cometkim

Maybe you can flat using dot-free style function calls too.

However, it unnecessarly add runtime overhead, pollute the callstack, and even not intuitive as direct-style control flows are.

cometkim

// Generated by ReScript, PLEASE EDIT WITH CARE

import * as Objects from "./Objects.res.mjs";
import * as Caml_obj from "rescript/lib/es6/caml_obj.js";
import * as Caml_int32 from "rescript/lib/es6/caml_int32.js";
import * as Environment from "./Environment.res.mjs";

var cTRUE = {
  TAG: "MBoolean",
  _0: {

cometkim

# Cargo binstall
# https://github.com/cargo-bins/cargo-binstall?tab=readme-ov-file#telemetry-collection
export BINSTALL_DISABLE_TELEMETRY=true

# Go Telemetry (not opt-in by default)
# https://go.dev/doc/telemetry
export GOTELEMETRY=off

# GatsbyJS

cometkim

export function encode(value) {
  const long = BigInt(value);
  const result = [];

  let shift = 0n;
  while (true) {
    const number = Number(BigInt.asUintN(7, long >> shift));
    if (number) {
      result.push(number);
      shift += 7n;

cometkim

보건의료빅데이터개방시스템(opendata.hira.or.kr) 공공데이터 상세 페이지 첨부파일 일괄 다운로드 스크립트

Usage

bun install

PAGE="https://opendata.hira.or.kr/op/opc/selectOpenData.do?sno=11925&publDataTpCd=&searchCnd=&searchWrd=%EC%A0%84%EA%B5%AD&pageIndex=1"
bun download-hira.ts "$PAGE"

cometkim

import * as http from 'node:http';
import { Readable } from 'node:stream';
import { setInterval } from 'node:timers/promises';

async function* chunksWhile(ms) {
  let count = 0;
  for await (const startTime of setInterval(1000, Date.now())) {
    yield `<div>Looking for your cat... (n=${++count})</div>`;
    if (Date.now() - startTime > ms) {
      return;

cometkim

JavaScript LRU library benchmark

There are too many LRU(Least Recently Used) implementations in JS.

I recommend to use [flru] which is the smallest one and fast enough. Unless you need more rich functionality.

However, it's performance is vary depend on the host environment. For example, flru loses on Bun.

If you need micro-optimization on it, measure it yourself in your environment.

cometkim

Call for study: Streaming runtime for GraphQL

Topics

1. Design format as a compact binary representation of the GraphQL trusted documents
   • The decoder should interface as a ReadableStream
   • It aligns the execution order of resolvers ahead-of-time
2. Pipe-able execute which compatible with exsting GraphQL resolvers
   • It should support e.g unpack(docs).pipe(executeStream)
3. Implement stream acceptance from the client

cometkim

#!/bin/bash

function healthcheck() {
  local domain=$1
  local total_tests=$2

  local total_failed_count=0
  local ip_list=()

  for ((i = 0; i < total_tests; i++)); do

cometkim

import { parseArgs } from 'node:util';
import { spawn } from 'node:child_process';
import { setTimeout } from 'node:timers/promises';

import prettyBytes from 'pretty-bytes';
import prettyMilliseconds from 'pretty-ms';

let { values, positionals } = parseArgs({
  args: process.argv.slice(2),
  allowPositionals: true,