[R/write] update cell dups

[JanMarvin]

why

This addresses an issue I observed when writing a file at work (unfortunately I could not yet trigger it at home. I've inserted a !duplicated() anyways). What I wanted to do is the following:

# load a file like this
wb <- wb_load("inst/extdata/oxlsx2_sheet.xlsx")

# modify some data
z <- wb_to_df(wb, 1, dims = "B5:I20", na.strings = "N/A", skipEmptyRows = TRUE)
sel <- c("V1_abs", "V2_abs", "V3_abs")
z[sel] <- z[sel] / 10

# clone the original sheet
wb$clone_worksheet(1, "clone")

# add the data back to the cloned sheet
wb$add_data(2, z, startCol = 2, startRow = 5)

# open the file again
wb$open()

notable changes

Previously we:

• wrote <v>NA</v> when writing to a sheet. This is not allowed unless we write to a cell that expects some formula evaluation text output (t=str) or to a inline character. This is now fixed
• used replaceCellStyle = TRUE this is now changed. I want this feature, therefore it is not negotiable. Nah, not really, but any objections?

ideally

We would replace the slow loop here and use something to create two cc and row_attr objects and merge them and only update what is needed from the merge. This would speed things up a lot. Unfortunately I'm to lazy right now to implement this.