using httr to get spotify track data

old_spotify_scrape.R

## thanks to http://www.r-bloggers.com/the-eurovision-2016-song-contest-in-an-r-shiny-app/ 

library(httr)
library(dplyr)
library(purrr)

## I Created a spotify developer account and was given these.
clientID = 'XXXXXXXXX'
secret = 'XXXXXXXXXX'

response = POST(
  'https://accounts.spotify.com/api/token',
  accept_json(),
  authenticate(clientID, secret),
  body = list(grant_type = 'client_credentials'),
  encode = 'form',
  verbose()
)

mytoken = content(response)$access_token

user_id = '129342083'
user_id = 'spotify'
#user_id = 'radioedit'

HeaderValue = paste0('Bearer ', mytoken)


URI = paste0('https://api.spotify.com/v1/users/', user_id, '/playlists')

## get the total number of playlists
response2 = GET(url = URI, add_headers(Authorization = HeaderValue), query = list(offset = 0, limit = 50))
Playlists = content(response2)
num_playlists <- Playlists$total

## function to pull the playlist href from the response content.
reffer <- function(n) {
  Playlists$items[[n]]$href
}


if (num_playlists >= 50) {
  offsets <- seq(0, num_playlists-50, by = 50)
  
  ## builds list of hrefs for everyone of users playlist.
  hrefs <- c()
  for (i in offsets) {
    response2 = GET(url = URI, add_headers(Authorization = HeaderValue), query = list(offset = i, limit = 50))
    Playlists = content(response2)
    
    reffer2 <- unlist(lapply(seq(1:50), FUN = reffer))
    hrefs <- c(hrefs, reffer2)
  }
  
  ## Gets reamining hrefs
  last_bit_offset =  max(offsets)+50
  last_bit_iter = num_playlists-last_bit_offset
  
  response2 = GET(url = URI, add_headers(Authorization = HeaderValue), query = list(offset = last_bit_offset, limit = 50))
  Playlists = content(response2)
  
  reffer2 <- unlist(lapply(seq(1:last_bit_iter), FUN = reffer))
  hrefs <- c(hrefs, reffer2)
} else {
  response2 = GET(url = URI, add_headers(Authorization = HeaderValue), query = list(offset = i, limit = 50))
  Playlists = content(response2)
  
  reffer2 <- unlist(lapply(seq(1:num_playlists), FUN = reffer))
  hrefs <- reffer2
}
## pulls track info from playlist given a track number.
tracker <- function(playlist,track_index) {
  i <- track_index
  x <- list(playlist_track_number = i,
            track_name = playlist$tracks$items[[i]]$track$name,
            track_id = playlist$tracks$items[[i]]$track$id,
            track_artist1 = playlist$tracks$items[[i]]$track$artists[[1]]$name,
            track_artist1_id = playlist$tracks$items[[i]]$track$artists[[1]]$id,
            track_artists_mess = playlist$tracks$items[[i]]$track$artists,
            track_album = playlist$tracks$items[[i]]$track$album$name,
            track_album_id = playlist$tracks$items[[i]]$track$album$id,
            track_added_at = playlist$tracks$items[[i]]$added_at,
            playlist_name = playlist$name,
            playlist_id = playlist$id,
            playlist_owner = playlist$owner$id
  )
  return(x)
}




## Gets all the tracks info from a playlist, and outputs a dataframe, currently only works on 100 tracks
playlist_tracks <- function(playlist) {
  if(is.null(playlist$tracks)) {return(data.frame())}
  if(playlist$tracks$total >  0) {
  tracks <- seq(1,min(100,playlist$tracks$total)) %>%
    map(~tracker(playlist = playlist, track_index = .)) 
  df <- do.call(rbind, tracks)
  return(df)}
}

# calls playlist_tracks on every href.
list_of_lists <- hrefs %>%
  map(~playlist_tracks(content(GET(url = ., add_headers(Authorization = HeaderValue) ))))
lol <- data.frame(do.call(rbind, list_of_lists))

#Drops track_artist_mess so that it can be written to csv.
lol <- lol %>% mutate(playlist_track_number =  as.integer(playlist_track_number),
                track_name = as.character(track_name),
                track_id = as.character(track_id),
                track_artist1 = as.character(track_artist1),
                track_artist1_id = as.character(track_artist1_id),
                #track_artists_mess = ?
                track_album = as.character(track_album),
                track_album_id = as.character(track_artist1),
                track_added_at = as.character(track_added_at),
                playlist_name = as.character(playlist_name),
                playlist_id = as.character(playlist_id),
                playlist_owner = as.character(playlist_owner))



# lol %>% select(-track_artists_mess) %>% write.csv(paste0('spotifiy_', Sys.Date(),'.csv'))

lol <- lol %>% mutate(track_artist_mess = jsonlite::toJSON(.$track_artists_mess))

# lol %>%  select(-track_artists_mess)