C++ *

There are two articles on Russian, the author of which writes a virtual machine (interpreter) for executing a simple bytecode and then applies different optimizations to make this virtual machine faster. Besides that, there is a compiler of a simple C-like language into this bytecode. After reading this article and getting familiar with the compiler, I thought that it would be interesting to try writing a virtual machine for this language that would be able to apply JIT-compilation to this bytecode with the libjit library. This article describes the experience of doing that.

I found several articles online that describe the usage of this library, but those that I saw, describe the compilation of concrete programs with libjit, while I was interested in compiling arbitrary bytecode. For people interested in further reading, there is an official titorial, a series of articles and a series of comparisons (in Russian).

The implementation was done in C++ because we aren`t playing games here. All my code is in my repository. The "main" branch has just the interpreter of the PigletVM bytecode; "labels-with-fallbacks" has a partial JIT compilation implementation (that doesn`t support JUMP instructions), "full-jit" has fully working JIT-compilationl; "making-jit-code-faster" makes code generated by JIT work faster and "universal-base-vm*" branches merge the interpreter and JIT-compilation implementations, by implementing a base generalised executor, which can be used for different implementations of PigletVM (both the interpreter and libjit compilation)

According to the description,

Tree-sitter is a parser generator tool and an incremental parsing library. It can build a concrete syntax tree for a source file and efficiently update the syntax tree as the source file is edited.

But how does Tree-sitter handle languages that require a preprocessing stage?

Network programming in C++ can be challenging. But even a greater challenge is to find educational content that will arm you with the knowledge on how to apply your networking skills in real applications.

In this article you can learn the basics of socket communication and many ways how you can design your internal messaging protocols.

Asynchronous programming is commonly employed for efficient implementation of network interactions in C++. The essence of this approach lies in the fact that the results of socket read/write functions are not immediately available but become accessible after some time. This approach allows for loading the processor with useful work during the wait for data. Various implementations of this approach exist, such as callbacks, actors, future/promise, coroutines. In C++, these implementations are available as libraries from third-party developers or can be implemented independently.

Coroutines are the most challenging to implement as they require writing platform-dependent code. However, the recent version of the C++ 20 language standard introduces support for coroutines at the compiler and standard library levels. Coroutines are functions that can suspend their execution, preserving their state, and later return to that state to resume the function's work. The compiler automatically creates a checkpoint with the coroutine's state.

For a comprehensive understanding of C++ 20 coroutines, refer to this article. Below, we examine a code example using coroutines and describe important points applied during implementation.

Hash, salt, SHA-1, SHA-2, std::hash.. To a non-programming person that may come up as some kind of a recipe that just does not seem to add up. In a sense, this is indeed supposed to be a gibberish to any third party and a strong, helpful mechanism for us, programmers. 

At the start of writing this article, I had one clear idea to get across the table: to finally unveil this mystery of hashing in C++ for beginners. I, a beginner myself, also wanted to solidify my knowledge in this area; so let’s get started.

The point of this article is to explore Lambda functions, their dirrerences from regular functions and how they are implemented, based on C++, Python and Java programming languages.

Throughout this article I will be using godbolt.org to compile code and see machine code or byte code.

I always pay attention to assessing the complexity of programming in a particular language. Programming is indeed not an easy task and this is perceived as a fact and usually does not require any confirmation.

But the concept of “complexity” is akin to the term “heap”. For some, five coconuts is not so much, but for someone who ate one and “didn’t want any more,” this means that even one coconut will be too much for him.

The same goes for the complexity of programs. It seems that the constant increase in the complexity of programs is obvious to everyone and is observed in all areas of application of IT technologies, and programming languages themselves become more and more complex as they develop, but assessing “complexity” using numerical metrics is a problem. obviously a thankless task, but also “You can’t manage what you can’t measure...”

Typically, talk of “complexity” only implies value judgments without any numerical evaluation. And since I am personally interested in the issue of the complexity of programming languages, I decided to calculate the complexity of implementing the gcc compiler on some conditional “parrots”. What if we could see some patterns of difficulty changing over time?