tests/testthat/test-backend-common.R

In torchaudio: R Interface to 'pytorch''s 'torchaudio'

filepath_mp3 <- system.file("sample_audio_1.mp3", package = "torchaudio")

test_that("set_audio_backend and torchaudio_load works", {

  expect_equal(get_audio_backend(), av_loader)

  set_audio_backend("tuneR")
  expect_equal(get_audio_backend(), tuneR_loader)
  audio <- torchaudio_load(filepath_mp3, offset = 1, duration = 100, unit = "samples")
  expect_equal(class(audio)[1], "Wave")
  expect_equal(length(audio), 100)

  set_audio_backend("av")
  expect_equal(get_audio_backend(), av_loader)
  audio <- torchaudio_load(filepath_mp3, offset = 1, duration = 100, unit = "samples")
  expect_equal(class(audio)[1], "av")
  expect_equal(length(audio), 100)

  expect_error(set_audio_backend("sox"))
})