nealrichardson commented on a change in pull request #9143:
URL: https://github.com/apache/arrow/pull/9143#discussion_r555288000
##########
File path: r/tests/testthat/test-dataset.R
##########
@@ -303,11 +303,68 @@ test_that("Other text delimited dataset", {
filter(integer > 6) %>%
summarize(mean = mean(integer))
)
+})
+
+test_that("readr parse options", {
+ arrow_opts <- names(formals(CsvParseOptions$create))
+ readr_opts <- names(formals(readr_to_csv_parse_options))
+
+ # Arrow and readr parse options must be mutually exclusive, or else the code
+ # in `csv_file_format_parse_options()` will error or behave incorrectly. A
+ # failure of this test indicates that these two sets of option names are not
+ # mutually exclusive.
+ expect_equal(
+ intersect(arrow_opts, readr_opts),
+ character(0)
+ )
- # Now with readr option spelling (and omitting format = "text")
- ds3 <- open_dataset(tsv_dir, partitioning = "part", delim = "\t")
+ # With not yet supported readr parse options
+ # (remove this after ARROW-8631)
+ if (!"na" %in% readr_opts) {
+ expect_error(
+ open_dataset(tsv_dir, partitioning = "part", delim = "\t", na = "\\N"),
+ "supported"
+ )
+ }
Review comment:
Provided that you agree, here's a suggestion version you can just accept
```suggestion
# With not yet supported readr parse options (ARROW-8631)
expect_error(
open_dataset(tsv_dir, partitioning = "part", delim = "\t", na = "\\N"),
"supported"
)
```
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
