tiledb_dim() prints argument 'filter_list' as 'filters'

[cgiachalis]

This is not functional issue but the print of tiledb_dim object is not matching the construction; 'filter_list' argument appears as 'filters'. This is traced back to an internal function `.as_text_dimension().

library(tiledb)
# TileDB R 0.25.0 with TileDB Embedded 2.21.0 on Windows 10 x64 (build 19045).

tiledb_dim(
  name = "a",
  domain = c(1L, 10L),
  tile = 10L,
  type = "INT32",
  filter_list = tiledb_filter_list(c(
    tiledb_filter_set_option(tiledb_filter("ZSTD"), "COMPRESSION_LEVEL", -1)
  )))
#> tiledb_dim(name="a", domain=c(1L,10L), tile=10L, type="INT32", filters=tiledb_filter_list(c(tiledb_filter_set_option(tiledb_filter("ZSTD"),"COMPRESSION_LEVEL",-1))))
#>                                                                      ^----- here

# attr looks OK
tiledb_attr(
  name = "model",
  type = "INT32",
  ncells = 1,
  nullable = FALSE,
  filter_list = tiledb_filter_list(c(
    tiledb_filter_set_option(tiledb_filter("ZSTD"), "COMPRESSION_LEVEL", -1)
  ))
) |> show() # it doesn't print; we're doing manually
#> tiledb_attr(name="model", type="INT32", ncells=1, nullable=FALSE, filter_list=tiledb_filter_list(c(tiledb_filter_set_option(tiledb_filter("ZSTD"),"COMPRESSION_LEVEL",-1))))

TileDB-R/R/Dim.R

Line 129 in 2dd3485

txt <- paste0(txt, "filters=", .as_text_filter_list(fl), ")")

For attributes:

TileDB-R/R/Attribute.R

Line 110 in 2dd3485

if (nfilters(fl) > 0) paste0(", filter_list=", .as_text_filter_list(fl)),

On top of that, a tiny typo was found : ?tiledb_dim and ?tiledb_attr (hint: Contructs -> Constructs)

Thanks

[eddelbuettel]

Thanks! And really appreciate your attentiveness to detail here. Would you want to propose a PR to correct either or both shortcomings?

[cgiachalis]

Sure! It might good chance to address another one I have in my notes (doc related). I will send a PR over the weekend.

[eddelbuettel]

I also have not forgotten about your open (and dated, by my inaction) issue about 0- and 1- indexing. It's a tricky one!

[cgiachalis]

Oh, thanks - I thought about it back then and concluded that needs clear documentation wherever it needs one; for example there is one function that has one argument that accepts (0-based) and another with (1-based). But it's not a deal if it's documented it.

But there is something else that I am about to open soon and is similar to #524.