSoft\Rowcast\Connection; use AsceticSoft\Rowcast\DataMapper; class UserDto { public int $id; public string $email; public bool $isActive; } $connection = Connection::create('sqlite::memory:'); $mapper = new DataMapper($connection); $user = new UserDto(); $user->email = 'alice@example.com'; $user->isActive = true; $id = $mapper->insert('users', $user); $found = $mapper->findOne(UserDto::class, ['id' => (int) $id]);
Core Concepts
Rowcast supports two mapping styles:
- Auto mapping — pass
class-stringfor reads and table name for writes. Names are converted viaNameConverterInterface(default:SnakeCaseToCamelCase). - Explicit mapping — pass
Mappingto control table name, column/property pairs, and ignored properties.
Auto Mapping
Table name is derived from DTO class name for reads:
| Class | Table |
|---|---|
User |
users |
UserProfile |
user_profiles |
Column/property conversion (default):
| Column | Property |
|---|---|
created_at |
createdAt |
is_active |
isActive |
Explicit Mapping
use AsceticSoft\Rowcast\Mapping; $mapping = Mapping::auto(UserDto::class, 'custom_users') ->column('usr_email', 'email') ->ignore('internalNote'); $user = $mapper->findOne($mapping, ['id' => 1]);
Use Mapping::explicit(...) when only declared columns must be used:
$mapping = Mapping::explicit(UserDto::class, 'custom_users') ->column('id', 'id') ->column('usr_email', 'email');
Connection
Connection wraps PDO and provides query helpers, transaction API, nested transaction support (savepoints), and query-builder factory.
Create Connection
use AsceticSoft\Rowcast\Connection; // From DSN $connection = Connection::create( dsn: 'mysql:host=localhost;dbname=app', username: 'root', password: 'secret', nestTransactions: true, ); // From existing PDO $pdo = new \PDO('sqlite::memory:'); $connection = new Connection($pdo, nestTransactions: true);
Raw Queries
$stmt = $connection->executeQuery('SELECT * FROM users WHERE id = ?', [1]); $affected = $connection->executeStatement('UPDATE users SET email = ? WHERE id = ?', ['a@x.com', 1]); $rows = $connection->fetchAllAssociative('SELECT * FROM users'); $row = $connection->fetchAssociative('SELECT * FROM users WHERE id = ?', [1]); $count = $connection->fetchOne('SELECT COUNT(*) FROM users');
Transactions
$connection->transactional(function (Connection $conn) { $conn->executeStatement('INSERT INTO users (email) VALUES (?)', ['alice@example.com']); });
Nested mode creates savepoints for inner transactions.
DataMapper API
Main methods:
insert(string|Mapping $target, object $dto): string|falsebatchInsert(string|Mapping $target, array $dtos, ?int $maxBindParameters = null): voidupdate(string|Mapping $target, object $dto, array $where): intbatchUpdate(string|Mapping $target, array $dtos, array $identityProperties, ?int $maxBindParameters = null): voiddelete(string|Mapping $target, array $where): intfindAll(string|Mapping $target, array $where = [], array $orderBy = [], ?int $limit = null, ?int $offset = null): arrayiterateAll(string|Mapping $target, array $where = [], array $orderBy = [], ?int $limit = null, ?int $offset = null): iterablefindOne(string|Mapping $target, array $where = []): ?objectsave(string|Mapping $target, object $dto, string ...$identityProperties): voidupsert(string|Mapping $target, object $dto, string ...$conflictProperties): intbatchUpsert(string|Mapping $target, array $dtos, array $conflictProperties, ?int $maxBindParameters = null): voidhydrate(...),hydrateAll(...),extract(...)
CRUD Example
$dto = new UserDto(); $dto->email = 'alice@example.com'; $dto->isActive = true; $id = $mapper->insert('users', $dto); $one = $mapper->findOne(UserDto::class, ['id' => (int) $id]); $one->isActive = false; $mapper->update('users', $one, ['id' => $one->id]); $mapper->delete('users', ['id' => $one->id]);
save(...) Example
save(...) checks row existence by identity columns, then performs insert or update.
$mapper->save('users', $dto, 'id');
upsert(...) Example
$affected = $mapper->upsert('users', $dto, 'email');
Batch operations
$mapper->batchInsert('users', [$dto1, $dto2, $dto3]); $mapper->batchUpsert('users', [$dto1, $dto2, $dto3], ['id']); $mapper->batchUpdate('users', [$dto1, $dto2, $dto3], ['id']); // Optional override for chunk sizing by bind parameter limit $mapper->batchInsert('users', [$dto1, $dto2, $dto3], 500); $mapper->batchUpsert('users', [$dto1, $dto2, $dto3], ['id'], 500);
batchInsert and batchUpsert automatically split large input into chunks based on DB parameter limits (for example, SQLite: 999 bind params), while executing all chunks inside one transaction.
Advanced where in DataMapper
DataMapper passes where arrays to the same QueryBuilder condition engine, so advanced operators are available there as well:
$users = $mapper->findAll(UserDto::class, where: [ 'deleted_at' => null, '$or' => [ ['status' => ['active', 'pending']], ['role' => 'admin'], ], 'age >=' => 18, ]);
Type Conversion
Rowcast converts DB values to declared PHP property types on hydrate, and PHP values to DB-safe values on extract/write.
Built-in converters:
ScalarConverter(int,float,string)BoolConverter(bool<->0/1)DateTimeConverter(DateTimeInterface<-> formatted UTC string)JsonConverter(array<-> JSON)EnumConverter(BackedEnum<-> backing value)
Custom Type Converter
Implement TypeConverterInterface and pass a custom registry to DataMapper:
use AsceticSoft\Rowcast\DataMapper; use AsceticSoft\Rowcast\TypeConverter\TypeConverterInterface; use AsceticSoft\Rowcast\TypeConverter\TypeConverterRegistry; final class UuidConverter implements TypeConverterInterface { public function supports(string $phpType): bool { return $phpType === Uuid::class; } public function toPhp(mixed $value, string $phpType): mixed { return new Uuid((string) $value); } public function toDb(mixed $value): mixed { return (string) $value; } } $converters = TypeConverterRegistry::defaults()->add(new UuidConverter()); $mapper = new DataMapper($connection, typeConverter: $converters);
Custom Name Converter
Implement NameConverterInterface and pass it to DataMapper:
use AsceticSoft\Rowcast\DataMapper; use AsceticSoft\Rowcast\NameConverter\NameConverterInterface; final class PrefixedConverter implements NameConverterInterface { public function toPropertyName(string $columnName): string { return lcfirst(str_replace('usr_', '', $columnName)); } public function toColumnName(string $propertyName): string { return 'usr_' . $propertyName; } } $mapper = new DataMapper($connection, nameConverter: new PrefixedConverter());
Query Builder
Connection::createQueryBuilder() provides a fluent SQL builder.
SELECT
$rows = $connection->createQueryBuilder() ->select('u.id', 'u.email') ->from('users', 'u') ->where('u.is_active = :active') ->orderBy('u.id', 'DESC') ->setOffset(20) ->setLimit(10) ->setParameter('active', 1) ->fetchAllAssociative();
For pagination, use:
setOffset(int $offset)— start rowsetLimit(int $limit)— max rows
You can also pass associative arrays to where(), andWhere(), and orWhere():
$rows = $connection->createQueryBuilder() ->select('*') ->from('users') ->where(['email' => 'alice@example.com', 'is_active' => 1]) ->fetchAllAssociative(); // SQL: SELECT * FROM users WHERE email = :w_email AND is_active = :w_is_active
array predicates are converted to field = :param expressions joined by AND:
where(['a' => 1, 'b' => 2])->a = :w_a AND b = :w_bandWhere(['a' => 1])appends anotherANDblockorWhere(['a' => 1])wraps previous predicate:(prev OR a = :w_a)
Parameter names are generated automatically and made unique (:w_id, :w_id_1, ...).
Supported array operators:
// IS NULL / IS NOT NULL ->where(['deleted_at' => null]) // deleted_at IS NULL ->where(['deleted_at !=' => null]) // deleted_at IS NOT NULL // IN / NOT IN ->where(['status' => ['active', 'pending']]) // status IN (...) ->where(['status !=' => ['banned']]) // status NOT IN (...) ->where(['status IN' => ['active']]) // explicit IN ->where(['status NOT IN' => ['banned']]) // explicit NOT IN // BackedEnum in WHERE (direct QueryBuilder usage) ->where(['status' => UserStatus::Active]) // status = :w_status, parameter value: 'active' ->where(['status' => [UserStatus::Active, UserStatus::Inactive]]) // status IN (...), parameters: 'active', 'inactive' // Comparison operators ->where(['age >' => 18, 'age <=' => 65, 'score !=' => 0]) // LIKE / ILIKE / NOT LIKE / NOT ILIKE ->where(['name LIKE' => '%alice%']) ->where(['name ILIKE' => '%alice%']) // useful for PostgreSQL ->where(['name NOT LIKE' => '%bot%']) ->where(['name NOT ILIKE' => '%bot%']) // PostgreSQL only // BETWEEN ->where(['age BETWEEN' => [18, 65]])
Operator reference:
| Input | Example | SQL fragment (shape) |
|---|---|---|
| Equality | ['id' => 10] |
id = :w_id |
IS NULL |
['deleted_at' => null] |
deleted_at IS NULL |
IS NOT NULL |
['deleted_at !=' => null] |
deleted_at IS NOT NULL |
IN (auto) |
['status' => ['active', 'pending']] |
status IN (:w_status, :w_status_1) |
NOT IN (auto) |
['status !=' => ['banned']] |
status NOT IN (:w_status, ...) |
IN (explicit) |
['status IN' => ['active']] |
status IN (:w_status) |
NOT IN (explicit) |
['status NOT IN' => ['banned']] |
status NOT IN (:w_status) |
| Comparison | ['age >=' => 18] |
age >= :w_age |
LIKE |
['name LIKE' => 'A%'] |
name LIKE :w_name |
ILIKE |
['name ILIKE' => 'a%'] |
name ILIKE :w_name |
NOT LIKE |
['name NOT LIKE' => '%bot%'] |
name NOT LIKE :w_name |
NOT ILIKE |
['name NOT ILIKE' => '%bot%'] |
name NOT ILIKE :w_name |
BETWEEN |
['age BETWEEN' => [18, 65]] |
age BETWEEN :w_age AND :w_age_1 |
Notes:
- Empty
INarray compiles to1 = 0(always false). - Empty
NOT INarray compiles to1 = 1(always true). ILIKEandNOT ILIKEare PostgreSQL-specific.BackedEnumvalues are normalized to backing scalar values inWHEREparameters (includingIN/NOT INarrays).
Dialect-specific operator support:
| Dialect | Extra operators over base set |
|---|---|
PostgreSQL (pgsql) |
ILIKE, NOT ILIKE |
MySQL (mysql) |
none |
SQLite (sqlite) |
none |
| Generic/other drivers | none |
Base set for all dialects: >, >=, <, <=, LIKE, NOT LIKE.
OR Conditions
You can compose OR logic in two ways.
Method-based OR groups:
// (status = 'active' AND age > 18) OR (role = 'admin') $rows = $connection->createQueryBuilder() ->select('*') ->from('users') ->whereOr( ['status' => 'active', 'age >' => 18], ['role' => 'admin'], ) ->fetchAllAssociative();
Combine with existing filters:
// deleted_at IS NULL AND ((status = 'active') OR (role = 'admin')) $rows = $connection->createQueryBuilder() ->select('*') ->from('users') ->where(['deleted_at' => null]) ->andWhereOr(['status' => 'active'], ['role' => 'admin']) ->fetchAllAssociative();
Nested-key style in one array:
$rows = $connection->createQueryBuilder() ->select('*') ->from('users') ->where([ 'age >' => 18, '$or' => [ ['status' => 'active'], ['$and' => [ ['role' => 'admin'], ['verified' => true], ]], ], ]) ->fetchAllAssociative();
OR composition reference:
| Pattern | Example | SQL fragment (shape) |
|---|---|---|
whereOr(...groups) |
->whereOr(['status' => 'active'], ['role' => 'admin']) |
((status = :w_status) OR (role = :w_role)) |
andWhereOr(...groups) |
->where(['deleted_at' => null])->andWhereOr(['status' => 'active'], ['role' => 'admin']) |
deleted_at IS NULL AND ((status = :w_status) OR (role = :w_role)) |
$or inside where([...]) |
->where(['age >' => 18, '$or' => [['status' => 'active'], ['role' => 'admin']]]) |
age > :w_age AND ((status = :w_status) OR (role = :w_role)) |
$and inside $or |
->where(['$or' => [['status' => 'active'], ['$and' => [['role' => 'admin'], ['verified' => true]]]]]) |
((status = :w_status) OR ((role = :w_role) AND (verified = :w_verified))) |
| Mixed operators in OR groups | ->whereOr(['status' => ['active', 'pending'], 'deleted_at' => null], ['name LIKE' => 'A%', 'age BETWEEN' => [18, 65]]) |
((status IN (...) AND deleted_at IS NULL) OR (name LIKE :w_name AND age BETWEEN :w_age AND :w_age_1)) |
INSERT / UPDATE / DELETE
$connection->createQueryBuilder() ->insert('users') ->values(['email' => ':email', 'is_active' => ':is_active']) ->setParameter('email', 'alice@example.com') ->setParameter('is_active', 1) ->executeStatement(); $connection->createQueryBuilder() ->update('users') ->set('is_active', ':is_active') ->where('id = :id') ->setParameter('is_active', 0) ->setParameter('id', 1) ->executeStatement(); $connection->createQueryBuilder() ->delete('users') ->where('id = :id') ->setParameter('id', 1) ->executeStatement();
UPSERT
$sql = $connection->createQueryBuilder() ->upsert('users') ->values(['email' => ':email', 'name' => ':name']) ->onConflict('email') ->doUpdateSet(['name']) ->getSQL();
Upsert is compiled via SQL dialects:
MysqlDialectPostgresDialectSqliteDialectGenericDialect(throws for unsupported UPSERT)
WHERE array operator support is also dialect-aware (for example, ILIKE/NOT ILIKE only for PostgreSQL).
Architecture
AsceticSoft\Rowcast\
├── ConnectionInterface
├── Connection
├── DataMapper
├── Hydrator
├── Extractor
├── Mapping
├── TargetResolver
├── QueryHelper
├── NameConverter\
│ ├── NameConverterInterface
│ └── SnakeCaseToCamelCase
├── TypeConverter\
│ ├── TypeConverterInterface
│ ├── TypeConverterRegistry
│ ├── ScalarConverter
│ ├── BoolConverter
│ ├── DateTimeConverter
│ ├── JsonConverter
│ └── EnumConverter
└── QueryBuilder\
├── QueryBuilder
├── QueryType
├── Dialect\
│ ├── DialectInterface
│ ├── DialectFactory
│ ├── AbstractStandardDialect
│ ├── AbstractOnConflictDialect
│ ├── MysqlDialect
│ ├── PostgresDialect
│ ├── SqliteDialect
│ └── GenericDialect
└── Compiler\
├── SqlCompilerInterface
├── SqlFragments
├── SelectCompiler
├── InsertCompiler
├── UpsertCompiler
├── UpdateCompiler
└── DeleteCompiler
Development
composer install vendor/bin/phpunit vendor/bin/phpstan analyse
License
MIT