Regression tests for writeObject & writeType refactor

write_type_regression_test.R

old_env = new.env()
URL = 'https://raw.githubusercontent.com/apache/spark/master/R/pkg/R/serialize.R'
source(URL, old_env)

new_env = new.env()
# run from ./R/pkg on the write_type_object_s3 branch
source('R/serialize.R', new_env)

# missing: jobj
test_objects = list(
  1L, 1, as.raw(0x01),      # integer, numeric, raw
  'a', TRUE,                # character, logical
  list2env(                 
    list(int = 1L, num = 4) # environment (different types)
  ),
  list2env(list(a = 'b')),   # length-1 environment
  list('a', TRUE, 0L),       # list
  as.list(1:5),              # integer array
  as.list(letters),          # character array
  'abcdefg', '',             # length>1 character, length-0 character
  .Date(0), .POSIXct(0),     # Date, POSIXct
  list(), list(1L),          # empty list, length-1 list
  list(1:2),                 # length-2 list (so x[[1]] is an array)
  list2env(list(
    a = list(1:2),
    b = list(
      list(NULL), 'c'        # complicated env with nesting
    ),
    c = c('a', 'b')
  )),
  NA, NA_real_,              # NA logical, NA numeric
  NULL                       # NULL
)

for (ii in seq_along(test_objects)) {
  old_env$conn = rawConnection(raw(), 'wb')
  eval(quote(writeObject(conn, test_objects[[ii]])), old_env)
  oldBin = rawConnectionValue(old_env$conn)
  close(old_env$conn)
  
  new_env$conn = rawConnection(raw(), 'wb')
  eval(quote(writeObject(test_objects[[ii]], conn)), new_env)
  newBin = rawConnectionValue(new_env$conn)
  close(new_env$conn)
  
  if (!identical(oldBin, newBin)) {
    stop('Failed at object ', ii, 
         '\n\noldBin:\n', format(oldBin), 
         '\nnewBin:\n', format(newBin))
  }
}

# ** "EMPIRICAL" TESTS" **
# add this in writeObject S3 method:
  # TEST_FILE = 'pkg/writeObject.inputs'
  # repro_env = new.env()
  # if (file.exists(TEST_FILE)) {
  #   load(TEST_FILE, envir = repro_env)
  #   repro_env$obj_count = repro_env$obj_count + 1L
  # } else repro_env$obj_count = 1L
  # repro_env[[sprintf('object_%d', repro_env$obj_count)]] = object
  # save(list = ls(repro_env), envir = repro_env, file = TEST_FILE)
# then run tests & every time writeObject is called, its input will be
# added to the writeObject.inputs file; here, we test that all of these
# inputs produce the same output as the old version
test_env = new.env()
load('writeObject.inputs', envir = test_env)

# mock function
isValidJobj = function(...) TRUE
for (nm in ls(test_env)) {
  # internal class in refactor only
  if (class(get(nm, test_env)) == 'ArrayList') next
  old_env$conn = rawConnection(raw(), 'wb')
  eval(quote(writeObject(conn, test_env[[nm]])), old_env)
  oldBin = rawConnectionValue(old_env$conn)
  close(old_env$conn)
  
  new_env$conn = rawConnection(raw(), 'wb')
  eval(quote(writeObject(test_env[[nm]], conn)), new_env)
  newBin = rawConnectionValue(new_env$conn)
  close(new_env$conn)
  
  if (!identical(oldBin, newBin)) {
    stop('Failed at object ', nm, 
         '\n\noldBin:\n', format(oldBin), 
         '\nnewBin:\n', format(newBin))
  }
}