Fixes to the date column formatting.

R/imos_to_otn_receivers.R

@@ -43,7 +43,7 @@ imos_to_otn_receivers <- function(rcvr_dataframe) {
     rename(
       OTN_ARRAY = receiver_project_name,
       STATION_NO = station_name,
-      DEPLOY_DATE_TIME = receiver_deployment_datetime,
+      `DEPLOY_DATE_TIME (yyyy-mm-ddThh:mm:ss)` = receiver_deployment_datetime,
       DEPLOY_LAT = receiver_deployment_latitude,
       DEPLOY_LONG = receiver_deployment_longitude,
       INSTRUMENT_DEPTH = depth_below_surface,
@@ -56,6 +56,23 @@ imos_to_otn_receivers <- function(rcvr_dataframe) {
       cols = receiver_name,
       delim = "-",
       names = c("INS_MODEL_NO", "INS_SERIAL_NO")
+    ) 
+
+  #We have to do some cleaning up of date formats in the appropriate columns.
+  #First we'll hit them with Lubridate to make sure they're real dates, formatted correctly.
+  #Then, since they need a "T" in there, we'll cast them back to character strings and slot a 'T'
+  #in where once there was whitespace. Simple! I don't love calling multiple near-identical mutates, but I also
+  #don't like how unreadable and cursed it is to do the requisite jiggery-pokery to get multiple functions into a single
+  #mutate(across())
+  rcvr_return <- rcvr_return %>%
+    mutate(
+        across(ends_with("_DATE_TIME"), lubridate::ymd_hms)
+    ) %>%
+    mutate(
+      across(ends_with("_DATE_TIME"), as.character)
+    ) %>%
+    mutate(
+      across(ends_with("_DATE_TIME"), ~stringr::str_replace(.x, " ", "T"))
     )
 
   # Have to do a little extra manipulation on the dataframe to give "RECOVERED"