This is an automated email from the ASF dual-hosted git repository.
thisisnic 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 f0303652b4 ARROW-17699: [R] Add better error message for if a
non-schema passed into open_dataset() (#14108)
f0303652b4 is described below
commit f0303652b4934a9f767dca88268016c69375687d
Author: Nic Crane <[email protected]>
AuthorDate: Wed Sep 21 15:46:58 2022 +0100
ARROW-17699: [R] Add better error message for if a non-schema passed into
open_dataset() (#14108)
Example below shows what happens if the `schema()` function is passed in as
the `schema` argument via the `...` argument to `open_dataset()`.
Before (a later check for something else catches it, this giving a
misleading message):
```
library(dplyr)
library(arrow)
tf <- tempfile()
dir.create(tf)
write_dataset(mtcars, tf, format = "csv")
open_dataset(tf, format = "csv", schema = schema) %>% collect()
#> Error in `CsvFileFormat$create()`:
#> ! Values in `column_names` must match `schema` field names
#> ✖ `column_names` and `schema` field names match but are not in the same
order
```
After (more accurate error message):
```
library(dplyr)
library(arrow)
tf <- tempfile()
dir.create(tf)
write_dataset(mtcars, tf, format = "csv")
open_dataset(tf, format = "csv", schema = schema) %>% collect()
#> Error in `CsvFileFormat$create()`:
#> ! `schema` must be an object of class 'Schema' not 'function'.
```
Authored-by: Nic Crane <[email protected]>
Signed-off-by: Nic Crane <[email protected]>
---
r/R/dataset-format.R | 7 +++++++
r/tests/testthat/test-dataset-csv.R | 10 ++++++++++
2 files changed, 17 insertions(+)
diff --git a/r/R/dataset-format.R b/r/R/dataset-format.R
index 948abf2829..ad2a38a4f4 100644
--- a/r/R/dataset-format.R
+++ b/r/R/dataset-format.R
@@ -128,6 +128,13 @@ CsvFileFormat$create <- function(...,
options <- list(...)
schema <- options[["schema"]]
+ if (!is.null(schema) && !inherits(schema, "Schema")) {
+ abort(paste0(
+ "`schema` must be an object of class 'Schema' not '",
+ class(schema)[1],
+ "'."
+ ))
+ }
column_names <- read_options$column_names
schema_names <- names(schema)
diff --git a/r/tests/testthat/test-dataset-csv.R
b/r/tests/testthat/test-dataset-csv.R
index 0718746624..969bb82e40 100644
--- a/r/tests/testthat/test-dataset-csv.R
+++ b/r/tests/testthat/test-dataset-csv.R
@@ -380,3 +380,13 @@ test_that("skip argument in open_dataset", {
)
expect_equal(collect(ds), tbl)
})
+
+test_that("error message if non-schema passed in as schema to open_dataset", {
+
+ # passing in the schema function, not an actual schema
+ expect_error(
+ open_dataset(csv_dir, format = "csv", schema = schema),
+ regexp = "`schema` must be an object of class 'Schema' not 'function'.",
+ fixed = TRUE
+ )
+})
