Maciej Biłas maciej

Inverted worker pool

Inverted worker thread pool pattern from Rethinking Classical Concurrency Patterns

var sem = make(chan struct{}, limit)

func start() {
    for _, task := range hugeSlice {
 sem &lt;- struct{}{}

Go sets the default buffer size of na new bufio.Scanner to 4096 bytes. I've scanned a ~155MB file with an average line length of 62 bytes with smaller and larger buffer sizes. Here are the results running on a 2019 x86 MacBookPro:

file size: 155605069
lines: 2501619
avg line length: 62.20
  1024: 182.103886ms
 2048: 116.351501ms

Redis HLL has a 16 bytes header:

8 bytes are used to cache the set cardinality
3 are not used
4 are used to store the magic "HYLL"
1 is used to represent the type of the set (sparse or dense)

struct hllhdr {
 char magic[4]; /* "HYLL" */

	#!/usr/bin/env bash
	set -euo pipefail

	# Step 1: Extract the last element of the module name from go.mod
	if [[ ! -f "go.mod" ]]; then
	echo "Error: go.mod not found in current directory." >&2
	exit 1
	fi

	module_name=$(awk '/^module / {print $2}' go.mod)

	#!/usr/bin/python

	# This is used for calculating of birthday paradox for large values
	# We're using approximation explained here: http://preshing.com/20110504/hash-collision-probabilities/
	# Beware that approximation isn't very precise at smaller params N and K, but gets more precise with large numbers

	# see https://docs.python.org/3/library/decimal.html#module-decimal
	from decimal import *

	# setting decimal precision

	package main

	import (
	"sync"
	"sync/atomic"
	"testing"
	)

	func BenchmarkClosedChannel(b *testing.B) {
	c := make(chan struct{})

	package main

	import (
	"fmt"
	"time"
	)

	func main() {
	defer func() {
	if r := recover(); r != nil {

	package main

	import "fmt"

	func defer1() {
	var s = "Hello!"
	defer fmt.Println(s)
	s = "Ahoy!"
	}

	package main

	import (
	"bufio"
	"github.com/mediocregopher/radix/v3"
	"github.com/mediocregopher/radix/v3/resp"
	"net"
	"testing"
	)

	Hello, Gist!

	(this Gist is used for testing purposes)