need to know fixed sboe_id when writing out the parquet files

process/R/fix-sboe_id-missing.R

@@ -1,9 +1,19 @@
 fix_sboe_id_missing <- function(sboe_id, report_id) {
   idx_missing <- which(sboe_id == "No Id")
+
   if (!length(idx_missing)) {
     return(sboe_id)
   }
 
+  # can't fix `sboe_id` if we don't have a `report_id`
+  idx_has_report_id <- which(!is.na(report_id))
+
+  if (!length(idx_has_report_id)) {
+    return(sboe_id)
+  }
+
+  idx_missing <- union(idx_missing, idx_has_report_id)
+
   sboe_id[idx_missing] <- paste0("NOID-", report_id[idx_missing])
   sboe_id
 }

process/R/read_report_file.R

@@ -146,8 +146,25 @@ report_data_set_column_type <- function(data) {
 write_processed_report_export <- function(dir_sboe_id, report_list = tar_read(report_list)) {
   reports <- process_report_export(dir_sboe_id, report_list)
 
-  sboe_id <- report_path_info(dir_sboe_id)$sboe_id
+  info <- report_path_info(dir_sboe_id)
 
+  if (!any(info$sboe_id == "No Id") && length(unique(info$sboe_id)) == 1) {
+    return(write_processed_report_export_parquet(reports, unique(info$sboe_id)))
+  }
+
+  sboe_ids <- map_dfr(reports, select, sboe_id) |> pull(sboe_id) |> unique()
+
+  reports <-
+    sboe_ids |>
+    set_names() |>
+    map(\(id) map(reports, \(d) filter(d, sboe_id == id)))
+
+  map(sboe_ids, \(id) write_processed_report_export_parquet(reports[[id]], id)) |>
+    flatten_chr() |>
+    unname()
+}
+
+write_processed_report_export_parquet <- function(reports, sboe_id) {
   base_dir <- here::here("..", "data")
   sboe_id_param <- sprintf("sboe_id=%s", sboe_id)