Cookbooks
Go
Chapter 2. Strings
Deleting characters matching a pattern

Go:
Deleting characters matching a pattern

How to:

In Go, deleting characters that match a pattern can be efficiently accomplished using the regexp package. Here, we’ll show how to remove all digits, then all non-alphanumeric characters from a string as examples.

  1. Removing All Digits:
package main

import (
    "fmt"
    "regexp"
)

func main() {
    text := "Go1 is cool, but Go2 will be cooler! Now: 2023."
	
    // Compile the regular expression for digits
    re, err := regexp.Compile("[0-9]+")
    if err != nil {
        fmt.Println("Error compiling regex:", err)
        return
    }
	
    // Replace digits with an empty string
    result := re.ReplaceAllString(text, "")
	
    fmt.Println(result) // Output: Go is cool, but Go will be cooler! Now: .
}
  1. Removing All Non-Alphanumeric Characters:
package main

import (
    "fmt"
    "regexp"
)

func main() {
    text := "Go is #1 @ programming languages!"
	
    // Compile the regular expression for non-alphanumeric characters
    re, err := regexp.Compile("[^a-zA-Z0-9]+")
    if err != nil {
        fmt.Println("Error compiling regex:", err)
        return
    }
	
    // Replace non-alphanumeric characters with an empty string
    result := re.ReplaceAllString(text, "")
	
    fmt.Println(result) // Output: Gois1programminglanguages
}

Deep Dive

The regexp package in Go provides a powerful interface for pattern matching and manipulation with regular expressions. Its implementation is derived from RE2, a regular expression library designed to guarantee a linear time execution, avoiding the possibility of “catastrophic backtracking” issues present in some other regex engines. This makes Go’s regex relatively safe and efficient for a wide array of applications.

While the regexp package is a comprehensive solution for dealing with patterns, it’s worth noting that for simpler or highly specific string manipulations, other string functions like strings.Replace(), strings.Trim(), or slicing might offer more performant alternatives. Regular expressions are a powerful tool, but their relative computational expense means that for operations that can be specified without them, exploring standard library alternatives can sometimes lead to simpler and more efficient code.

Converting a string to lower caseExtracting substrings