Category: pet projects
-
DocumentorDatabaseExtensionsAspire–part 3–usage
The usage is pretty simple : add the package https://www.nuget.org/Packages/DocumentorDatabaseExtensionsAspire to the Aspire host project. Add an SqlServer database and write //accepts also relative paths var res = db.AddDocumentationOnFolder(@”D:\documentation”); That will be all ! And here it the Aspire Dashboard with documentation for Pubs database And the diagram Here you will find a video about…
-
DocumentorDatabaseExtensionsAspire–part 2–implementation
The Aspire project as already a database resource – and a executing one ( sqlpad or dbgate or adminer ) So the answer to the several problems to be solved: 1. How to obtain the tables definition ? We can obtain the definition by handling ourselves the tables- but why do not profit from the…
-
DocumentorDatabaseExtensionsAspire–part 1–idea
The Aspire project as already a database resource – and a executing one ( sqlpad or dbgate or adminer ) What if we can generate documentation for the database ? There are several problems to be solved: 1. How to obtain the tables definition ? 2. How to generate documents for the site ? 3.…
-
SqlExtensionsAspire–drop create database
Another improvement for https://www.nuget.org/account/Packages/SqlExtensionsAspire : Drop Database , Create the database, run init scripts . Demo with Video : https://ignatandrei.github.io/aspireExtensions/images/SqlServerExtensions/video-Recreate_Database_With_Scripts-20251102225854.mp4 The code is pretty simple: executing drop create and initialization
-
Conclusions: Translating fp-filters from TypeScript to C# with AI Assistance
1. AI Accelerates Porting, but Human Oversight Is Essential AI tools (GPT-4, GPT-5, Claude, Copilot, CodeRabbit) greatly speed up translation and refactoring, but human review is crucial for correctness, especially for edge cases and build issues. 2. Iterative Improvements Work Best Incremental changes—such as improving code coverage, refactoring namespaces, and enhancing documentation—yield better results than…
-
Deploy to nuget
GPT 4.1 was incredibly usefull. But having the 2 AI reviewers ( Copilot and Code Rabbit )were best : improving the .csproj improving the github action to publish with the name of the build and many others See https://github.com/ignatandrei/Filters/pull/11 Discussion Human Please add readme and other things to deploy to nuget the Filters csproj from…
-
Cosmetic fixes
Before deploy to Nuget I think how the programmer will use it . And the problem is the generated code – it repeats the name of the class twice .So AI will do some cosmetic fixes : Remove from the namespace the name of the class namespace FpFilters.NumberFilters { public static class NumberFilters modified to…
-
Interlude–AI Code Review
CodeRabbit ( https://www.coderabbit.ai/ ) provides a free tier for OSS to do code reviews Also GitHub Copilot does that. So I tried both . And it is nice to see their suggestions ( for example, see https://github.com/ignatandrei/Filters/pull/1 and the other PR ) and then telling to another AI to do the changes It was pretty…
-
Adding Linq functions
LINQ is a powerful feature in C#, and this post showcases its integration into filter tests. By adding LINQ-friendly overloads and corresponding BDD tests, the filters became more versatile and idiomatic for C# developers. I give the general instruction that, << for any function with 2 parameters, add a function that can be used in…
-
Pareto principle for CodeCoverage
The repeated discussion with AI about each filter ( string, boolean , number and so on) bring code coverage is 87.28 %. I do aim to reach 100% . Why ? Because I haven’t created or teste the project manually , but it is MY project under MY name . Resume of how I have…