Update export.R

Refactored export_ris function to enhance flexibility. The function now accepts a user-defined mapping of data frame columns to RIS fields. Replaced specific field parameters with a general 'user_mapping' parameter. Added user confirmation step before proceeding with the export.

R/export.R

@@ -54,61 +54,61 @@ export_csv <- function(citations, filename = "citations.csv", separate = NULL, t
 }
 
 
-#' Export deduplicated citations to RIS file
+#' Export data frame to RIS file
 #'
-#' This function saves deduplicated citations as a RIS file with sources, labels and strings
-#' included as custom fields (if they were initially provided for any of the citations). Note that
+#' This function saves a data frame as a RIS file with specified columns mapped to RIS fields. Note that
 #' *existing files are overwritten without warning.*
 #'
-#' @param citations Dataframe with unique citations, resulting from `dedup_citations()`
+#' @param citations Dataframe to be exported to RIS file
 #' @param filename Name (and path) of file, should end in .ris
-#' @param source_field Character. Which RIS field should cite_sources be saved to? NULL to exclude.
-#' @param label_field Character. Which RIS field should cite_labels be saved to? NULL to exclude.
-#' @param string_field Character. Which RIS field should cite_strings be saved to? NULL to exclude.
+#' @param user_mapping List. Custom mapping of RIS fields to data frame columns. If NULL, a default mapping is used.
 #' @export
 #' @examples
 #' if (interactive()) {
 #'   # Load example data from the package
 #'   examplecitations_path <- system.file("extdata", "examplecitations.rds", package = "CiteSource")
 #'   examplecitations <- readRDS(examplecitations_path)
 #'   dedup_results <- dedup_citations(examplecitations, merge_citations = TRUE)
-#'   export_ris(dedup_results$unique, "cite_sources.ris", string_field = NULL)
+#'   export_ris(dedup_results$unique, "cite_sources.ris", user_mapping = list("DB" = "cite_source_include", "C7" = "cite_label_include"))
 #' }
 
-export_ris <- function(citations, filename = "citations.ris", source_field = "DB", label_field = "C7", string_field = "C8") {
+export_ris <- function(x, filename = "output.ris", user_mapping = NULL) {
 
   if (tolower(tools::file_ext(filename)) != "ris") warning("Function saves a RIS file, so filename should (usually) end in .ris. For now, name is used as provided.")
 
-  if (!is.null(source_field) && source_field %in% names(citations)) {
-    citations <- citations %>% dplyr::rename(cite_source_include = .data[[source_field]])
-  }
+  # Default mapping of RIS fields to x columns
+  default_mapping <- list(
+    "DB" = "cite_source_include",
+    "C7" = "cite_label_include",
+    "C8" = "cite_string_include",
+    "C1" = "duplicate_id",
+    "C2" = "record_ids"
+  )
 
-  if (!is.null(label_field) && label_field %in% names(citations)) {
-    citations <- citations %>% dplyr::rename(cite_label_include = .data[[label_field]])
+  # If user_mapping is provided, override the default mapping
+  if (!is.null(user_mapping)) {
+    for (field in names(user_mapping)) {
+      default_mapping[[field]] <- user_mapping[[field]]
+    }
   }
 
-  if (!is.null(string_field) && string_field %in% names(citations)) {
-    citations <- citations %>% dplyr::rename(cite_string_include = .data[[string_field]])
+  # Rename the x columns according to the final mapping
+  for (field in names(default_mapping)) {
+    if (default_mapping[[field]] %in% names(x)) {
+      x <- x %>% dplyr::rename(!!field := .data[[default_mapping[[field]]]])
+    }
   }
 
-  select_cols <- c("source_type", "duplicate_id") %in% names(citations)
-  citations <- citations %>%
-    dplyr::select(c("source_type", "duplicate_id")[select_cols], dplyr::everything(), -tidyselect::any_of(c("cite_source", "cite_string", "cite_label", "record_id")))
-
-  synthesisr_codes <- dplyr::bind_rows(
-    tibble::tribble(
-      ~code, ~field, ~ris_synthesisr,
-      source_field, "cite_source_include", TRUE,
-      string_field, "cite_string_include", TRUE,
-      label_field, "cite_label_include", TRUE,
-      "C1", "duplicate_id", TRUE,
-      "C2", "record_ids", TRUE
-    ),
-    synthesisr_code_lookup %>% dplyr::filter(.data$ris_synthesisr)
-  ) %>% dplyr::distinct(.data$code, .keep_all = TRUE) # Remove fields from synthesisr specification used for CiteSource metadata
+  # Print the final mapping and ask the user if they want to proceed
+  print(default_mapping)
+  proceed <- readline("Do you want to proceed with this mapping? (yes/no): ")
 
-  # Currently, write_refs does not accept tibbles, thus converted
-  write_refs(as.data.frame(citations), file = filename, tag_naming = synthesisr_codes)
+  if (tolower(proceed) == "yes") {
+    # Currently, write_refs does not accept tibbles, thus converted
+    write_refs(as.data.frame(x), file = filename)
+  } else {
+    message("Export cancelled. Please modify your user_mapping argument and try again.")
+  }
 }