RE: Unable to match octal character [SOLVED/WORKAROUND]

RE: Unable to match octal character [SOLVED/WORKAROUND]

[Boylan, Ross]

When I use M-x find-file-literally the characters do match octal 203 (and decribe-char produces same resul as shown immediately below).  I was able to search and replace them.

That gets me what I need.  I still wonder a bit what was going on.  It seems strange to have the display should \203 but for a match on \203 to fail.

Thanks, Drew, for your help.

Ross
________________________________________
From: Boylan, Ross
Sent: Wednesday, April 13, 2016 2:14 PM
To: Drew Adams; help-gnu-emacs@gnu.org
Subject: RE: Unable to match octal character

The \203 I insert manually has a much different describe-char than the one from the file.  Here's the manually inserted one:
            position: 212 of 48737 (0%), column: 0
            character: \203 (displayed as \203) (codepoint 131, #o203, #x83)
    preferred charset: unicode (Unicode (ISO10646))
code point in charset: 0x83
               syntax: w        which means: word
             category: l:Latin
             to input: type "C-x 8 RET HEX-CODEPOINT" or "C-x 8 RET NAME"
          buffer code: #xC2 #x83
            file code: #xC2 #x83 (encoded by coding system raw-text-unix)
              display: terminal code #xC2 #x83

Character code properties: customize what to show
  old-name: NO BREAK HERE
  general-category: Cc (Other, Control)
  decomposition: (131) ('')

So, different character set and codepoint.

I examined the file in binary and there are a bunch of hex 83 = octal 203.  I thought maybe the fact that there is a string of such characters was making a difference, since x83 is not a single byte in UTF-8.  But inserting a space after one of the characters doesn't help.  Perhaps this is because emacs has already encoded it.
Ross
________________________________________
From: Drew Adams [drew.adams@oracle.com]
Sent: Wednesday, April 13, 2016 2:07 PM
To: Boylan, Ross; help-gnu-emacs@gnu.org
Subject: RE: Unable to match octal character

> > I have a file with some characters that display in the emacs buffer as
> > \203.  Yet when I search for that, using C-s C-q 203 <ret> I can't match
> > it.  Likewise if I use search and replace.
>
> Works for me.  Do you see the same thing if you start Emacs using
> `emacs -Q' (no init file)?
>
> YES

So you see the same problem even without your init file.
But when you do it in *scratch* (below) you don't see the problem.

> What happens if you do this, starting from emacs -Q:
> 1. In *scratch*, move point into the text somewhere and use
>    `C-q 203 RET' to insert the \203 character.
> 2. `M-<'
> 3. `C-s C-q 2 0 3 RET'
>
> THE SEARCH SUCCEEDS.

Then I guess you will need to explore the difference between
the context where you see the problem, even starting from `emacs -Q',
and the above context (recipe using *scratch*).

Try to provide a step-by-step recipe, which starts from `emacs -Q',
to reproduce the problem.