repair bug that reads a character in a numeric excel field (a version number like "9.3") and coinverts it into a character with too many significant digits

Author: douwe

R/read.R

@@ -80,7 +80,8 @@ read_keyvalue <- function(
       drfile,
       sheet = sheet,
       range = range,
-      col_names = c("key", "value")
+      col_names = c("key", "value"),
+      col_types = c("text", "list") # Key column is read as text; value column is read as "list" to preserve native Excel types (numeric, text, etc.), so that R's own as.character() is used for conversion instead of readxl's internal conversion which yields "9.3000000000000007" for 9.3
     )
   }) |>
     dplyr::bind_rows()
@@ -279,7 +280,7 @@ short_to_longnames <- gentranslator('short-long')
 #' Read all data from a spreadsheet according to a reporting template guide. The
 #' data will be returned as a list with the optional elements keyvalue, table and
 #' platedata if defined in the guide. The minimal and maximal template versions
-#' of the guide mustbe compatible with that of the template in which the data
+#' of the guide must be compatible with that of the template in which the data
 #' were recorded. Furthermore, the name of the template must match the template
 #' name in the guide when when `checkname` is `TRUE`.
 #' @details

man/read_data.Rd

@@ -1,5 +1,5 @@
 test_that("Reading a spreadsheet map works", {
-  expect_no_error(read_guide(test_path("testdata/guide_competition_1_0.yml")))
+  expect_no_error(read_guide(test_path("fixtures/guide_competition_1_0.yml")))
 })
 
 test_that("Function check_dim works", {
@@ -12,8 +12,8 @@ test_that("Function check_dim works", {
 })
 
 #test_that("check_guide works for correct guides", {
-  # expect_no_error(read_guide(test_path("testdata/goodguides/guide_competition_9_3.yml")))
-  # expect_no_error(read_guide(test_path("testdata/goodguides/guide_without_platedata.yml")))
-  # expect_no_error(read_guide(test_path("testdata/goodguides/guide_with_two_plates.yml")))
-  # expect_no_error(read_guide(test_path("testdata/goodguides/guide_without_templatemetadata.yml")))
+  # expect_no_error(read_guide(test_path("fixtures/goodguides/guide_competition_9_3.yml")))
+  # expect_no_error(read_guide(test_path("fixtures/goodguides/guide_without_platedata.yml")))
+  # expect_no_error(read_guide(test_path("fixtures/goodguides/guide_with_two_plates.yml")))
+  # expect_no_error(read_guide(test_path("fixtures/goodguides/guide_without_templatemetadata.yml")))
 #})

tests/testthat/test-guide.R

@@ -1,6 +1,6 @@
 
 test_that("Translation from long to short names works", {
-  guide <- read_guide(test_path("testdata/guide_competition_1_0.yml"))
+  guide <- read_guide(test_path("fixtures/guide_competition_1_0.yml"))
   expect_equal(long_to_shortnames("Version", guide$translations), "template.version")
 })
 
@@ -10,7 +10,7 @@ test_that("Function long_to_short yields correct result for missing translations
 })
 
 test_that("Function short_to_longnames works", {
-  guide <- read_guide(test_path("testdata/guide_competition_1_0.yml"))
+  guide <- read_guide(test_path("fixtures/guide_competition_1_0.yml"))
   expect_equal(short_to_longnames("template.version", guide$translations), "Version")
 })
 
@@ -20,66 +20,87 @@ test_that("Function short_to_long yields correct result for missing translations
 })
 
 test_that("reading a single key-value pair works", {
-  expect_no_error(read_keyvalue(drfile = test_path('testdata/test1.xlsx'), sheet = 'description', range = "A2:B2"))
+  expect_no_error(read_keyvalue(drfile = test_path('fixtures/test1.xlsx'), sheet = 'description', range = "A2:B2"))
   expected_result <- list('template.version' = '9.3')
-  guide <- read_guide(test_path("testdata/guide_competition_1_0.yml"))
-  expect_equal(read_keyvalue(drfile = test_path('testdata/test1.xlsx'), sheet = 'description', range = "A2:B2", translate = TRUE,
+  guide <- read_guide(test_path("fixtures/guide_competition_1_0.yml"))
+  expect_equal(read_keyvalue(drfile = test_path('fixtures/test1.xlsx'), sheet = 'description', range = "A2:B2", translate = TRUE,
                              translations = guide$translations), expected_result)
 })
 
 test_that("reading and coercing key-value pairs works", {
   expected_result_single <- list('max.spread' = 0.2)
-  guide <- read_guide(test_path("testdata/guide_competition_1_0.yml"))
-  expect_equal(read_keyvalue(drfile = test_path('testdata/test1.xlsx'), sheet = '_parameters', range = "A24:B24",
+  guide <- read_guide(test_path("fixtures/guide_competition_1_0.yml"))
+  expect_equal(read_keyvalue(drfile = test_path('fixtures/test1.xlsx'), sheet = '_parameters', range = "A24:B24",
                              translate = FALSE, translations = guide$translations, atomicclass = "numeric"), expected_result_single)
   expected_result_multiple <- list('ic0.min' = 65, ic0.max=135)
-  expect_equal(read_keyvalue(drfile = test_path('testdata/test1.xlsx'), sheet = '_parameters', range = "A27:B28",
+  expect_equal(read_keyvalue(drfile = test_path('fixtures/test1.xlsx'), sheet = '_parameters', range = "A27:B28",
                              translate = FALSE, translations = guide$translations, atomicclass = "numeric"), expected_result_multiple)
 })
 
 test_that("reading multiple key-value pairs works", {
   expected_result <- list('studyID' = 'MyProject', 'exptID' = 'TTRfitc-020', 'plateID' = '1', 'instrID' = 'Instr1')
-  expect_no_error(read_keyvalue(drfile = test_path('testdata/test1.xlsx'), sheet = 'description', range = "A10:B13"))
-  guide <- read_guide(test_path("testdata/guide_competition_1_0.yml"))
-  expect_equal(read_keyvalue(drfile = test_path('testdata/test1.xlsx'), sheet = 'description', range = "A10:B13", translate = TRUE,
+  expect_no_error(read_keyvalue(drfile = test_path('fixtures/test1.xlsx'), sheet = 'description', range = "A10:B13"))
+  guide <- read_guide(test_path("fixtures/guide_competition_1_0.yml"))
+  expect_equal(read_keyvalue(drfile = test_path('fixtures/test1.xlsx'), sheet = 'description', range = "A10:B13", translate = TRUE,
                              translations = guide$translations), expected_result)
 })
 
 test_that("Reading cells works", {
   variables <- list(list(name = "spread.itm1", cell = "G6"),
                     list(name = "spread.itm2", cell = "G33"))
   result <- list(spread.itm1 = "0.0463713477851084", spread.itm2 = "0.0154571159283695")
-  expect_no_error(read_cells(drfile = test_path('testdata/test1.xlsx'), sheet = 'BGfluo', variables = variables))
-  expect_equal(read_cells(drfile = test_path('testdata/test1.xlsx'), sheet = 'BGfluo', variables = variables), result)
+  expect_no_error(read_cells(drfile = test_path('fixtures/test1.xlsx'), sheet = 'BGfluo', variables = variables))
+  expect_equal(read_cells(drfile = test_path('fixtures/test1.xlsx'), sheet = 'BGfluo', variables = variables), result)
 })
 
 test_that("reading a table works", {
-  expect_no_error(read_table(drfile = test_path('testdata/test1.xlsx'), sheet = '_data', range = "A101:B111"))
+  expect_no_error(read_table(drfile = test_path('fixtures/test1.xlsx'), sheet = '_data', range = "A101:B111"))
 })
 
 test_that("Function read_data works", {
-  guide <- read_guide(test_path("testdata/guide_competition_1_0.yml"))
-  expect_no_error(read_data(drfile = test_path('testdata/test1.xlsx'), guide = guide))
-  expect_no_error(read_data(drfile = test_path('testdata/test1.xlsx'), guide = test_path("testdata/guide_competition_1_0.yml")))
+  guide <- read_guide(test_path("fixtures/guide_competition_1_0.yml"))
+  expect_no_error(read_data(drfile = test_path('fixtures/test1.xlsx'), guide = guide))
+  expect_no_error(read_data(drfile = test_path('fixtures/test1.xlsx'), guide = test_path("fixtures/guide_competition_1_0.yml")))
 })
 
 
 # test_that("Function read_data using guide with two plates returns two plates", {
-#   guide <- read_guide(test_path("testdata/goodguides/guide_with_two_plates.yaml"))
-#   result <- read_data(drfile = test_path('testdata/test1.xlsx'), guide = guide)
+#   guide <- read_guide(test_path("fixtures/goodguides/guide_with_two_plates.yaml"))
+#   result <- read_data(drfile = test_path('fixtures/test1.xlsx'), guide = guide)
 #   expect_equal(length(result$platedata), 2)
 # })
 
 # test_that("Function read_data using guide split table returns single table", {
-#   # guide <- read_guide(test_path("testdata/goodguides/guide_with_split_table.yml"))
-#   # result <- read_data(drfile = test_path('testdata/test2.xlsx'), guide = guide)
+#   # guide <- read_guide(test_path("fixtures/goodguides/guide_with_split_table.yml"))
+#   # result <- read_data(drfile = test_path('fixtures/test2.xlsx'), guide = guide)
 #   # expect_equal(length(result$table), 1)
 # })
 
 test_that("Version incompatibilities yield an error.", {
-  guide <- read_guide(test_path("testdata/erroneousguides/conflicting_min_version.yml"))
-  expect_error(read_data(drfile = test_path('testdata/test1.xlsx'), guide = guide))
-  guide <- read_guide(test_path("testdata/erroneousguides/conflicting_max_version.yml"))
-  expect_error(read_data(drfile = test_path('testdata/test1.xlsx'), guide = guide))
+  guide <- read_guide(test_path("fixtures/erroneousguides/conflicting_min_version.yml"))
+  expect_error(read_data(drfile = test_path('fixtures/test1.xlsx'), guide = guide))
+  guide <- read_guide(test_path("fixtures/erroneousguides/conflicting_max_version.yml"))
+  expect_error(read_data(drfile = test_path('fixtures/test1.xlsx'), guide = guide))
 })
 
+# testing reading of dates
+test_that("Reading key-value pairs with different atomic classes yields correct results", {
+  excel_file <- test_path("fixtures/test0.xlsx")
+  sheet <- "kvpairs"
+  expect_no_error(
+    result <- read_keyvalue(
+      drfile = excel_file, sheet = sheet, range = "A2:B4",
+      translate = FALSE, atomicclass = c("character","character","numeric")
+    )
+  )
+  expect_equal(result, list(char1 = "9.3", char2 = "test", num1 = 1.2))
+  expect_no_error(
+    result <- read_keyvalue(
+      drfile = excel_file, sheet = sheet, range = "A7:B7",
+      translate = FALSE, atomicclass = c("character")
+    )
+  )
+  expect_equal(
+    result, list(char1 = "9.3")
+  )
+})

tests/testthat/test-read.R

@@ -48,18 +48,21 @@ test_that("function coerce works", {
   expect_equal(coerce("true", "logical"), TRUE)
 })
 
+
+# testing reading of dates
 test_that("Reading dates from an excel file yield correct dates", {
-  excel_file <- test_path("testdata/dates_linux.xlsx")
-  date1 <- readxl::read_excel(excel_file, sheet=1, range="A1:A2")
-  date2 <- readxl::read_excel(excel_file, sheet=1, range="B1:B2")
-  date3 <- readxl::read_excel(excel_file, sheet=1, range="C1:C2")
-  expect_equal(coerce(date1 |> dplyr::pull(1), "date"), as.Date("1962-10-27"))
-  expect_equal(coerce(date2 |> dplyr::pull(1), "date"), as.Date("1962-10-27"))
-  expect_equal(coerce(date3 |> dplyr::pull(1), "date"), as.Date("1962-10-27"))
-  date4 <- readxl::read_excel(excel_file, sheet=1, range="D1:D2")
-  date5 <- readxl::read_excel(excel_file, sheet=1, range="E1:E2")
-  date6 <- readxl::read_excel(excel_file, sheet=1, range="F1:F2")
-  expect_equal(coerce(date4 |> dplyr::pull(1), "date"), as.Date("2025-04-01"))
-  expect_equal(coerce(date5 |> dplyr::pull(1), "date"), as.Date("2025-04-01"))
-  expect_equal(coerce(date6 |> dplyr::pull(1), "date"), as.Date("2025-04-01"))
+  excel_file <- test_path("fixtures/test0.xlsx")
+  sheet <- "dates"
+  date1 <- readxl::read_excel(excel_file, sheet = sheet, col_names = "date", range = "B1:B1")
+  date2 <- readxl::read_excel(excel_file, sheet = sheet, col_names = "date", range = "B2:B2")
+  date3 <- readxl::read_excel(excel_file, sheet = sheet, col_names = "date", range = "B3:B3")
+  expect_equal(coerce(date1 |> dplyr::pull(date), "date"), as.Date("1962-10-27"))
+  expect_equal(coerce(date2 |> dplyr::pull(date), "date"), as.Date("1962-10-27"))
+  expect_equal(coerce(date3 |> dplyr::pull(date), "date"), as.Date("1962-10-27"))
+  date4 <- readxl::read_excel(excel_file, sheet = sheet, col_names = "date", range="B4:B4")
+  date5 <- readxl::read_excel(excel_file, sheet = sheet, col_names = "date", range="B5:B5")
+  date6 <- readxl::read_excel(excel_file, sheet = sheet, col_names = "date", range="B6:B6")
+  expect_equal(coerce(date4 |> dplyr::pull(date), "date"), as.Date("2025-04-01"))
+  expect_equal(coerce(date5 |> dplyr::pull(date), "date"), as.Date("2025-04-01"))
+  expect_equal(coerce(date6 |> dplyr::pull(date), "date"), as.Date("2025-04-01"))
 })