Introduction

This brief how-to guide covers the steps I follow to:

• Create an API client to access data from Genius.com,
• Store API credentials in an .Renviron file, and
• Use the geniusr package to retrieve song lyrics from Genius.com.

These steps have been tested on my PC running Windows 11. On other machines and operating systems, your mileage may vary!

Here’s a general explainer on APIs that you may find helpful if you haven’t set up an API client before.

1. Create a Genius API client

1. Go to genius.com and sign up for an account or log in if you already have one.

2. Once logged in, click here to go to the page for creating a Genius API client. It will look something like this:

3. Fill in the details for your API client - required fields are App name and App website URL. I filled in the details shown below - when you’re ready, click “Save”.

4. Once your API client has been created, you will be redirected to a page that looks something like the screenshot below, but with the relevant details you filled in for your own API client.

Important: API keys are not inherently secure, so they should be treated like passwords.

5. Keep your web browser open for now - we will come back to grab the relevant API keys later on.

2. Store your API credentials in an .Renviron file

Before we press on, it’s important to think about how you organise files on your machine for your data science projects. My general approach is:

• Create a new repository on GitHub.
• Clone the repository to my local machine using RStudio, via the “New Project” method described here in the “Happy Git and GitHub for the useR” ebook. I clone the repo so that it is a subdirectory of a folder on my computer called “GitHub”.

I do not have my “GitHub” folder set to sync to any cloud storage services (e.g., Dropbox, OneDrive, and the like). This is very important for the next steps - we want to store the API keys on your local machine only and nowhere else, to reduce the opportunity for someone to steal your API keys!

6. Log in to Github.com and create a new repository. After creating a repo, you will be redirected to the home page for that repo. Here’s what the repo home page looks like for jacquietran/geniusr, just after creating the repo:

7. From the home page from your new repo, click the “Code” button and copy the URL to your clipboard.

8. Open RStudio. From the top nav bar, go to the “File” menu and click “New Project…” which will open the New Project Wizard.

9. Click on “Version control”.

10. Click on “Git”.

11. Into the “Repository URL” text box, paste the URL that you copied from your repo home page. This action will auto-populate the “Project directory name” text box. You can also click the “Browse” button if needed to navigate to a location on your machine where you would like to clone your repo as a subdirectory. When you’re ready, click “Create Project”.

12. RStudio will now open you up in a session that corresponds to the GitHub repo you’ve cloned. For more on how RStudio Projects work, so for more info, check out Chapter 8 - Workflow: projects from the “R for Data Science” book.

13. Now we’re ready to locally store your API keys! In RStudio, from the top nav bar, go to the “File” menu, hover over “New file”, and from the sub-menu select “Text File”.

14. Save this file to your cloned repo, with the filename .Renviron. Note: This file should be saved in the “top” level of your cloned repo, where the .Rproj file is also stored.

15. Go back to your web browser and to the page where your Genius API client details are recorded. If you happen to have closed the page, click here. Click on the link “Generate Access Token”. After a second or two, a text box will appear with your “Client Access Token”. Copy the string inside this text box to your clipboard.

15. Go back to RStudio. Edit your .Renviron file to add your Genius API token by pasting your Client Access Token where indicated in the screenshot below. Save your changes and close your .Renviron file.

Do not include the [ ] square brackets. They are shown below only to indicate where to paste your Client Access Token.

GENIUS_API_TOKEN = [paste your Client Access Token here]

16. Go to your “Git” tab in RStudio. At this stage, you will see the .Renviron you have just saved is listed here; this means that this file is being tracked by GitHub - we need to change this because we do not want to accidentally share our API keys with the world by uploading the .Renviron file to GitHub.

17. In RStudio, from your “Files” pane, navigate to the top level of the folder (subdirectory) where you have cloned your repo. You will see a .gitignore file in this subdirectory, which was created in the process of cloning the repo to your local machine by creating a new RStudio project. Click on this .gitignore file to open it up in the editor pane.

18. In the .gitignore file, add a line for .Renviron and save your changes. My .gitignore file now looks like this:

19. In RStudio, check your “Git” tab again - you should see that the .Renviron file is no longer listed, which means that that file is no longer being tracked. Yay!

3. Retrieve song lyrics using R

20. Install the geniusr package from CRAN:

install.packages("geniusr")

21. Load the geniusr library into your session.

library(geniusr)

22. At the time of preparing this tutorial, there is a bug in the geniusr package (described here). Until the bug is fixed in the package itself, you can apply a local patch as a workaround (thanks to MalcolmMashig). Copy the code below and run it in your console to apply the local patch:

# Load libraries for the workaround patch
library(rvest)
library(xml2)
library(rlang)

get_lyrics <- function (session) {
  lyrics <-  session %>% html_nodes(xpath = '//div[contains(@class, "Lyrics__Container")]')
  song <-  session %>% html_nodes(xpath = '//span[contains(@class, "SongHeaderVariantdesktop__")]') %>% html_text(trim = TRUE)
  artist <-  session %>% html_nodes(xpath = '//a[contains(@class, "SongHeaderVariantdesktop__Artist")]') %>% html_text(trim = TRUE)
  xml_find_all(lyrics, ".//br") %>% xml_add_sibling("p", "\n")
  xml_find_all(lyrics, ".//br") %>% xml_remove()
  lyrics <- html_text(lyrics, trim = TRUE)
  lyrics <- unlist(strsplit(lyrics, split = "\n"))
  lyrics <- grep(pattern = "[[:alnum:]]", lyrics, value = TRUE)
  if (is_empty(lyrics)) {
    return(tibble(line = NA, section_name = NA, section_artist = NA, 
                  song_name = song, artist_name = artist))
  }
  section_tags <- nchar(gsub(pattern = "\\[.*\\]", "", lyrics)) == 0
  sections <- geniusr:::repeat_before(lyrics, section_tags)
  sections <- gsub("\\[|\\]", "", sections)
  sections <- strsplit(sections, split = ": ", fixed = TRUE)
  section_name <- sapply(sections, "[", 1)
  section_artist <- sapply(sections, "[", 2)
  section_artist[is.na(section_artist)] <- artist
  tibble(line = lyrics[!section_tags], section_name = section_name[!section_tags], 
         section_artist = section_artist[!section_tags], song_name = song, 
         artist_name = artist)
}
assignInNamespace("get_lyrics", get_lyrics, "geniusr")

23. From the geniusr package, use the get_lyrics_search() function to retrieve lyrics from a song based on the artist name and song title. For example, the code below returns a tibble containing the lyrics for the song “Don’t Wanna Try”, by the band Saskwatch (an awesome Melbourne band that is awesome).

# Load {dplyr} for data wrangling
library(dplyr)

get_lyrics_search(
  artist_name = "Saskwatch",
  song_title = "Don't Wanna Try")

## # A tibble: 33 x 5
##    line                section_name       section_artist song_name   artist_name
##    <chr>               <chr>              <chr>          <chr>       <chr>      
##  1 Another day in lov~ Another day in lo~ Saskwatch      Don’t Wann~ Saskwatch  
##  2 Don't you know it'~ Another day in lo~ Saskwatch      Don’t Wann~ Saskwatch  
##  3 And I guess that i~ Another day in lo~ Saskwatch      Don’t Wann~ Saskwatch  
##  4 To give up all tha~ Another day in lo~ Saskwatch      Don’t Wann~ Saskwatch  
##  5 Don't you know, I ~ Another day in lo~ Saskwatch      Don’t Wann~ Saskwatch  
##  6 Each and every day  Another day in lo~ Saskwatch      Don’t Wann~ Saskwatch  
##  7 But I'm tired of b~ Another day in lo~ Saskwatch      Don’t Wann~ Saskwatch  
##  8 So boy, what's it ~ Another day in lo~ Saskwatch      Don’t Wann~ Saskwatch  
##  9 I've given all my ~ Another day in lo~ Saskwatch      Don’t Wann~ Saskwatch  
## 10 But I don't wanna ~ Another day in lo~ Saskwatch      Don’t Wann~ Saskwatch  
## # ... with 23 more rows