Have only a single space character between elements and their separators?

[IndrajeetPatil]

For example, here I've deliberately added two spaces before "m" and "o" elements, but they are not converted to a single space by {styler}, which is what I would expect. Does that make sense?

observed

styler::style_text(
'c(
  "x", "y", "z",  "m", "n",  "o", "p",
  "c", "d"
)'
)
#> c(
#>   "x", "y", "z",  "m", "n",  "o", "p",
#>   "c", "d"
#> )

^{Created on 2022-01-14 by the reprex package (v2.0.1.9000)}

expected

styler::style_text(
'c(
  "x", "y", "z",  "m", "n",  "o", "p",
  "c", "d"
)'
)
#> c(
#>   "x", "y", "z", "m", "n", "o", "p",
#>   "c", "d"
#> )

[lorenzwalthert]

Thanks @IndrajeetPatil for your perseverance regarding alignment 😄. The reason for the behavior is that the current implementation requires all columns to be aligned for taking code as aligned, and that is the case in your example. Column 1 and 2 are aligned, and for the remainder of the columns, there is always just one instance, and an instance is always aligned with itself. For example for column 3, there is one line in the set we check, and all elements in this set (just one) match by comma. We need to adapt the rules that

For code to be aligned, columns that have only one row must have only one space after the most right comma from the previous column.

Like in the below example, the first column that has one row is column 2. So the previous column is column 1, and for this column, row 3 has the most right comma, so from that position, we add one space and then, column 2 must start.

c(
  "x",     "z",
  "cgjhg",  
)        ^ most right comma

This should also work when things are right-aligned:

c(
  "x",       "z",
  "cgjhg", "thi", "z"
)               ^ most right comma

If you can't follow, it's ok, took me a while to figure out, and I mainly wrote it down for future reference. :D

[IndrajeetPatil]

Wow, thank you so much for resolving this so quickly!

I can only imagine how complex it must be getting to cover all these edge cases.