Incorrect table cell width with Unicode soft-hyphen character

[billdenney]

This may be related to #18, #280, or #296

When I add several Unicode soft-hyphen characters (\u00ad), pandoc.table.return() misaligns the text in the cells after the soft-hyphen. My specific example has the issue in the row names.

library(pander)

d <-
  data.frame(
    a = LETTERS[1:2],
    b = paste(letters, collapse = "")
  )
# insert soft hyphen
rownames(d) <- c("a", "w\u00adx\u00ady\u00adz")

# Second row of table does not align
cat(
  pandoc.table.return(d)
)
#> 
#> ----------------------------------------------
#>    &nbsp;      a               b              
#> ------------- --- ----------------------------
#>     **a**      A   abcdefghijklmnopqrstuvwxyz 
#> 
#>  **w­x­y­z**   B   abcdefghijklmnopqrstuvwxyz 
#> ----------------------------------------------

^{Created on 2023-12-09 with reprex v2.0.2}

The soft-hyphen character is detected as having a width of 1 with nchar(), but it appears to result in a width of zero later in the process. My guess is that difference is the source of the issue.