Refactor to make reading files easier to test.

R/read_in_files.R

@@ -67,27 +67,17 @@ read_documents_data <- function(project_name,
     if (length(dir(paths$data)) != 0){
       file_list <- dir(paths$data)
       doc_text  <- character()
-      # This is because not all users will be able to install textreadr.
-      if (!requireNamespace("textreadr", quietly = TRUE)){
-        for (i in 1:length(file_list)){
-          doc_text[i] <- readr::read_file(paste0(paths$data, file_list[i]))
-        }
-      } else {
-        for (i in 1:length(dir(paths$data))){
-           doc_text[i] <- textreadr::read_document(
-                                 file.path(paths$data, file_list[i]),
-                                  combine = TRUE)
-        }
-      }
+      doc_text <- read_files(file_list, doc_text, paths$data,
+                             requireNamespace("textreadr", quietly = TRUE))
 
-        data_set <- data.frame( doc_id = seq_along(1:length(file_list)),
+      data_set <- data.frame( doc_id = seq_along(1:length(file_list)),
                           document_text = doc_text,
                           doc_path = file_list,
                           stringsAsFactors = FALSE)
 
-        # validate column names etc here
+        # validate column names here
         actualNames <- names(data_set)
-        expectedNames <- c("doc_id", "document_text", "doc_path") #GOOD
+        expectedNames <- c("doc_id", "document_text", "doc_path")
         if (sum(expectedNames %in% actualNames) != length(data_set)){
           warning("Required variables for documents_data are not present")
         }
@@ -100,6 +90,22 @@ read_documents_data <- function(project_name,
   invisible(TRUE)
 }
 
+read_files <- function(file_list, doc_text, path_data, textreadr_available){
+  # This is because not all users will be able to install textreadr.
+  if (!textreadr_available){
+    for (i in 1:length(file_list)){
+      doc_text[i] <- readr::read_file(file.path(path_data, file_list[i]))
+    }
+  } else {
+    for (i in 1:length(file_list)){
+      doc_text[i] <- textreadr::read_document(
+        file.path(path_data, file_list[i]),
+        combine = TRUE)
+    }
+  }
+  doc_text
+}
+
 #' Create an empty documents data set
 #'
 #' Used to create a codes data frame with no data but that can
@@ -130,7 +136,7 @@ create_empty_docs_file <-function(path){
 #' @param files  file tibble produced by ShinyFiles
 #' @param file_path  Full path to the data set of documents including
 #' trailing slash
-#' @param docs_df_path  Existing data frame of text documents
+#' @param docs_df_path  Path to existing data frame of text documents
 #' @examples
 #' create_qcoder_project(project_name = "my_qcoder_project", sample = TRUE)
 #'
@@ -139,30 +145,19 @@ create_empty_docs_file <-function(path){
 #' @export
 add_new_documents <- function(files, docs_df_path = "", file_path = ""){
         text_df <- readRDS(docs_df_path)
-        file_list <- files[["name"]]
+        file_list <- as.character(files[["name"]])
         old_docs <- text_df[["doc_path"]]
         if (length(intersect(file_list, old_docs)) != 0){
           warning("One or more files are already imported")
           return()
         }
         doc_text  <- character()
-        if (!requireNamespace("textreadr", quietly = TRUE)){
-          for (i in 1:length(file_list)){
-            doc_text[i] <- readr::read_file(paste0(file_path,
-                                                           file_list[i]))
-          }
-        } else {
-          for (i in 1:length(file_list)){
-                       if (length(file_list) == 0){
-                         return()
-                       }
-                       doc_text[i] <- textreadr::read_document(
-                                    paste0(file_path, file_list[i]))
-          }
-        }
+        doc_text <- read_files(file_list, doc_text, file_path,
+                               requireNamespace("textreadr", quietly = TRUE))
+
         ids <- integer(length(file_list))
         new_rows <- data.frame(doc_id = ids, document_text = doc_text,
-                               doc_path = file_list)
+                               doc_path = as.character(file_list))
         text_df <- rbind( text_df, new_rows)
         row_n <- row.names(text_df)
         text_df$doc_id <- ifelse(text_df$doc_id == 0, row_n,
@@ -269,8 +264,6 @@ read_unit_data <- function(data_path = "units/units.csv",
                              data_frame_name = data_frame_name)
 
   if (file.exists(paths[["data"]])){
-        # units_df_path <- file.path(df_path,
-        #                   paste0(data_frame_name, "_", project_name, ".rds" ))
         units <- readr::read_csv(file = paths[["data"]],
                            col_types = "ic" )
 

tests/testthat/test-read_in_files.R

@@ -2,15 +2,17 @@ context("Reading files into data frames for analysis")
 
 test_that("A new file is successfully added to the documents", {
   new_file_name <- "newfile.txt"
-  file_list <- data.frame(name = new_file_name, size = c(NA), type = c(""), datapath = c("NA/newfile6.txt"))
-  data_path <- "./data/"
-  save_path <- paste0(tempdir(), "/rqcoder_documents_my_qcoder_project.rds")
+  file_list <- data.frame(name = new_file_name, size = c(NA), type = c(""),
+                          datapath = c("NA/newfile6.txt"), stringsAsFactors = FALSE)
+  file_path <- file.path(getwd(), "data")
+  data_path <- file.path(getwd(),"data/qcoder_documents_my_qcoder_project.rds")
+  save_path <- paste0(tempdir(), "/qcoder_documents_my_qcoder_project.rds")
   file.copy("./data/qcoder_documents_my_qcoder_project.rds", save_path)
-  add_new_documents(files = file_list, file_path = data_path, docs_df_path = save_path)
+  add_new_documents(files = file_list, docs_df_path = save_path, file_path = file_path)
   new_df <- readRDS(save_path)
   expect_equal(nrow(new_df), 6)
   expect_equal(new_df[6,"doc_path"], new_file_name)
-  unlink(save_path)
+  unlink(save_path, recursive = TRUE)
 })
 
 test_that("A new file with the same name as an existing file generates a warning", {

tests/testthat/test-utils.R

@@ -37,18 +37,21 @@ test_that("Project with sample data is created correctly.", {
 test_that("Updating documents works", {
   # Set up data
   qcoder_documents_my_qcoder_project <- readRDS("./data/qcoder_documents_my_qcoder_project.rds")
-  save_path <- paste0(tempdir(), "/", "qcoder_documents_my_qcoder_project.rds")
+  test_path <- file.path(tempdir(), "test")
+  # Should not be necessary
+  unlink(test_path, recursive = TRUE)
+  dir.create(test_path)
+  save_path <- file.path(test_path, "qcoder_documents_my_qcoder_project.rds")
   # Do the test on a copy of the data.
   saveRDS(qcoder_documents_my_qcoder_project, file = save_path)
   updated_data <- as.character("A B C D")
   do_update_document(updated_data, docs_df_path = save_path, "CoC_Example1_MU.txt")
   input <- readRDS(file = save_path)
   expect_equal(updated_data, input[2, 2])
   # This should be 2 because we started with an empty directory.
-  expect_equal(length(list.files(tempdir())), 2)
+  expect_equal(length(list.files(test_path)), 2)
 
-  unlink(save_path)
-  setwd(basewd)
+  unlink(test_path, recursive = TRUE)
 })
 
 test_that("Converting to HTML works", {