Refactor convert_input to Perform tasks via helper function

CHANGELOG.md

@@ -1,4 +1,5 @@
 # Change Log
+
 All notable changes are kept in this file. All changes made should be added to the section called
 `Unreleased`. Once a new release is made this file will be updated to create a new `Unreleased`
 section for the next release.
@@ -9,6 +10,8 @@ For more information about this file see also [Keep a Changelog](http://keepacha
 
 ### Added
 
+- Refactor `convert_input` to Perform tasks via helper function. Subtask of [#3307](https://github.com/PecanProject/pecan/issues/3307)
+
 ### Fixed
 - updated github action to build docker images
 

base/db/R/add.database.entries.R

@@ -0,0 +1,130 @@
+#' Return new arrangement of database while adding code to deal with ensembles
+#'
+#' @param result list of results from the download function
+#' @param con database connection
+#' @param start_date start date of the data
+#' @param end_date end date of the data
+#' @param write whether to write to the database
+#' @param overwrite Logical: If a file already exists, create a fresh copy?
+#' @param insert.new.file whether to insert a new file
+#' @param input.args input arguments obtained from the convert_input function
+#' @param machine machine information
+#' @param mimetype data product specific file format
+#' @param formatname format name of the data
+#' @param allow.conflicting.dates whether to allow conflicting dates
+#' @param ensemble ensemble id
+#' @param ensemble_name ensemble name
+#' @param existing.input existing input records
+#' @param existing.dbfile existing dbfile records
+#' @param input input records
+#' @return list of input and dbfile ids
+#'
+#' @author Betsy Cowdery, Michael Dietze, Ankur Desai, Tony Gardella, Luke Dramko
+
+add.database.entries <- function(
+    result, con, start_date,
+    end_date, overwrite,
+    insert.new.file, input.args,
+    machine, mimetype, formatname,
+    allow.conflicting.dates, ensemble,
+    ensemble_name, existing.input,
+    existing.dbfile, input) {
+    # Setup newinput. This list will contain two variables: a vector of input IDs and a vector of DB IDs for each entry in result.
+    # This list will be returned.
+    newinput <- list(input.id = NULL, dbfile.id = NULL) # Blank vectors are null.
+
+    for (i in 1:length(result)) { # Master for loop
+        id_not_added <- TRUE
+
+        if (!is.null(existing.input) && nrow(existing.input[[i]]) > 0 &&
+            (existing.input[[i]]$start_date != start_date || existing.input[[i]]$end_date != end_date)) {
+            # Updating record with new dates
+            db.query(
+                paste0(
+                    "UPDATE inputs SET start_date='", start_date, "', end_date='", end_date,
+                    "' WHERE id=", existing.input[[i]]$id
+                ),
+                con
+            )
+            id_not_added <- FALSE
+
+            # The overall structure of this loop has been set up so that exactly one input.id and one dbfile.id will be written to newinput every iteration.
+            newinput$input.id <- c(newinput$input.id, existing.input[[i]]$id)
+            newinput$dbfile.id <- c(newinput$dbfile.id, existing.dbfile[[i]]$id)
+        }
+
+        if (overwrite) {
+            # A bit hacky, but need to make sure that all fields are updated to expected values (i.e., what they'd be if convert_input was creating a new record)
+            if (!is.null(existing.input) && nrow(existing.input[[i]]) > 0) {
+                db.query(
+                    paste0(
+                        "UPDATE dbfiles SET file_path='", dirname(result[[i]]$file[1]),
+                        "', file_name='", result[[i]]$dbfile.name[1],
+                        "' WHERE id=", existing.dbfile[[i]]$id
+                    ),
+                    con
+                )
+            }
+
+            if (!is.null(existing.dbfile) && nrow(existing.dbfile[[i]]) > 0) {
+                db.query(paste0(
+                    "UPDATE dbfiles SET file_path='", dirname(result[[i]]$file[1]),
+                    "', file_name='", result[[i]]$dbfile.name[1],
+                    "' WHERE id=", existing.dbfile[[i]]$id
+                ), con)
+            }
+        }
+
+        # If there is no ensemble then for each record there should be one parent
+        # But when you have ensembles, all of the members have one parent !!
+        parent.id <- if (is.numeric(ensemble)) {
+            ifelse(is.null(input[[i]]), NA, input[[1]]$id)
+        } else {
+            ifelse(is.null(input[[i]]), NA, input[[i]]$id)
+        }
+
+
+        if ("newsite" %in% names(input.args) && !is.null(input.args[["newsite"]])) {
+            site.id <- input.args$newsite
+        }
+
+        if (insert.new.file && id_not_added) {
+            dbfile.id <- dbfile.insert(
+                in.path = dirname(result[[i]]$file[1]),
+                in.prefix = result[[i]]$dbfile.name[1],
+                "Input",
+                existing.input[[i]]$id,
+                con,
+                reuse = TRUE,
+                hostname = machine$hostname
+            )
+
+            newinput$input.id <- c(newinput$input.id, existing.input[[i]]$id)
+            newinput$dbfile.id <- c(newinput$dbfile.id, dbfile.id)
+        } else if (id_not_added) {
+            # This is to tell input.insert if we are writing ensembles
+            # Why does it need it? Because it checks for inputs with the same time period, site, and machine
+            # and if it returns something it does not insert anymore, but for ensembles, it needs to bypass this condition
+            ens.flag <- if (!is.null(ensemble) || is.null(ensemble_name)) TRUE else FALSE
+
+            new_entry <- dbfile.input.insert(
+                in.path = dirname(result[[i]]$file[1]),
+                in.prefix = result[[i]]$dbfile.name[1],
+                siteid = site.id,
+                startdate = start_date,
+                enddate = end_date,
+                mimetype = mimetype,
+                formatname = formatname,
+                parentid = parent.id,
+                con = con,
+                hostname = machine$hostname,
+                allow.conflicting.dates = allow.conflicting.dates,
+                ens = ens.flag
+            )
+
+            newinput$input.id <- c(newinput$input.id, new_entry$input.id)
+            newinput$dbfile.id <- c(newinput$dbfile.id, new_entry$dbfile.id)
+        }
+    } # End for loop
+    return(newinput)
+}

base/db/R/check.missing.files.R

@@ -0,0 +1,47 @@
+#' Function to check if result has empty or missing files
+#'
+#' @param result A list of dataframes with file paths
+#' @param outname Name of the output file
+#' @param existing.input Existing input records
+#' @param existing.dbfile Existing dbfile records
+#' @return A list of dataframes with file paths, a list of strings with the output file name, a list of existing input records, and a list of existing dbfile records
+#'
+#' @author Betsy Cowdery, Michael Dietze, Ankur Desai, Tony Gardella, Luke Dramko
+
+check_missing_files <- function(result, outname, existing.input = NULL, existing.dbfile = NULL) {
+  result_sizes <- purrr::map_dfr(
+    result,
+    ~ dplyr::mutate(
+      .,
+      file_size = purrr::map_dbl(file, file.size),
+      missing = is.na(file_size),
+      empty = file_size == 0
+    )
+  )
+
+  if (any(result_sizes$missing) || any(result_sizes$empty)) {
+    log_format_df <- function(df) {
+      formatted_df <- rbind(colnames(df), format(df))
+      formatted_text <- purrr::reduce(formatted_df, paste, sep = " ")
+      paste(formatted_text, collapse = "\n")
+    }
+
+    PEcAn.logger::logger.severe(
+      "Requested Processing produced empty files or Nonexistent files:\n",
+      log_format_df(result_sizes[, c(1, 8, 9, 10)]),
+      "\n Table of results printed above.",
+      wrap = FALSE
+    )
+  }
+
+
+  # Wrap in a list for consistant processing later
+  if (exists("existing.input") && is.data.frame(existing.input)) {
+    existing.input <- list(existing.input)
+  }
+
+  if (exists("existing.dbfile") && is.data.frame(existing.dbfile)) {
+    existing.dbfile <- list(existing.dbfile)
+  }
+  return(list(existing.input, existing.dbfile))
+}