How to interact with firebase from a shinyapp

Find how to use httr to access or modify stored data in firebase.

At the planning phase of creating a shiny app you would find the importance of having storing data and interact with the stored system.

You might have interaction with data bases using dbplyr, DBI, among others packages. Reading this blog you will find how to use httr in order to access or modify stored data in firebase.

Interacting with firebase from R

Firstly, you need to have all setup in firebase so R can connect trough API, it is recommended to store the API_KEY, firebase_url and password in the .Renviron file in the root of the app (where ui.R and server.R are stored or app.R).

Secondly, it is very important to be familiar with JSON structures in order to design how you are going to store your data in firebase.

From previous, you might want to bring the information, update, delete, insert, among others. Let’s build together the select.

Define JSON structure to store data

In my learning polish app, I design to have a list of words, this list would contains a list of categories, and each of this would have the register with the word, translation in spanish and date of insertion in a simple text.

For the purpose of the blog I am going to use a fragment of data from my learning polish shiny app.

  "words" : {
    "animals" : [
        "pies: perro: 2022-01-23",
        "kot: gato: 2022-01-23",
        "biedronka: mariquita: 2022-01-23",
        "Ptak: Pájaro: 2022-01-23",
        "Komar: Mosquito: 2022-01-23", ],
    "clothes" : [
        "buty: zapatos:2022-01-26",
        "spodnie: pantalón: 2022-01-30",
        "sweter: sueter: 2022-02-22",
        "krawat: corbata: 2022-02-22",
        "koszula: camisa: 2022-02-22" ],

Selecting data from firebase

The firebase url given by google is the place where your data is stored. It will look somethis as followwing: “”

If you would like to access to the words inside the category clothes, you might add the list words and clothes in the previous link, as follors:


In the documentary from firebase, you can find that you need to add .json when you are using an API (I invite you to read documentation to more detail).

select_words <- function(categories) {
    words <- httr::content(
          Sys.getenv("FIREBASE_URL"), "/words/", categories, ".json")
    ) %>%
    purrr::flatten() %>%
select_categories <- function() {
  categories <- content(GET(
    paste0(Sys.getenv("FIREBASE_URL"), "/words/.json")

Insert data to firebase:

For inserting data you can use the function PUT from httr. Into words variable we bring the words from category so we can add the new word to already stored words and converted to json with jsonlite package.

add_words <- function(categories, word) {
  if (word != "") {
    words <- select_words(categories)
    body <- jsonlite::toJSON(c(words, word),
      pretty = TRUE
    response <- httr::PUT(
      paste0(Sys.getenv("FIREBASE_URL"), "/words/", categories, ".json"),
      body = body

Delete data stored in firebase from R.

The following function receives the name of the category and one or more words (example: ptak and komar).

The first purrr::map compares each word with the list of words inside the category and save the position where it is stored in firebase.

The second purrr::map iterates over positions and tells firebase wich position to DELETE.

delete_words <- function(categories, word) {
  words_delete <- purrr::map(
    .x = stringr::str_to_lower(word),
    .f = function(.x) {
        paste0(Sys.getenv("FIREBASE_URL"), "/words/", categories, ".json")
      )) %>%
        stringi::stri_trans_tolower(.) %>%
        unique() %>%
        stringr::str_starts(.x) %>%
        which() - 1
    .x = words_delete,
    .f = function(.x) {
          Sys.getenv("FIREBASE_URL"), "words/", categories, "/", .x, ".json"

Thanks for reading. Any comments or feedback I would love to hear from you, you can have my info from contact.