Some updates to read_file() (#727)

* Use `!!` injection operator as a simpler workaround apache/arrow#36658 * Add explicit `.rds.gz` and `.csv.gz` extension handling Previously this assumed any file ending in `.gz` was a CSV. * Style code --------- Co-authored-by: Moohan <[email protected]> Co-authored-by: Jennit07 <[email protected]>
Public-Health-Scotland · Jul 17, 2023 · 5272eed · 5272eed
1 parent 3f86f89
commit 5272eed
Show file tree

Hide file tree

Showing 2 changed files with 35 additions and 13 deletions.
diff --git a/R/read_file.R b/R/read_file.R
@@ -16,10 +16,27 @@
 #' @return the data a [tibble][tibble::tibble-package]
 #' @export
 read_file <- function(path, col_select = NULL, as_data_frame = TRUE, ...) {
-  valid_extensions <- c("rds", "fst", "sav", "zsav", "csv", "gz", "parquet")
+  valid_extensions <- c(
+    "rds",
+    "rds.gz",
+    "fst",
+    "sav",
+    "zsav",
+    "csv",
+    "csv.gz",
+    "parquet"
+  )
 
   ext <- fs::path_ext(path)
 
+  if (ext == "gz") {
+    ext <- paste(
+      fs::path_ext(fs::path_ext_remove(path)),
+      "gz",
+      sep = "."
+    )
+  }
+
   if (!(ext %in% valid_extensions)) {
     cli::cli_abort(c(
       "x" = "Invalid extension: {.val {ext}}",
@@ -36,17 +53,19 @@ read_file <- function(path, col_select = NULL, as_data_frame = TRUE, ...) {
   }
 
   data <- switch(ext,
-    "rds" = readr::read_rds(path),
-    "fst" = fst::read_fst(path),
-    "sav" = haven::read_spss(path, ...),
-    "zsav" = haven::read_spss(path, ...),
-    "csv" = readr::read_csv(path, ..., show_col_types = FALSE),
-    "gz" = readr::read_csv(path, ..., show_col_types = FALSE),
-    "parquet" = if (is.null(col_select)) {
-      arrow::read_parquet(path, as_data_frame = as_data_frame, ...)
-    } else {
-      arrow::read_parquet(path, col_select = col_select, as_data_frame = as_data_frame, ...)
-    }
+    "rds" = readr::read_rds(file = path),
+    "rds.gz" = readr::read_rds(file = path),
+    "fst" = tibble::as_tibble(fst::read_fst(path = path)),
+    "sav" = haven::read_spss(file = path, ...),
+    "zsav" = haven::read_spss(file = path, ...),
+    "csv" = readr::read_csv(file = path, ..., show_col_types = FALSE),
+    "csv.gz" = readr::read_csv(file = path, ..., show_col_types = FALSE),
+    "parquet" = arrow::read_parquet(
+      file = path,
+      col_select = !!col_select,
+      as_data_frame = as_data_frame,
+      ...
+    )
   )
 
   return(data)

diff --git a/tests/testthat/test-read_file.R b/tests/testthat/test-read_file.R
@@ -1,5 +1,6 @@
 test_that("read_file works", {
   rds_path <- tempfile(fileext = ".rds")
+  rds_gz_path <- tempfile(fileext = ".rds.gz")
   fst_path <- tempfile(fileext = ".fst")
   sav_path <- tempfile(fileext = ".sav")
   zsav_path <- tempfile(fileext = ".zsav")
@@ -10,6 +11,7 @@ test_that("read_file works", {
   aq_data <- tibble::as_tibble(datasets::airquality)
 
   readr::write_rds(aq_data, rds_path)
+  readr::write_rds(aq_data, rds_gz_path)
   fst::write_fst(aq_data, fst_path)
   haven::write_sav(aq_data, sav_path)
   haven::write_sav(aq_data, zsav_path, compress = "zsav")
@@ -18,7 +20,8 @@ test_that("read_file works", {
   arrow::write_parquet(aq_data, parquet_path)
 
   expect_equal(aq_data, read_file(rds_path))
-  expect_equal(aq_data, tibble::as_tibble(read_file(fst_path)))
+  expect_equal(aq_data, read_file(rds_gz_path))
+  expect_equal(aq_data, read_file(fst_path))
   expect_equal(aq_data, haven::zap_formats(read_file(sav_path)))
   expect_equal(aq_data, haven::zap_formats(read_file(zsav_path)))
   expect_equal(aq_data, read_file(csv_gz_path))