by Joche Ojeda | May 17, 2023 | AlchemyDotNet, C#
Alright, it’s time to start writing some code, but first, let’s decide how this project will be organized.
So far, the repository structure that I’ve found most appealing is the one I used for the SyncFramework (https://github.com/egarim/SyncFramework). Here is a representation with bullet points:
- Repo Folder: This is the parent folder that will contain all the code from our project. o Git Files: These are Ignore and attributes files.
- Clean.bat: This is a batch file that deletes all child BIN and OBJ folders to ensure our repository does not contain binary files (sometimes the ‘clean’ command from Visual Studio does not clear the outputs completely).
- CHANGES.MD: This is a Github markdown file that contains the history of changes in the code.
- README.MD: This is the landing page of the current repository where we can write some basic instructions and other important information about the project.
- src (folder): This is where the actual code resides.
Before I conclude this post, I want to discuss versioning. The main idea here is this: the project AlchemyDotNet.Core will start at version 1.0.0, and the version will change when there is a fix or an update to the core specification. We will only move to a new version like 2.0.0 when the specification introduces breaking changes. This means that we will be on version 1 for a very long time.
Link to the repo
https://github.com/egarim/Alchemy.Net
Previous posts
Alchemy Framework: 1 – Creating a framework for import data
by Joche Ojeda | May 15, 2023 | Uncategorized
Last Friday, I received a message from a dear friend and colleague, Pedro Hernandez. He asked me if I had the latest compiled version of the XPO import framework we created in our office. As it turned out, I did not have it readily available and had to search extensively for it.
While conducting this search across my computers, repositories, and virtual machines, I was inspired to create another import framework — yes, another piece of code to maintain.
Most of the time, my research projects begin in the same manner, typically after a conversation with my close friend Jose Javier Columbie. As always, he would say something like this: “Jose, do you think this can be possible? Why don’t you give it a try? If we succeed, that will be el batazo (like hitting a home run).”
In this specific case, that conversation didn’t happen. However, I could hear his words echoing in my mind.
Furthermore, I want this project to be community-driven, not just a technical experiment I have to maintain alone. I truly believe in the power of a community.
First Step
So, let’s begin. I’ll start with my favorite part: naming the project. I spent all weekend pondering this, attempting to condense the concept and associate it with a literary term or a Latin word (these are my preferred methods for naming a project).
Let’s define what the ultimate goal is. In an import process, the aim is to take information from a source ‘A’, translate or transform the information, and then store it in a target ‘B’.
Growing up, I was an avid reader – and I mean, I read a lot. (Now, I’ve switched to audiobooks.) Therefore, after defining what this project is about, naming it became incredibly easy.
noun
noun: alchemy
- the medieval forerunner of chemistry, concerned with the transmutation of matter, in particular with attempts to convert base metals into gold or find a universal elixir.
“occult sciences, such as alchemy and astrology”
Origin
late Middle English: via Old French and medieval Latin from Arabic al-kīmiyā’, from al ‘the’ + kīmiyā’ (from Greek khēmia, khēmeia ‘art of transmuting metals’).
The alchemy Framework
Alchemy is a framework created for DotNet, designed to import data from a data source to a data target. These sources and targets can be anything from a text file, CSV file, a database, ORMs, and so on.
The framework consists of a set of contracts, interfaces, and base classes. When implemented, these allow you to import and transform data between various sources and targets.
The requirement in a few words
As stated in the framework’s description, the requirement is only to define the contracts that represent the sources and targets, as well as a job configuration that describes how the information flows from one source to a target. The concrete implementations are not important at this point and should be discussed individually for each case.
The design patterns.
For this project, we will use the SOLID design principles and dependency injection. This will enable us to easily replace small functionalities, allowing us to mix and match different implementations depending on the data source and data target.
by Joche Ojeda | May 5, 2023 | Application Framework, XAF, XPO, XPO Database Replication
I will explain what XAF is just for the sake of the consistency of this article, XAF is a low code application framework for line of business applications that runs on NET framework (windows forms and web forms) and in dotnet (windows forms, Blazor and Web API)
XAF is laser focus on productivity, DevExpress team has created several modules that encapsulate design patterns and common tasks needed on L.O.B apps.
The starting point in XAF is to provide a domain model using an ORMs like XPO or Entity framework and then XAF will create an application for you using the target platform of choice.
It’s a common misunderstanding that you need to use and ORM in order to provide a domain model to XAF
DevExpress team has created ObjectSpace abstraction so XAF can be extended to use different data access technologies ( you can read more about it here https://docs.devexpress.com/eXpressAppFramework/DevExpress.ExpressApp.BaseObjectSpace)
Out of the box XAF provide 3 branches of object spaces as show is the graph below.
XPObjectSpace: this is the object space that allows you to use XPO as a data access technology.
EfCoreObjectSpace: this is the object space that allows you to use Microsoft Entity Framework as a data access technology.
NonPersistenObjectSpace: this object space is interesting as it provides the domain model needed for XAF to generate the views and interact with the data is not attached to an ORM technology so it’s up to us to provide the data, also this type of object space can be used in combination with XPObjectSpace and EfCoreObjectSpace
When querying external data sources, you also need to solve the problem of filtering and sorting data in order to provide a full solution, for that reason DevExpress team provide us with the DynamicCollection class, that is a proxy collection that allows you to filter and sort an original collection without changing it.
Now that we know the parts involved in presenting data in a XAF application, we can define the required flow.
[DefaultClassOptions]
[DefaultProperty(nameof(Article.Title))]
[DevExpress.ExpressApp.ConditionalAppearance.Appearance("", Enabled = false, TargetItems = "*")]
[DevExpress.ExpressApp.DC.DomainComponent]
public class Article : NonPersistentObjectBase {
internal Article() { }
public override void Setup(XafApplication application) {
base.Setup(application);
// Manage various aspects of the application UI and behavior at the module level.
application.SetupComplete += Application_SetupComplete;
}
- Wire the application object space created event.
private void Application_SetupComplete(object sender, EventArgs e) {
Application.ObjectSpaceCreated += Application_ObjectSpaceCreated;
}
private void Application_ObjectSpaceCreated(object sender, ObjectSpaceCreatedEventArgs e) {
var npos = e.ObjectSpace as NonPersistentObjectSpace;
if (npos != null) {
new ArticleAdapter(npos);
new ContactAdapter(npos);
}
}
public ArticleAdapter(NonPersistentObjectSpace npos) {
this.objectSpace = npos;
objectSpace.ObjectsGetting += ObjectSpace_ObjectsGetting;
}
private void ObjectSpace_ObjectsGetting(object sender, ObjectsGettingEventArgs e) {
if(e.ObjectType == typeof(Article)) {
var collection = new DynamicCollection(objectSpace, e.ObjectType, e.Criteria, e.Sorting, e.InTransaction);
collection.FetchObjects += DynamicCollection_FetchObjects;
e.Objects = collection;
}
}
private void DynamicCollection_FetchObjects(object sender, FetchObjectsEventArgs e) {
if(e.ObjectType == typeof(Article)) {
e.Objects = articles;
e.ShapeData = true;
}
}
Full source code here
In conclusion the ObjectSpace abstraction ensures that different data access technologies can be employed, while the DynamicCollection class allows for seamless filtering and sorting of data from external sources. By following the outlined steps, developers can create robust, adaptable, and efficient applications with XAF, ultimately saving time and effort while maximizing application performance.
