At least for this specific example I don't know why I wouldn't use null instead of option and ?? As it's more clear what's happening as it's standard C#
Also in your example does the function to the right of | execute always?
At least for this specific example I don't know why I wouldn't use null instead of option and ?? As it's more clear what's happening as it's standard C#
Also in your example does the function to the right of | execute always?
The example is simplified, but I dislike returning null in my own code. The function will always execute, left or right doesn't matter it's mapped across in the ResultObject class.
The function must return an IResult, the ResultObject analyzes the IResult checking for IFail or IOk. If it's IOk the value of type T is retrieved from the Value property of the IOk object and returned, the Some property defaults to true. If the IResult is an IFail, Some is set to false, it copies the message from the IFail object into the ResultObject, and returns the value the was supplied to its constructor.
I'm just sharing something I find useful, and I hope I can make it useful for others as well. Thanks for the questions.
Nulls are famously called the billion dollar mistake, and for good reason.
Option types are the answer to that problem, because they make the optionality explicit and require one to handle it or propagate it.
That being said: as someone that does functional programming professionally, this looks kinda janky, to me. But the good news is that C# is actually adding support for discriminated unions finally (seriously, it's been waaaay too damn long): https://github.com/dotnet/csharplang/blob/18a527bcc1f0bdaf542d8b9a189c50068615b439/proposals%2FTypeUnions.md
With discriminated unions, you can finally comfortably work with Option/Result types natively.
Null pointers are one thing, C# nulls (with nullable reference types enabled) are another. They behave a lot like an Option monad with the caveat that the static analysis can technically be tricked by incorrect hints.
I very much disagree with this, Null Reference Exceptions have been a huge problem in c#. Nullable reference types are a partial fix, but the question of "how do I 'return' an error from a statically typed method" is not answered there.
Are there any updates on DUs in C#? I feel like the linked proposal was opened a decade ago by now.
Check the edit on the post, I thought I had linked to the GitHub page but I guess the image overrides that.
I'm not that familiar with newer c# code and only recently started with result pattern but tbh, I can't tell what is this code supposed to do. Does opt resolve to true or false in this case? Why do you want TestStringFail to always execute, and what should it return? Why is opt.None true when it was initialized with a valid string value, what does None even mean in this context?
The operator being applied to the ResultObject will always resolve to the Generic type that was specified as 'T in IResult. If the function is not successful the resolved value will be whatever value was supplied to the ResultObject constructor, the opt.None property will true and the opt.Some property will be false.
Check the edit on the post, I thought I had linked to the GitHub page but I guess the image overrides that.
You can't disagree with the fact that Nullable works a lot like an Option. Returning an error is not idiomatic C# code (which would be to throw an exception usually) but if you wanted that, you'd use a Result<T, TError> type or similar.
Seems an interesting solution. I'll try it when have some time
Thank you, I'd love to hear back from you.
A community about the C# programming language
Getting started
Useful resources
IDEs and code editors
Tools
Rules
Related communities