This is an automated email from the ASF dual-hosted git repository.
npr pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/arrow.git
The following commit(s) were added to refs/heads/master by this push:
new dfc1480 ARROW-11253 [R]: Make sure that large metadata tests are
reproducible
dfc1480 is described below
commit dfc1480fe8e93f98fc06981bc2f45f7abc31a8f5
Author: Jonathan Keane <[email protected]>
AuthorDate: Thu Jan 14 14:58:55 2021 -0800
ARROW-11253 [R]: Make sure that large metadata tests are reproducible
Don't assert that the large strings are the same (since they might vary
acorss R versions if those versions change the way the seed works e.g. 3.6.0)
Closes #9205 from jonkeane/ARROW-11253-metadata-failures
Authored-by: Jonathan Keane <[email protected]>
Signed-off-by: Neal Richardson <[email protected]>
---
.../golden-files/data-arrow-extra-meta_3.0.0.parquet | Bin 13263 -> 7862 bytes
r/tests/testthat/helper-data.R | 6 +++++-
r/tests/testthat/test-metadata.R | 2 +-
3 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/r/tests/testthat/golden-files/data-arrow-extra-meta_3.0.0.parquet
b/r/tests/testthat/golden-files/data-arrow-extra-meta_3.0.0.parquet
index bf95f23..3394be2 100644
Binary files
a/r/tests/testthat/golden-files/data-arrow-extra-meta_3.0.0.parquet and
b/r/tests/testthat/golden-files/data-arrow-extra-meta_3.0.0.parquet differ
diff --git a/r/tests/testthat/helper-data.R b/r/tests/testthat/helper-data.R
index ecce773..15ea0fc 100644
--- a/r/tests/testthat/helper-data.R
+++ b/r/tests/testthat/helper-data.R
@@ -68,9 +68,13 @@ make_big_string <- function() {
rep(purrr::map_chr(2047:2050, ~paste(sample(letters, ., replace = TRUE),
collapse = "")), 2^18)
}
-make_string_of_size <- function(size = 1) {
+make_random_string_of_size <- function(size = 1) {
purrr::map_chr(1000*size, ~paste(sample(letters, ., replace = TRUE),
collapse = ""))
}
+make_string_of_size <- function(size = 1) {
+ paste(rep(letters, length = 1000*size), collapse = "")
+}
+
example_with_extra_metadata <- example_with_metadata
attributes(example_with_extra_metadata$b) <- list(lots =
rep(make_string_of_size(1), 100))
diff --git a/r/tests/testthat/test-metadata.R b/r/tests/testthat/test-metadata.R
index 17c43bb..5abd3d2 100644
--- a/r/tests/testthat/test-metadata.R
+++ b/r/tests/testthat/test-metadata.R
@@ -105,7 +105,7 @@ test_that("Metadata serialization compression", {
# However strings where compression + serialization is not effective are no
# worse than only serialization alone
- large_few_strings <- as.list(rep(make_string_of_size(50), 2))
+ large_few_strings <- as.list(rep(make_random_string_of_size(50), 2))
large_few <- .serialize_arrow_r_metadata(large_few_strings)
expect_equal(
object.size(large_few),
