Joe Cheng jcheng5

To opt into a more recent version of jQuery, you can add an htmlDependency object pointing to your desired version, somewhere (anywhere) in your UI:

htmltools::htmlDependency("jquery", "3.3.1",
  src = c(href = "https://code.jquery.com/"),
  script = "jquery-3.3.1.min.js")

This example will serve jQuery 3.3.1 from the jQuery CDN. If your app needs to work with clients that won't be able to connect to the wider internet, you'll need to download the jquery-3.3.1.min.js file, put it in an app subdirectory (say, jquery), and point to the directory using the src argument.

Why use modules?
- Reuse
- Organization/wrangling complexity (small pieces reliably composed)
Functions vs. modules
- Namespacing Shiny inputs/outputs
Passing values in/getting values out
- Reactives by name vs. by value
- ^ in both directions (arguments and return values)
Connecting modules

How can module UI compose? ("higher-order" modules)

This snippet shows how you could use later to create an API that can be called on a background thread, to perform an arbitrary task on the main R thread, and return the result to the background thread. The background thread blocks until the result is returned from the main R thread.

	library(shiny)

	bigvec <- paste0("a", 1:1e5)
	named_bigvec <- setNames(bigvec, bigvec)
	nested_biglist <- lapply(named_bigvec, function(item) setNames(list(item), item))

	test_set <- list(
	"Unnamed vector" = c(1, 2),
	"Named vector" = c(a = 1, B = 2),
	"Partially named vector" = c(a = 1, B = 2, 3),

	// Before running this script, set the LDAP_PASSWORD environment variable in
	// the shell. You can do this without echoing the password by running:
	//
	// read -s LDAP_PASSWORD && export LDAP_PASSWORD
	//
	// You'll also need to modify the variable definitions at the top of this file.

	// == BEGIN configuration variables ===============

	var url = "ldap://adtest.rstudio.com/dc=adtest,dc=rstudio,dc=com";

	base_task_iterator <- function(should_continue, iter_body) {
	if (should_continue()) {
	iter_body()
	later::later(~base_task_iterator(should_continue, iter_body))
	}

	invisible()
	}

	while_task_iterator <- function(cancelExpr, whileExpr, func) {

	library(shiny)
	library(ggplot2)

	ui <- fluidPage(
	sidebarLayout(
	sidebarPanel(
	selectInput("dataset", "Dataset", c("cars", "pressure", "mtcars")),
	uiOutput("input_ui")
	),
	mainPanel(

	library(shiny)
	library(dplyr)
	library(ggplot2)

	dedupe <- function(r) {
	x <- reactiveVal(isolate(r()))

	o <- observe({
	x(r())
	})

	snippet shinymod
	${1:name}UI <- function(id) {
	ns <- NS(id)
	tagList(
	${0}
	)
	}

	${1:name} <- function(input, output, session) {