I was wondering whether it is possible to replace a capturing group with same amount of a constant character (e.g. blanks).
I stumbled over this question with two tasks lately:

1. Replacing the first block of blanks in any line of a given file with a block of 'X' of the same length:
   "blabla   blabla blabla      bla" -> "blablaXXXblabla blabla      bla"
   "blablablablabla blabla      bla" -> "blablablablablaXblabla      bla"
   
   The match is easy (Vim Regex): s/^\([^ ]*\)\(  *\)/\1X/
   But what should go instead of the single 'X' ?
   
   
2. Replace the characters between two specific offsets (n-th and m-th character of a line)  with blanks:
   This can be done easily, but it's annoying to hardcode  (m-n) blanks in the replace-expression.
   
   Easy but annoying solution: s/^\(.\{50\}\).\{20\}/\1                    /
   Is there a more elegant way to accomplish this?

 As environment any of the *nix-Engines will do. If I have to pick: Vim-Regex or Perl.

Thanks in advance,
Stefan