My #bubblebobble cereal bowl is done. It looks all shiny and ready to be used!
Took me multiple hours to paint, but I'm very happy with the result. 😍
My #bubblebobble cereal bowl is done. It looks all shiny and ready to be used!
Took me multiple hours to paint, but I'm very happy with the result. 😍
Currently my custom language compiler is doing one-pass when compiling.
So it reads a source file, and it generates the abstract syntax tree as it goes. This is quite efficient, but it feels like it's getting hard to maintain.
I wonder if doing a multi-pass will make it easier to make the compiler more modular and easier to extend.
I want to add #multithreading in my custom language compiler.
When run it currently compiles all source files one-by-one, but I feel like I could run 3 threads and each can compile a third of the source files, similar to the "-j" argument in #make .
I don't want to deal with thread-related errors though ..
I ended up putting a flag in the class table to keep track of functions being virtual (so they can be used in polymorphism) and them being methods (so the object they modify is automatically passed).
My compiler can now convert something like "myvar.myfunc()" to "VarClass_myfunc(&myvar)".
The fun part of making your own programming language is you get to design all the features that you like.
No one told me I also have to implement them 
I can feel the urge to get a #linux phone returning... My last purchase was a PinePhone some years ago.
Curious if there are more mature options nowadays.
My programming language can do either compile-time or run-time function resolution, similar to #C++ 's "virtual" keyword.
I've been doing some rewriting and I can't figure out what's the best place to save that information. The class table sounds like the most obvious choice, but I can't help but feel like it belongs somewhere else.
I just found out that the parser for my programming language is considered a "recursive descent parser".
Of course I found that out while struggling to implement a feature, which apparently is harder to do in these kind of parsers 😅
I've been rewriting my abstract syntax tree implementation, for my programming language.
The goal is to isolate each node, so it is easier to work on, but also for the whole tree to work as the intermediate representation, so each node can easily be translated into the final result.
The idea feels good in my head, only one way to find out if it works 😅
I had a delicious meal with my partner for Valentine's day ❤️
Delicious rose latte and yummy dessert 🤤