fix missing sboe_id values that are "No Id" in the database

By filling these in with `NOID-{report_id}`. This creates a unique sboe_id
for the committee for the report, otherwise all "No Id" committees would
end up being impossible to differentiate.

process/R/fix-sboe_id-missing.R

@@ -0,0 +1,9 @@
+fix_sboe_id_missing <- function(sboe_id, report_id) {
+  idx_missing <- which(sboe_id == "No Id")
+  if (!length(idx_missing)) {
+    return(sboe_id)
+  }
+
+  sboe_id[idx_missing] <- paste0("NOID-", report_id[idx_missing])
+  sboe_id
+}

process/R/process_report_list.R

@@ -9,5 +9,6 @@ process_report_list <- function(path_report_list) {
       report_id = col_integer()
     )
   ) |>
+    mutate(sboe_id = fix_sboe_id_missing(sboe_id, report_id)) |>
     arrow::write_parquet(out)
 }

process/R/report_path_info.R

@@ -5,10 +5,12 @@ report_path_info <- function(report_path) {
 
   idx_reports <- map_int(x, \(x) which(x == "reports"))
 
-  tibble(
-    sboe_id = map2_vec(x, idx_reports, \(x, i) x[i + 1]),
-    report_id = map2_int(x, idx_reports, \(x, i) as.integer(strsplit(x[i + 3], "_")[[1]][1])),
-  )
+  sboe_id <- map2_vec(x, idx_reports, \(x, i) x[i + 1])
+  report_id <- map2_int(x, idx_reports, \(x, i) as.integer(strsplit(x[i + 3], "_")[[1]][1]))
+
+  sboe_id <- fix_sboe_id_missing(sboe_id, report_id)
+
+  tibble(sboe_id, report_id)
 }
 
 report_info_in_report_list <- function(files, report_list = tar_read(report_list)) {

reports/2023-10-07_organizing/2023-10-07_organizing.R

@@ -1,7 +1,7 @@
 #+ setup
 library(tidyverse)
 library(fs)
-load_all(here::here("process"))
+pkgload::load_all(here::here("process"))
 
 #+ candidates
 john_bell <- "STA-8S285O-C-001"
@@ -57,4 +57,31 @@ cf$report_list |>
 #> * `../../data-raw/reports/STA-O079OC-C-001/all/200919_2021-07-17.txt`: no records
 #> * `../../data-raw/reports/STA-XD82JF-C-001/all/210255_2022-10-24.txt`: no records
 
-# TODO: Identify reports that are missing a cover entry...
+# Identify reports that are missing a cover entry...
+no_records <-
+  cf$report_list |>
+  anti_join(distinct(cf$receipts, sboe_id, report_id)) |>
+  anti_join(distinct(cf$expenditures, sboe_id, report_id)) |>
+  anti_join(distinct(cf$summary, sboe_id, report_id))
+
+missing_cover <-
+  cf$report_list |>
+  anti_join(cf$cover, by = c("sboe_id", "report_id")) |>
+  anti_join(no_records, by = c("sboe_id", "report_id"))
+
+cf$report_list |> semi_join(missing_cover)
+cf$summary |>semi_join(missing_cover)
+
+
+# Some committees have "No Id" as their SBOE ID
+# => I'll make these "NOID-{report_id}"
+unique_committees <- cf$report_list |> distinct(sboe_id, committee_name)
+
+unique_committees |>
+  semi_join(
+    unique_committees |> count(sboe_id) |> filter(n > 1),
+    by = "sboe_id"
+  )
+
+cf$cover |> distinct(sboe_id, committee_name) |> count()
+cf$cover |> distinct(sboe_id) |> count()