Re: [PATCH 2/2] gdbarch.py: Fix indentation in the generated set_gdbarch_* definitions

[Simon Marchi <simark@simark.ca>]

> diff --git a/gdb/gdbarch.py b/gdb/gdbarch.py
> index 7ac7140e1452..ce430318eb49 100755
> --- a/gdb/gdbarch.py
> +++ b/gdb/gdbarch.py
> @@ -479,10 +479,10 @@ with open("gdbarch.c", "w") as f:
>              print(file=f)
>              print("void", file=f)
>              print(f"set_gdbarch_{c.name} (struct gdbarch *gdbarch,", file=f)
> -            print(
> -                f"            {' ' * len(c.name)}  gdbarch_{c.name}_ftype {c.name})",
> -                file=f,
> -            )
> +            total_indentation = len("set_gdbarch_") + len(c.name) + 2
> +            tabs = '\t' * (total_indentation // 8)
> +            spaces = ' ' * (total_indentation % 8)
> +            print(f"{tabs}{spaces}gdbarch_{c.name}_ftype {c.name})", file=f)

For clarity and to factor out the implementation, I think it would be
worth putting the computation of the indentation characters in its own
function (untested):

  def indentation(n_columns):
    return '\' * (n_columns // 8) + ' ' * (n_columns % 8)

It can then be used directly in the format string:

  setter_name = f"set_gdbarch_{c.name}"
  ftype_name = f"gdbarch_{c.name}_ftype"
  print(f"{setter_name} (struct gdbarch *gdbarch,", file=f)
  indent_columns = len(f"{setter_name} (")
  print("f{indentation(indent_columns)}{ftype_name})", file=f)

Simon