swc VS kataw

First, we switched the default compiler for new projects from Babel to SWC (Speedy Web Compiler). SWC is dramatically faster than Babel and requires zero configuration. We’ll continue to support Babel in any project currently using it.

SWC

As the reference explains “**SWC** (Speedy Web Compiler) is an extensible Rust-based platform that can be used for both compilation and bundling. Using SWC with Nest CLI is a great and simple way to significantly speed up your development process.”

This is specifically about breaking the myth that performing expensive self-contained operations (e.g, parsing GraphQL) in a native extension (C, Rust, etc.) is always faster than the interpreted language.

The JS ecosystem has the same problem, people think rewriting everything in Rust will be a magic fix. In practice, there's always the problem highlighted in the post (transitioning is expensive, causes optimization bailouts), as well as the cost of actually getting the results back into Node-land. This is why SWC abandoned the JS API for writing plugins - constantly bouncing back and forth while traversing AST nodes was even slower than Babel (e.g https://github.com/swc-project/swc/issues/1392#issuecomment-...)

Why Speedy Web Compiler ?

FTA is a TypeScript static analysis tool built on the speedy foundations of swc. FTA is fast; capable of analyzing more than 150 files per second on typical hardware, it offers a powerful addition to your code quality toolkit.

Very cool! I'm curious, is this intended for dev tooling?

For example, I could see this (or something similar) being useful as the engine for a typescript language server that would be faster than the standard one

But if it's not aimed at 1:1 with tsc, would it be intended more for something like swc[1]?

Or what would you expect people to use this for, besides just being a cool project to learn from?

[1] https://github.com/swc-project/swc

This is... good news, but I still cannot fathom using the default Typescript compiler for regular development. Seriously, leave the type-checking to your IDE and CICD chain, and switch to using tsx (https://www.npmjs.com/package/tsx) or swc (https://swc.rs/) and you will _immediately_ notice the difference in speed and productivity.

It then builds on the assumption that all of Javascript performance scales linearly based on the results of that fishy benchmark. Never mind that in C you can stack allocate, or use arena allocators, or use a million of other tricks, where in JS, once you write {}, you got a heap allocation that you don't have any control over (and if you think this doesn't matter, this is one of the reasons why esbuild is so much faster than babel, or heck even kataw, the fastest JS parsing toolkit I'm aware of).

esbuild is the current darling leading the pack, but there are also various other projects in the space (swc[0] is written in Rust, fjb[1] is written in C, bun[2] is written in zig, leveraging JavascriptCore's engine).

The most significant performance-oriented effort in this space still leveraging JS that I know of is kataw[3], and while that's quite fast compared to babel, it's still within an order of magnitude from babel. Kataw itself is a CST-based implementation that was created to outperform seafox (a AST-based parser by the same developer).

Babel gained popularity due to the crazy amount of churn in grammar over the past few years, but more and more I think the dust is settling, and flexibility is no longer the name of the game, making an AST-based implementation less appealing. The Rome team must be feeling the heat if the data structure design choices are being informed by performance. I highly doubt someone will be able to compete in performance using a JS implementation in today's landscape.

[0] https://github.com/swc-project/swc

[1] https://github.com/sebbekarlsson/fjb

[2] https://bun.sh/

[3] https://github.com/kataw/kataw

Kataw is a *work in progress* javascript toolchain with high focus on performance, and it's main goal is to unify functionality that has previously been separate tools. It's core is an ECMAScript friendly CST that allows you to parse `ECMAScript® 2022 (ECMA-262 12th Edition) language specification`. Each CST node contains several properties, and the `flags` property contains the CST info. The CST info can be extracted from the CST node through public API methods. Click [here](https://github.com/kataw/kataw/tree/main/src/parser#public-api-methods-to-extract-info-from-cst-nodes) for a complete list over all the public API methods. Many of these APIs have the advantage that they allow you to "retrieve" info that is not otherwise available with a standard AST parser. One example is that you only need to use `kataw.isStatementNode` to find out if the current CST node is a statement node. With an AST parser you must use a `switch statement` with 60 `switch cases`. Another example is how easy it is to find out if a keyword node contains an `escaped keyword`. You only need to use `kataw.hasUnicodeEscape`. You can narrow this down further if you use `kataw.isChildLess`. A keyword doesn't have any child nodes. A third benefit with this CST parser is that it is running in `recovery mode` by `default` and can be used in any editor. A build-in diagnostic system reports diagnostics if an `error handler` have been used. The diagnostics are dynamic. It means all the diagnostics are informative, and they will will change based on the context you are parsing in. The diagnostics have been designed like this so you can quickly understand what the problem is and correct it. These features used together gives you more options to adjust, modify and customize the CST tree compared to a regular AST parser and you can also write fewer code lines and at the same time experience insane performance.