Meadow.Framework 1.2.3

Meadow

M<nothing>anaged/<nothing>E<nothing>nhanced<nothing>Ado<nothing>W<nothing>rapper

This library is designed to help using ado.net easier and with some good practices consistently.

Features

• It Supports MS-SqlServer, My Sql, SQLite and Postgre database frameworks.
• The application would create/buildup and update database by itself after deployments.
• Your data access logic will be fully implemented in SQL language, therefore:
  • You would have access to what ever feature your database framework provides with no limitations.
  • Non of your data-access business would be transported to server as pure sql text
• Like Migrations in EntityFramework, each change in the database would be maintained as the project source code.
• There will be no object-tracing and each data-access operation closes the connection which helps reducing the chance of concurrency and improves performance in load peaks.
• The conversion of data to/from POCOs are internally covered so working with data is easy.

The Practice

Meadow is designed to help implementing a practice as follows:

• Design and Maintain the database, step by step, by creating build-up scripts (They would be treated like source code)
• For each data access operation, Create a Stored Procedure in a build-up script.
• For each data access operation, drive a well named request class, from MeadowRequest class.
• Perform the request and receive the results (if any) whenever needed, using MeadowEngine and an instance of corresponding request.

The MeadowEngine can take some configurations. It also performs some useful operations for you like creating/deleting and building up the database.

• NOTE: The best practice would be to use Meadow behind your Repository Pattern implementation.

In summery

1. You would use meadow for retrieving/manipulating data through the MeadowRequests
2. You would use meadow for creation/deletion or building up your database (usually at the startup section of the application) via MeadowEngine

How to Use

1. Prepare your connection string, regarding your database-framework of choice. (Ms-Sql-server, MySql, SQLite)
2. Setup "Build-up scripts"
3. Add Meadow library to your project. (You can install the NuGet package)
4. Create a MeadowRequest and it's required models To use each of your procedures in your c# code.
5. Use meadow!

Build-up Scripts

Create a directory in your source codes for your build-up scripts. Directory name does not matter. Make sure this directory will be copied alongside your binaries when you build/publish your project.

ex: your-project.csproj

  <ItemGroup>
    <Content Include="Scripts\**">
      <CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
    </Content>
  </ItemGroup>

For each set of changes during the process of evolving your database, Create a script inside the build-up scripts directory. It's best to create script files, regarding you data structure changes, and in a way that i's easy to read and find where to look for scripts related to a specific part of the data base. It this sense, it would be similar to migrations in ORMs. Also try to avoid manipulating last scripts, and always create new scripts for manipulation of already existing db-objects.

• ⚠️ NOTE: MANIPULATING PREVIOUSLY ADDED SCRIPTS, WOULD REQUIRE THE DATABASE TO BE RE-CREATED WHICH IS A BIG COST FOR PRODUCTION, VERSA KEEPING THE CHANGES IN A ALWAYS FORWARDING MANNER MAKES IT SMOOTHER AND SAFER FOR PRODUCT UPDATE WITHOUT DOWN-TIME.

Each build-up script file should has .sql extension and Start with 4 digit number. The numbers would represent the files order. smaller numbers will be applied before larger numbers. (ie.: 0000_initial-database.sql or 0001-AddUserInformationTables.sql)

^ [Back to How To Use]

Add Meadow library

Meadow libraries are available on Nuget. To use meadow, you would add

1. Add Meadow.Framework.

   PackageReference	<PackageReference Include="Meadow.Framework" Version="1.0.1" />
   PackageManager	Install-Package Meadow.Framework -Version 1.0.1
   Dotnet CLI	dotnet add package Meadow.Framework --version 1.0.1

2. Add Meadow.<Db-Framework>. For each database-framework supported by Meadow, you would add it's own db-framework adapter:

   1. SqlServer

      PackageReference	<PackageReference Include="Meadow.SqlServer" Version="1.0.0" />
      PackageManager	dotnet add package Meadow.SqlServer --version 1.0.0
      Dotnet CLI	dotnet add package Meadow.SqlServer --version 1.0.0

   2. SQLite

      PackageReference	<PackageReference Include="Meadow.SQLite" Version="1.0.0" />
      PackageManager	Install-Package Meadow.SQLite -Version 1.0.0
      Dotnet CLI	dotnet add package Meadow.SQLite --version 1.0.0

   3. MySql

      PackageReference	<PackageReference Include="Meadow.MySql" Version="1.0.0" />
      PackageManager	Install-Package Meadow.MySql -Version 1.0.0
      Dotnet CLI	dotnet add package Meadow.MySql --version 1.0.0

   4. Postgre

      Postgre Related Considerations	Postgre Readme
      PackageReference	<PackageReference Include="Meadow.Postgre" Version="1.0.0" />
      PackageManager	Install-Package Meadow.Postgre -Version 1.0.0
      Dotnet CLI	dotnet add package Meadow.Postgre --version 1.0.0

• ⚠️ NOTE: VERSIONS APPEARING IN THIS SECTION ARE NOT NECESSARILY LATEST. PLEASE CHECK OUT THE NUGET PAGE FOR EACH TO SEE THE LATEST VERSION

^ [Back to How To Use]

Creating MeadowRequests

For creating a request, you can drive from MeadowRequest class. Its a better practice to keep your request classes under a grouped and managed directory/namespace. You can use Conventions to write minimum code.

By Convention, a MeadowRequest named AbcRequest, would be resolved for to execute a procedure named: spAbc.

If you prefer not to rely on conventions, you can provide procedure name manually by overriding the the property RequestText in your meadow request.

Meadow also contains several pre-built requests for common operations that you can drive from:

1. InsertSpRequest<TModel> Inserts an Entity Of type TModel, Into the database. It expects the corresponding procedure (which you would write in build-up scripts) to (a) have one input parameter corresponding to each effectively-primitive [^1] property of TModel. (Except for the Id Parameter)
   (b) return newly inserted model.
2. UpdateSpRequest<TModel>
   Updates an Entity Of Type TEntity then returns the updated value. Therefore it also
   expects your stored-procedure, to (a) have one input parameter corresponding to each effectively-primitive property of TModel. (Except for the Id Parameter)
   (b) return newly inserted model.
3. ReadAllSpRequest<TModel>
   Reads all records of type TModel. If you pass 'True' for the constructor argument: fullTree, then this request expects your procedure to return all fields from all other tables (joined), so it can glue all together and construct an instance of TModel and all it's properties. If you pass 'False' for full-tree argument, then a simple procedure like select * from <table-name> ... would be all it needs.
4. ReadByIdSpRequest<TModel, TId>
   Reads the Entity identified by the given Id. This Request and it's descendants, would have an Id property of type: TId. You can set this property before performing the request. This request, expects its Corresponding procedure to have one argument named after the Identity Field of your Model. the full-tree argument in this request is the same as ReadAllSpRequest.
5. DeleteByIdSpRequest<TModel,TId>
   This request is for deleting a record by its Id. Like ReadById, It uses the Id property. It returns the success result of the deletion as DeletionResult. So it expects your procedure to return a true/false value with the field-name of Success.
6. ByIdRequestBase
   This request is the base class for ReadByIdSpRequest and DeleteByIdSpRequest. It can be used for creating any other ById operations of interest.
7. SaveSpRequest<TModel> This Request to have a data inserted when it's not present in the database, and updated if it's already there. The expected input parameters are the same as Insert or Update. Also no matter the insertion happened or an update took place, the procedure should return the inserted/updated model. More On Requests

Output/Input

When you drive MeadowRequest<TIn,TOut>, TIn would be the type of the data you send towards the database when you perform that request. in example when you are inserting a data. The same way, TOut would be the type of data returned from the database. In a case that you do not send any data towards the database, you should use the type MeadowVoid for TIn. Also for those requests that would not return any data, you will use this class.

Constructor

Another point to consider is MeadowRequest's Constructor. It takes a boolean argument to determine if this request is obligated to read returned data.

Name Conventions

In a case that you needed to know, what Db Object Names Meadow would prefer to be used for a specific Entity, you can instantiate NameConvention<TEntity> class. This class will provide predefined names for common procedures and tables regarding given type (TEntity).

^ [Back to How To Use]

Using Meadow In Your Project

Create and configure MeadowEngine by passing a MeadowConfiguration object to its constructor. A MeadowConfiguration object has a ConnectionString property and a BuildupScriptDirectory property which is where you save build-up scripts. Then all you need to do is to instantiate an object of your meadow request, and call MeadowEngine.PerformRequest(request). Meadow creates and disposes resources per request, therefore it's not necessary to keep the engine object. You can keep the engine object or instantiate it per request.

You can also instantiate the MeadowEngine also by passing an object of Type IMeadowConfigurationProvider. This would be helpful when you want to use different configurations controlled by your DI.

^ [Back to How To Use]

Examples

Meadow can be used to connect to Ms-SqlServer, My-Sql, SQLite and Postgre. For each of these databases, these is an examples in the project. You can see that using Meadow is the same in each example, but the only difference is in the actual sql scripts written for each database.

ℹ️ NOTE: SQLITE, DOES NOT SUPPORT STORED PROCEDURES, AND MEADOW IS BASED ON STORED PROCEDURES. SO MEADOW DOES STORE YOUR PROCEDURES SEAMLESSLY.

When The Darkness Caresses The Meadows...[^2] (Bugs and Issues)

Meadow For Ms-SqlServer is currently being tested in my other projects, but for MySql, SQLite and Postgre, It still has not being tested in practice. So Please consider dropping a mail or comment or opening a github issue if you faced a bug or an issue. Very Thanks.

Developing a new Database system Adapter

Meadow can recognize any database system through it's abstractions. The main enttry for any database system is the interface IMeadowDataAccessCoreProvider which connects your database system for the meadow engine and can make use of it in higher levels of your code seamlessly.

IMeadowDataAccessCoreProvider provides your implementation of IMeadowDataAccessCore. This interface defines a database system via several functions to be implemented. The Main method to be implemented, would be IMeadowSataAccessCore.Perform<.>(.). Other methods are for creating and also providing the basic functionalities meadow needs for it's internal buildup process.

In most cases, you might not need to implement IMeadowDataAccessCore from scratch. Instead, you can extend the MeadowDataAccessCoreBase class. This class handles a portion of implementation and breaks down the remaining into the implementation of:

• IStandardDataStorageAdapter
• IStorageCommunication

You still do not have to implement IStandardDataStorageAdapter from the scratch in all cases. For sql database systems, you can extend SqlDataStorageAdapterBase.

Easier Implementation For Ado wrapped database systems

If The database system of your interest, does already have an ADO implementation, you can write your adapter, just by extending the class AdoDataAccessCoreBase. and implement its methods. You also would have to provide an implementation of IDbTypeNameMapper which is a dictionary, mapping the c# System types to proper data types known in your database system.

Contact

Please feel free to contact me about meadow, or any related questions.

acidmanic.moayedi@gmail.com

https://www.linkedin.com/in/mani-moayedi

https://www.instagram.com/acidmanix/

Thanks And Regards. Mani.

.

[^1]: For Meadow, Effectively Primitive mean any type that is able to be inserted directly into the database. It involves all primitives, plus the String, DateTime and etc.... [^2]: IGNEA, Jahi.