---
title: "Custom logic"
date: "`r Sys.Date()`"
output:
  rmarkdown::html_vignette
vignette: >
  %\VignetteEncoding{UTF-8}
  %\VignetteIndexEntry{How to: implement custom logic.}
  %\VignetteEngine{knitr::rmarkdown}
---

When saving your modifications, eDT will just update the data hereby only
checking the data types and foreign tables. Sometimes however, you might want
some specific custom behavior or extra checks that `editbl` doesn't support out
of the box. There are two main solutions to write custom logic for the
modifications that were made.

```{r echo = TRUE, results = 'hide'}
library(shiny)
library(editbl)
```

## Option 1

Add your own implementations for `e_rows_update()`, `e_rows_insert()` and/or
`e_rows_delete`. This is arguably the most flexible approach since it nicely
integrates with the behavior of eDT.

```{r, }
ui <- fluidPage(eDTOutput("DT"))

# Define a custom update function
e_rows_update.mtcars <- function(
    x,
    y,
    by = NULL,
    ...,
    match = match,
    unmatched = c("error", "ignore"),
    copy = FALSE,
    in_place = FALSE){		
		
		# Extra checks
		if(any(y[,"mpg"] < 0)){
			stop("mpg should be a positive value!")	
		}
		
		# Logging
		print("Updated rows:")
		print(y)
		
		# Drop extra class and execute default code
		class(x) <- class(x)[class(x) != 'mtcars']
		e_rows_update(
	      x = x,
	      y = y,
	      by = by,
	      ... ,
	      match = match,
	      unmatched = unmatched,
	      copy = copy,
	      in_place = in_place
	  )
	}

server <- function(input, output, session){
	
	# Add a new class to your data object
	df = tibble::tibble(mtcars)
	class(df) <- c("mtcars", class(df))

	  result = eDT(
	    id = "DT",
	    data = df
	)
	   
}
shinyApp(ui, server)
```


## Option 2

You can capture the modifications made by editbl and handle them downstream in
the app. This is more straightforward, but doesn't support adding any logic
before saving.


```{r, }
ui <- fluidPage(eDTOutput("DT"))

server <- function(input, output, session){
	  result = eDT(
	    id = "DT",
	    data = mtcars
	  )
	  
	  # Some logging
	  observe({
		print('These rows have been permanently deleted:')
		print(result$deleted())
	  })
	   
}
shinyApp(ui, server)
```