Parsing JSON in C & C++: Singleton Tax
I’d argue that almost every open-source developer gets an extra spark of joy when someone reads the documentation and uses their tool in a way that goes beyond the classic 101 examples. It’s a rare treat even for popular projects like JSON parsers, but if you are building high-throughput software, such as analyzing millions of network packets per second, you’ll have to dig deeper.
The first thing I generally check in such libraries is the memory usage pattern and whether I can override the default memory allocator. It is the most common singleton of them all!
From my vantage point, singletons are a significant code smell. They might be convenient in small demos, but they come back to bite you in high-performance environments. In this article, we’ll dig into how some of the most popular JSON parsing libraries handle memory management, why singletons make me cringe, how hard it is to implement hybrid structures in C++ what C++ developers can learn from C, and what difference custom allocators can make — especially when you’re parsing short JSON objects on a tight budget.
Others also include (4.) Tencent’s RapidJSON and (5.) Stephen Berry’s Glaze, but we’ll focus on the first two for brevity. As the inputs get longer, they are generally dominated by string parsing, where SIMD plays a crucial role and has been broadly covered in this blog. Sadly, simdjson doesn’t currently support custom memory allocators, which is a big reason for many to explore the alternatives.
