Download files with names matching pattern from an Alveo list

search-n-download-alveo-list.R

############################## 1. Set up ######################################

# Make sure you have the necessary packages installed (see section 2).

# The last number from the list URL for example 1045 for:
# https://app.alveo.edu.au/item_lists/1045
alveo_list_id <- 904

# Regex pattern to search for 'speaker16.wav$' means 'ending with speaker16.wav'
alveo_search_pattern <- "speaker16.wav$"

# Directory to download the wav files into
output_path   <- "~/Desktop/"

########################### 2. Packages & config ##############################

# Install packages if you don't have them, e.g.:
# install.packages("pbapply")
library(purrr)
library(pbapply)
library(stringr)

# For alveo grab the latest copy from me (for now)
# devtools::install_github("fauxneticien/alveo-r")
library(alveo)

if(!file.exists("~/alveo.config")) {
   stop("Do you have alveo.config file in your home folder?")
}

client    <- RestClient(server_uri="app.alveo.edu.au")
item_list <- client$get_item_list_by_id(alveo_list_id)

################################ 3. Pipeline ##################################

# Note I've split the pipeline up to give some verbose output message() calls

message("Getting items from list at https://app.alveo.edu.au/item_lists/", alveo_list_id)

pblapply(X   = 1:length(item_list$items),
         FUN = function(an_item) { item_list$get_item(an_item) }) -> result

message("Getting metadata of each item in list...")

pblapply(X   = result,
         FUN = function(alveo_doc) {
            metadata <- alveo_doc$get_metadata()
            return(metadata$`ausnc:document`)
   }) -> result

message("Downloading files matching search string '", alveo_search_pattern,"' to ", output_path)

map(result, function(doc_string) {
   str_split(string = doc_string, pattern = ", ")[[1]] %>%
      keep(~ grepl(pattern = alveo_search_pattern, x = .))
   }) %>%

   map(~ Document(uri  = ., type = "audio/wav")) %>% 

   pblapply(FUN = function(alveo_obj) { alveo_obj$download(output_path) }) -> result