Resolve duplicate data issue in data preparation scripts

[hepplerj]

While working on #146 I've discovered an issue with the weekly data getting duplicate data, which I discovered with some new CONSTRAINT checks in PostgreSQL. I'll want to 1) make sure this isn't present in the DataScribe export first; and then, 2) resolve where the data is duplicated and ensure it's getting addressed properly.

[hepplerj]

A quick check of the data as it loads shows there are no duplicate items. I checked with the following:

# After initial data loads, check each source of weekly bills
check_weekly_source <- function(data, source_name) {
  message(sprintf("\n=== Checking %s Weekly Bills Source Data ===", source_name))
  
  # Check for complete duplicates
  total_rows <- nrow(data)
  distinct_rows <- nrow(distinct(data))
  duplicates <- total_rows - distinct_rows
  
  # Check for duplicates based on key fields
  key_dupes <- data %>%
    group_by(`DataScribe Item #`, Year) %>%
    filter(n() > 1)
  
  message(sprintf("Total rows: %d", total_rows))
  message(sprintf("Distinct rows: %d", distinct_rows))
  message(sprintf("Complete duplicate rows: %d", duplicates))
  message(sprintf("Rows with duplicate keys: %d", nrow(key_dupes)))
  
  if(nrow(key_dupes) > 0) {
    message("\nSample of rows with duplicate keys:")
    print(head(key_dupes, 5))
  }
  
  return(duplicates > 0 || nrow(key_dupes) > 0)
}

# Check each source
check_weekly_source(`1669_1670_Wellcome_weeklybills_parishes`, "Wellcome")
check_weekly_source(Laxton_1700_weeklybills_causes, "Laxton")
check_weekly_source(BLV1_weeklybills_parishes, "Bodleian V1")
check_weekly_source(BLV2_weeklybills_parishes, "Bodleian V2")
check_weekly_source(BLV3_weeklybills_parishes, "Bodleian V3")

Which leads me to think the rbind I do in the weeks data is causing some duplication of records.