Why are file names surrounded with "â" sign?

Question

After adding alias rm='rm -i' to my ~/.bashrc file (because, when I removed a file, it wasn't asking for confirmation), file names are surrounded with "â" signs as in the example below:

rm: cannot remove âfile1.txtâ: No such file or directory

List of aliases:

alias egrep='egrep --color=auto'
alias fgrep='fgrep --color=auto'
alias grep='grep --color=auto'
alias l.='ls -d .* --color=auto'
alias ll='ls -l --color=auto'
alias ls='ls --color=auto'
alias rm='rm -i'
alias vi='vim'
alias which='alias | /usr/bin/which --tty-only --read-alias --show-dot --show-tilde'

Note: I am ssh'ing to my CentOS machine with PuTTY from my Windows machine, so this is definitely a character encoding issue. Using Ubuntu guest in my VM, everything is fine. Smart quotes are showing as they need to.

Solution

These âs are UTF-8 quotes that your current terminal is unable to display properly, being configured in ISO-8859-1 or similar.

You can have a proper display with setting a matching locale or the POSIX one:

$ rm file.txt
rm: cannot remove â  file.txtâ : No such file or directory
$ LC_ALL=en_US.UTF-8 rm file.txt
rm: cannot remove â  file.txtâ : No such file or directory
$ LC_ALL=C rm file.txt
rm: cannot remove 'file.txt' : No such file or directory

---

$ rm foo 2>&1 | od -c
0000000   r   m   :       c   a   n   n   o   t       r   e   m   o   v
0000020   e     342 200 230   f   o   o 342 200 231   :       N   o    
0000040   s   u   c   h       f   i   l   e       o   r       d   i   r
0000060   e   c   t   o   r   y  \n
0000067
$ LC_ALL=C rm foo 2>&1 | od -c
0000000   r   m   :       c   a   n   n   o   t       r   e   m   o   v
0000020   e       '   f   o   o   '   :       N   o       s   u   c   h
0000040       f   i   l   e       o   r       d   i   r   e   c   t   o
0000060   r   y  \n
0000063

Comments

[ott--]

Is there another alias for rm already? Can you show all of them? Do you also have completion enabled?

[smw]

Is there actually a file called file1.txt? That looks like the normal response from rm for a non-existent file when there is a character set translation issue (a terminal set up for ISO-8859-1 rather than UTF-8 for example)

[Alan Kis]

@steeldriver, same for existing and non-existing file.

[Alan Kis]

@ott-- see edit.

[Admin]

How did you get the first rm to generate UTF-8 quotes? I'm having trouble duplicating that. I can't tell if my locale is off, or if I've got xterm doing something right, which is wrong for getting an erroneous condition.

[jlliagre]

GNU rm generates UTF-8 opening and closing quotes when it is running under an UTF-8 locale (see my updated answer). This is by design. There is nothing specific to configure.

[Admin]

Thanks. I'm not sure that all the steps I took to duplicate are necessary but... 1. Use PuTTY, setting "Translation" to ISO-8859-1. ssh to linux machine. 2. Unset environment variables LC_ALL, TERM, TERMINAL, XTERM_LOCAL. 3. set environment variable LAN=en_US.UTF-8. I finally got UTF-8 left- and right-single quotes, 4. Run dash shell. I got ASCII single-quote bytes for all the other combos of xterm, LC_ALL, LANG, TERM, TERMINAL, etc that I tried. I also got the a-circumflex-file1.txt-a-circumflex as the question has: no visible characters after the a-circumflex as in your example.

[jlliagre]

@BruceEdiger My tests were done using with a Solaris installation configured to use UTF-8 and in a Solaris text console, i.e. outside any graphic environment. The Solaris console doesn't support UTF-8.

[Alan Kis]

@jlliagre, It is possible that this is character encoding problem. Most important, I forgo to mention that I am usually ssh with Putty from my Win machine.

[jlliagre]

s/possible/certain/ !