loadtxt VS Ink

It really depends on what part of Numpy you're using. You can easily leave Numpy's text parsing in the dust. And if you're doing element-wise operations on arrays, you can easily see 2-3x improvement with just numba.

Amdahl's Law will catch up with you really fast as you add threads with this strategy, but it's simple and is amenable to formats where you may have a delimiter in the middle of a record. For situations where you need maximum scaling and don't have the possibility of delimiters scattered into records, you can use the strategy I used to implement a faster numpy.loadtxt: https://github.com/saethlin/loadtxt/blob/master/src/inner.rs#L84 The general idea is that you divide the file among thread boundaries by splitting it on byte boundaries, then seeking from that byte offset to the end of the next record. This gets you non-interleaved sections so there's no duplicate parsing.

I've always wanted to build a CLI tool, and when I realized that you can build one using React with Ink, I converted my Python script into a CLI tool.

⚠️ Git for Windows Terminal is currently not supported, and the tool is limited to ink. We will look for alternatives later. Please use CMD, Vscode terminal's Git... terminal

There is a good solution for some of those cases - ink. With ink, I can implement text-based UI with knowledge of React, which is neat but there are still some caveats for my usages:

Looks cool, right? Building a similar UI in the terminal without any library would be quite hard, though, thanks to Ink it's almost as easy as building any frontend UI with React.

ink?

I'm writing a CLI with ink. Writing async code is important as to not block the rendering and respond to user input. I have a few loading animations that update every 100ms. Synchronous operations can make the animation hang for >500ms, making the animation choppy.

You get the comfort of using react components instead of fighting with HTML tables to make your emails look nice. I think it's awesome! It's analog to what ink[0] does with CLI outputs. Sure, you could write fancy CLI outputs in bash, but ink takes the pain out of it and makes it easy.

[0] https://github.com/vadimdemedes/ink

Due to Node's asynchronous behavior, it makes Node great for long-running processes that make a lot of HTTP requests, database calls, and other async ops, like a web server or a REST API. However, if I am making a CLI tool for pretty much personal use only, with very minimal async operations, then blocking the event loop with a synchronous function that will resolve almost immediately will make no difference perceivable to a human brain or have any speed benefits that someone can actually observe (think `fs.readFileSync` or `require('dotenv') of 10 line config file, or a quick embedded db (sqlite) query with only ~100 records. I'm wondering what the best way to implement the database part of the app synchronous. I can read/write to JSON files but it would be tricky because the data is relational, and some complex joins and other data wrangling operations are required (complex to perform in JS but are easy to implement in a SQL statement). It's not important what the operations are, that's not the point of this post. This is mostly a personal project of interest: making this CLI tool completely avoiding any async operations/using no promises. I would like to use node tho, as I said this is just out of interest and I also want to experiment with several CLI libraries such as Ink or Cliffy.