context("Testing pdb.annotate()")


test_that("PDB annotation works", {
  skip_on_cran()
  skip_on_travis()
  skip_if_not(requireNamespace("httr", quietly=TRUE), 
              message = "Need httr installed to run the test")
  
  expected <- c('3R1C_X', '3R1C_B', '3R1C_C', '3R1C_D',
                '3R1C_E', '3R1C_F', '3R1C_G', '3R1C_H',
                '3R1C_I', '3R1C_J', '3R1C_K', '3R1C_L',
                '3R1C_M', '3R1C_N', '3R1C_O', '3R1C_P',
                '3R1C_Q', '3R1C_R', '3R1C_S', '3R1C_Y',
                '3R1C_T', '3R1C_U', '3R1C_W', '3R1C_A',
                '3R1C_V', '3R1C_Z', '3R1C_a', '3R1C_b',
                '3R1C_c', '3R1C_d', '3R1C_e', '3R1C_f',
                '3R1C_g', '3R1C_h', '3R1C_i', '3R1C_j')

  invisible(capture.output(anno <- pdb.annotate(expected)))
  expect_identical(rownames(anno), expected)

  expected <- c('3R1C_A', '3R1C_a', '3R1C_B', '3R1C_b',
                '3R1C_C', '3R1C_c', '3R1C_D', '3R1C_d',
                '3R1C_E', '3R1C_e', '3R1C_F', '3R1C_f',
                '3R1C_G', '3R1C_g', '3R1C_H', '3R1C_h',
                '3R1C_I', '3R1C_i', '3R1C_J', '3R1C_j',
                '3R1C_K', '3R1C_L', '3R1C_M', '3R1C_N',
                '3R1C_O', '3R1C_P', '3R1C_Q', '3R1C_R',
                '3R1C_S', '3R1C_Y', '3R1C_T', '3R1C_U',
                '3R1C_W', '3R1C_X', '3R1C_V', '3R1C_Z')
  
  invisible(capture.output(anno <- pdb.annotate('3R1C')))
  # 'sort' because sometimes PDB returns chain IDs in an uncertain way...
  expect_identical(sort(rownames(anno)), sort(expected))

  expected <- c('3R1C_A', '3R1C_a', '3R1C_B', '3R1C_b',
                '3R1C_C', '3R1C_c', '3R1C_D', '3R1C_d',
                '3R1C_E', '3R1C_e', '3R1C_F', '3R1C_f',
                '3R1C_G', '3R1C_g', '3R1C_H', '3R1C_h',
                '3R1C_I', '3R1C_i', '3R1C_J', '3R1C_j',
                '3R1C_K', '3R1C_L', '3R1C_M', '3R1C_N',
                '3R1C_O', '3R1C_P', '3R1C_Q', '3R1C_R',
                '3R1C_S', '3R1C_Y', '3R1C_T', '3R1C_U',
                '3R1C_W', '3R1C_X', '3R1C_V', '3R1C_Z',
                '1CDK_A', '1CDK_B', '1CDK_I', '1CDK_J')
  
  invisible(capture.output(anno <- pdb.annotate(c('3R1C', '1CDK'))))
  # 'sort' because sometimes PDB returns chain IDs in an uncertain way...
  expect_identical(sort(rownames(anno)), sort(expected))

  expect_warning(  
    invisible(capture.output(anno <- pdb.annotate(c('3R1C_A', '3r1c_a', '3r1c_q'))))
  )
  expect_identical(rownames(anno), expected[c(1, 2)])

  invisible(capture.output(anno <- pdb.annotate(c('3R1C_A', '3r1c_a'), unique=TRUE)))
  expect_identical(rownames(anno), "3R1C")
  expect_identical(anno$chainId, "A;a")
  
  
  expected <- sort(rep(c("ANP", "MN (2)", "MYR"), 2))
  invisible(capture.output(anno <- pdb.annotate(c('1cdk_A', '1cdk_B'), anno.terms="ligandId")))
  expect_identical(sort(unlist(strsplit(anno$ligandId, ","))), expected)

  ## Test cases without ligands
  ids <- c("6AXG_B", "1AA9_A", "7L0F_A", "6AMB_A", "121P_A", "5E95_A")
  invisible(capture.output(anno <- pdb.annotate(ids)))
  expect_identical(paste(anno[, "structureId"], anno[, "chainId"], sep="_"), ids)
  expect_true(is.na(anno[1, "ligandId"]))
  expect_identical(anno[2, "experimentalTechnique"], "NMR")

  ## Test cases without 'source'
  ids <- c("1P79", "4os7")
  invisible(capture.output(anno <- pdb.annotate(ids)))
  expect_true(is.na(anno[1, "source"]))
  expect_true(is.na(anno[3, "source"]))
  expect_identical(anno[1, "macromoleculeType"], "RNA")
  expect_identical(anno[3, "macromoleculeType"], "Protein")
  
  ## Test cases with duplicated 'sources'
  ids <- c("5TOK")
  invisible(capture.output(anno <- pdb.annotate(ids)))
  expect_identical(anno[1, "ligandId"], "NAG")

  ## Test error handling
  expect_error(pdb.annotate("1234"), regexp="No data retrieved")
})