duplicate name when reading csv from zenodo

[bart1]

I was testing with the german data from the uva zenodo page and extracted this file: https://zenodo.org/records/14711244/files/de.tgz

withr::with_options(list(
  "getRad.vpts_local_path_format"="{substr(radar, 1, 2)}/{radar}/{year}/{radar}_vpts_{year}{month}.csv.gz"),
  getRad::get_vpts("deboo", as.Date("2017-10-1"), source = "/media/bart/data_disk/tmp/"))
#> Called from: getRad::get_vpts("deboo", as.Date("2017-10-1"), source = "/media/bart/data_disk/tmp/")
#> debug: fetched_vpts <- radar_to_name(switch(dplyr::case_when(source == 
#>     "rmi" ~ "rmi", source %in% eval(formals("get_vpts_aloft")$source) ~ 
#>     "aloft", dir.exists(source) ~ "local"), rmi = purrr::map(radar, 
#>     ~get_vpts_rmi(.x, rounded_interval), .purrr_error_call = cl), 
#>     aloft = purrr::map(radar, ~get_vpts_aloft(.x, rounded_interval = rounded_interval, 
#>         source = source), .purrr_error_call = cl), local = get_vpts_local(radar, 
#>         rounded_interval, directory = source)))
#> Error in `purrr::map_chr()`:
#> ℹ In index: 1.
#> ℹ With name: deboo.
#> Caused by error:
#> ! Result must be length 1, not 2.

It seems that this is caused by two radar names in the file:

vroom::vroom("/media/bart/data_disk/tmp/de/deboo/2017/deboo_vpts_201710.csv.gz") |> dplyr::pull("radar") |> table()
#> Rows: 74200 Columns: 26
#> ── Column specification ────────────────────────────────────────────────────────
#> Delimiter: ","
#> chr   (2): radar, source_file
#> dbl  (21): height, u, v, w, ff, dd, sd_vvp, eta, dens, dbz, dbz_all, n, n_db...
#> lgl   (2): gap, vcp
#> dttm  (1): datetime
#> 
#> ℹ Use `spec()` to retrieve the full column specification for this data.
#> ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
#> 
#> 10132 deboo 
#> 53550 20650

It seems this change happened exactly in this month:

vroom::vroom("/media/bart/data_disk/tmp/de/deboo/2017/deboo_vpts_201709.csv.gz", show_col_types = F ) |> dplyr::pull("radar") |> table()
#> 
#> 10132 
#> 71400
vroom::vroom("/media/bart/data_disk/tmp/de/deboo/2017/deboo_vpts_201710.csv.gz", show_col_types = F ) |> dplyr::pull("radar") |> table()
#> 
#> 10132 deboo 
#> 53550 20650
vroom::vroom("/media/bart/data_disk/tmp/de/deboo/2017/deboo_vpts_201711.csv.gz", show_col_types = F ) |> dplyr::pull("radar") |> table()
#> 
#> deboo 
#> 71325

It breaks in this function as there are tow different names:

getRad:::radar_to_name
#> function (vpts_df_list) 
#> {
#>     purrr::set_names(vpts_df_list, purrr::map_chr(vpts_df_list, 
#>         function(df) unique(dplyr::pull(df, .data$radar))))
#> }
#> <bytecode: 0x5db748176680>
#> <environment: namespace:getRad>

@PietrH any suggestion for a good resolution? What name to pick I guess we here have the odim and wmo code both in one csv

[bart1]

ps for other months the local read works fine