Replace certain multibyte characters

Question

In my markdown file (coming from a Windows computer on which I have no control), I have several multi-byte characters.

Some text xx Some text yy Some text

In emacs, they are displayed as \221 and \222 in place of xx and yy respectively. And when I do describe character, I get this:

             position: 11 of 34 (29%), column: 10
            character:  (displayed as ) (codepoint 145, #o221, #x91)
    preferred charset: iso-8859-1 (Latin-1 (ISO/IEC 8859-1))
code point in charset: 0x91
               syntax: w    which means: word
             category: l:Latin
             to input: type "C-x 8 RET 91" or "C-x 8 RET PRIVATE USE ONE"
          buffer code: #xC2 #x91
            file code: #x91 (encoded by coding system iso-latin-1-dos)
              display: no font available

Character code properties: customize what to show
  old-name: PRIVATE USE ONE
  general-category: Cc (Other, Control)
  decomposition: (145) ('')

There are text properties here:
  charset              iso-8859-1
  fontified            t

How do I query-replace (or replace-regexp or replace-string) these characters?

I mean, when it asks me to enter the string to replace, how do I enter \221 or \222, etc.?

Answer 1

Most likely your file is actually not using latin-1, yet your Emacs tries to decode it as if it were using latin-1. Try C-x RET r windows-1252 RET to see if it fixes your problem.

If it does, you'll want to either add a -*- coding: windows-1252 -*- cookie on the first line of the file so Emacs knows how to decode it properly next time, or (the option I recommend) you can save it using another encoding such as utf-8 which has the advantage of being reasonably easy to auto-detect: with C-x RET f utf-8 RET you can set the coding-system to use when saving the file and then C-x C-s will save it using that encoding.