TinyORM

TinyORM is a modern ORM library that makes interacting with a database extremely simple.

The code is written in the modern c++20 way and is heavily tested with 3378 unit and functional tests. Almost all the query builder methods are unit tested. The TinyORM's query builder code and the code which is responsible for obtaining relationships, is tested by functional tests against all supported databases. The code coverage is good enough to guarantee API and behavior compatibility.

Donations 🪙

Bitcoin address  1NiF2cTvYxUj8FTZJnGn1ycN4yisWfo1vJ
PayPal                 https://paypal.me/silverzachara

Documentation

The whole library is documented as markdown documents:

• Prologue
• Dependencies
• Supported Compilers
• Database
  • Getting Started
  • Query Builder
  • Migrations
  • Seeding
• TinyORM
  • Getting Started
  • Relationships
  • Collections
  • Casts
  • Serialization
• TinyDrivers
  • Getting Started
• Building
  • TinyORM
  • Hello world
  • Migrations
• Stability
• Features Summary
• Donations

Features Summary

• Summary List
• Showcase Images

Summary List

The following list summarizes the main features of the TinyDrivers and TinyMySql libraries. 🆕

• can be used instead of the QtSql module
  • can be swapped at compile time 🤯
  • has 1:1 API as the QtSql module 😮
  • swapping is controlled by the qmake and CMake build system options
  • was designed to drop the QtSql dependency while maintaining backward compatibility and without the need for any code changes after the swap
• currently, only the MySQL database driver is supported and finished 😞
  • this is only true for our TinyDrivers library ❗
  • all database drivers described in the Database documentation are supported when linking against the QtSql module
• both, normal and prepared statements are supported
• TLS/SSL connections using MYSQL_OPT_SSL_MODE (verify_ca, verify_identity) 🔥
• setting many other MySQL connection options (see mysqldriver_p.cpp)
• building and linking against the MariaDB Connector/C 🕺
• transactions
• re-using the current SqlQuery instance to re-execute the same or another SQL query
• detaching from the result set (related to freeing/releasing memory)
• query size, number of affected rows, last inserted ID, testing isNull(), ...
• TinyDrivers library supports both build systems qmake and also CMake
• all 3378 unit tests passed 😮
• strictly using smart pointers (no new keyword in the whole TinyDrivers code base 😎)
• clear code 🤔

The following list quickly summarizes all the TinyORM features.

• simple database connections management 🧬
  • database manager that helps with the database connections management
  • Orm::DB facade class for nicer and shorter syntax
  • MySQL, MariaDB, SQLite, and PostgreSQL support for all features 💎
  • multi-threading support 👀
  • SSL connections support 🔒
• impressive query builder 🔧
  • allows passing sub-queries and raw expressions practically everywhere, to column names, values, and to every SQL clause as select, where, joins, group by, having, order by 🔥
  • a logical grouping that offers to wrap logical groups in parenthesis
  • chunked results for lower memory footprint ✨
  • raw methods for all SQL clauses
  • all join types (left, right, cross, inner) and also join where clause support 🫤
  • aggregate methods min, max, sum, increment, decrement, ...
  • whereExists and exists methods for an existence queries
  • transactions and pessimistic locking 🔒
  • of course, insert, update, and delete SQL clauses support
  • correct QDateTime time zone using the qt_timezone connection configuration option 📅 (returned QDateTime instances will have the correct time zone, and also works for an ORM)
    • this feature allows you to set up the database server time zone to the UTC and all returned QDateTime instances will have the correct UTC time zone
• clever ORM with all relation types support 🎉
  • one-to-one, one-to-many, and many-to-many relation types (also inverse relationships) 🧨
  • eager and lazy loading with custom select and constraints 🚀
  • fluent ModelsCollection that expose a variety of map / reduce operations that may be chained using an intuitive interface ✨
  • all query builder methods are proxied from the model instances and also from the relation instances back to the query builder 🤯 (everything that can be called on the query builder can also be called on the model and relation instances)
  • clean active record pattern
  • advanced features like timestamps, touching parent timestamps, soft deleting, default models, default model attributes, and attributes casting 🤓
  • querying relationships existence/absence using the has, whereHas, and hasNested methods (using dot notation for selecting nested relationships users.posts.comments)
  • serializing models and collection of models including all nested relations to JSON and converting to vectors and maps 🪡
    • supports controlling a custom date format during serialization
    • supports hiding and appending attributes
    • supports defining Accessors (transform attribute value when it is accessed; eg. building value from multiple attributes)
• compiled database migrations and seeders 🕺
  • create, update, drop, and rename database tables
  • create, update, drop, and rename table columns
  • extensive schema builder that allows creating of all possible column types
  • terser syntax for creating foreign keys and foreign key constraints
  • supports creating and dropping column indexes (primary, unique, fulltext, spatial)
• the tom console application with tab completion for all shells (pwsh, bash, zsh) 🥳
  • tab completion is as accurate as possible
    • completes commands, long and short parameters, argument values
    • even connection names can be completed
  • migrate, fresh, refresh, rollback, and reset migrate operations
  • install and uninstall migration repository (table that tracks migrations and batches; table name can be customized)
  • seed and wipe database operations
  • scaffolding of models, migrations, and seeders
  • impressive models scaffolding, every feature that is supported by models can be generated using the tom make:model cli command
• others 🛝
  • a huge amount of code is unit tested, currently 3378 unit tests 🤯
  • C++20 only, with all the latest features used like concepts/constraints, ranges, smart pointers (no new keyword in the whole code base 😎), folding expressions
  • qmake and CMake build systems support
    • CMake FetchContent module support 🤙
    • LTO support (backed up by CI)
  • vcpkg support (official tinyorm vcpkg port)
    • official port doesn't allow to have alternative implementations so it doesn't contain TinyDrivers
    • unofficial tinyorm vcpkg port with TinyDrivers support (currently MySQL only)
  • it's really fast, you can run 1000 complex queries in 500ms (heavily DB dependent, the PostgreSQL is by far the fastest) ⌚
  • extensive documentation 📃
  • ...

:::info See the TinyDrivers Features summary. :::

Showcase Images

Tom console application

Passed all unit tests 🥳

TinyOrmPlayground

The TinyOrmPlayground project is my personal project where I have tested all the TinyORM database queries in the early development phases, currently, it executes ~1600 database queries across the whole TinyORM framework. Every query has a nice title header, is logged to the console, and is counted and measured (elapsed time). Every query also runs on all Supported Databases.

The TinyOrmPlayground project can be compiled in a single-threaded or multi-threaded mode. In the multi-threaded mode, every database connection runs in its own thread. The connection summary is logged at the end of every database connection and the application summary is logged before an exit. Whole TinyOrmPlayground application is configurable through the src/configuration.hpp class.

TinyOrmPlayground single-threaded

TinyOrmPlayground multi-threaded