Decorator Design Pattern

[harell]

Intent

Attach additional responsibilities to an object dynamically. Decorators provide a flexible alternative to subclassing for extending functionality.

Also Known As

Wrapper

Example

print_command_and_execute <- function(command) {
    function_name <- as.character(substitute(command))
    function_name <- function_name[length(function_name)]
    wrapper <- function(...){
        args <- list(...)
        args_as_string <- character(0)
        for(i in seq_len(length(args))) 
            args_as_string[i] <- paste(names(args)[i],"=",as.character(args[i]))
        args_as_string <- paste(args_as_string, collapse = ", ")
        
        
        message("\033[43m\033[44m", function_name, "(", args_as_string, ")", "\033[43m\033[49m")
        return(command(...))
    }
    
    return(wrapper)
}

setdiff <- print_command_and_execute(base::setdiff)

> setdiff(x = 1:3, y = 1)
setdiff(x = 1:3, y = 1)
[1] 2 3

[harell]

Real-world examples

• Time a function setdiff <-timer(base::setdiff)
• passing a block: avoiding temporal coupling
• memorize (issue Memento Design Pattern #16)
• tidylog
• ViewPipeSteps
• purrr: safely, possibly, quietly, auto_browser, slowly, insistently
• Timeout decorator

[harell]

Further reading

• How to create decorators in R
• tinsel: Decorating functions in R
• distr6 Decorators article
• purrr function operators = decorators