provide wb_dims()

[JanMarvin]

@olivroy something I quickly typed down in addition to #681, thinking about ggplot2::aes() and how clean this approach is. This PR provides wb_dims() which can take either row/col, rows/cols or objects that can be coerced via as.data.frame().

  library(openxlsx2)
  
  # take data frames and col_names & row_names
  wb_dims(mtcars)
  wb_dims(mtcars, col_names = TRUE, row_names = TRUE)
  
  # take vectors
  wb_dims(letters)
  wb_dims(t(letters))
  
  # or time series
  wb_dims(AirPassengers)
  
  # single values
  wb_dims(1)
  
  # named 
  wb_dims(rows =  1:10, cols = 5:7)
  # wb_dims(rows =  1:10, row = 5:7)
  
  # single cell
  wb_dims(row = 5, col = 7)
  
  # two vectors  
  wb_dims(1:10, LETTERS)
  wb_dims(1:10, 1:26)
  
  # in a wb chain
  wb <- wb_workbook()$
    add_worksheet()$
    add_data(x = mtcars)$
    add_fill(
      dims = wb_dims(mtcars, col_names = TRUE),
      color = wb_color("yellow")
    )

This could be tweaked further, but might provide a simpler helper for users. A name to indicate that it is coming from openxlsx could be wb_dims().

Please let me know what you think!

[Edit:] Updated

[JanMarvin]

wb_dims() might be better, because dims is just an s away from dim and a small typo might quickly cause confusion.

[olivroy]

Oh yes, I like that, especially with the prefix!

With this, there wouldn't be a use-case for rowcol_to_dims() in the exported API?

could inform users of the new syntax.

But I think it's another step forward for a consistent, streamlined API

[JanMarvin]

Yes, this would be a more convenient and possibly simpler to remember function that creates the same output. It's a wrapper around functions.

[olivroy]

Hi @JanMarvin ,

after reviewing, wb_dims() may be enhanced

• provide data_only option maybe?
• For example in a vignette

wb_dims(mtcars)
"A1:K33"
# I propose:
wb_dims(mtcars, data_only = TRUE)
"A2:K33"
# it could be very useful when applying styling to the data.
# Edit2
wb_dims(mtcars, col_names = FALSE)
"A1:K32"
# you have to add
wb_dims(mtcars, col_names = FALSE, start_row = 2)
"A2:K33"
# to give the correct result. Which one do you think is correct?
# it would be good if this worked too
wb_dims(mtcars, cols = 4) 
# but you have to specify to make it work.
wb_dims( cols = 4, rows = seq_len(nrow(mtcars)) 

• Error if only cols or rows is provided.

# now
wb_dims(cols = 4)
"A1"
# I suggest 
wb_dims(cols = 4)
"D1"

I can file a new issue if it's easier.

Edit:

I think that all arguments in ... should be named to avoid problems.

so this should fail

wb_dims(mtcars)

should be replaced by

wb_dims(x = mtcars) # or data = 

Also a personal wish (as I would love to think about the Excel result as little as possible.

Could something like this work?

wb_dims(x = mtcars, cols = "vs")

[JanMarvin]

wb_dims(x = mtcars, cols = "vs")

like wb_dims(mtcars, start_cols = "VS")?

But yes, I though about the same (regarding columns, rownames and data selection). Go for it if you want to improve it. I usually use the cell directly. Open a new issue or a pull request, I currently don't have time for this.