11

Is there a command to display message “Yes” if a particular file exists? No need to give functionality if the file does not exist.

5 Answers 5

25

Use this simple Bash one-liner:

if [ -e FILENAME ] ; then echo Yes ; fi

The -e check evaluates to true if FILENAME exists, no matter what it is (file, directory, link, device, ...).

If you only want to check regular files, use -f instead, as @Arronical said.

5
  • 18
    [ -e FILENAME ] && echo Yes Jun 14, 2016 at 18:11
  • @ZebMcCorkle Correct, that works too. You might post that as separate answer.
    – Byte Commander
    Jun 14, 2016 at 18:28
  • 3
    If FILENAME is a variable expansion, or contains any special characters, then don't forget to quote it. e.g. in a script, [ -e "$1" ] && echo Yes. Jun 15, 2016 at 11:12
  • @PeterCordes if you're using bash you can use [[ and ]] instead of quoting the variable.
    – Holloway
    Jun 15, 2016 at 14:47
  • 1
    @Holloway: yup, but sometimes you need to write portable scripts that only require POSIX sh. That's why I gave a safe version of this instead of downvoting it and upvoting the bash answer. Also note that it's not a problem to quote expansions inside [[ ]], always quoting is a good habit for beginners that don't remember the special rules for inside the [[ ]] operator, or for arithmetic contexts. Jun 15, 2016 at 14:53
10

You can use this simple script:

#!/bin/bash

if [[ -f $1 ]]; then
    echo "Yes"
    exit 0
else
    exit 1
fi

Save it as file-exists.sh. Then, in the Terminal, type chmod +x file-exists.sh.

Use it like: ./file-exists.sh FILE where you replace FILE with the file you want to check, for example:

./file-exists.sh file.txt

If file.txt exists, Yes will be printed to the Terminal, and the program will exit with status 0 (success). If the file does not exist, nothing will be printed and the program will exit with status 1 (failure).

If you're curious why I included the exit command, read on...


What's up with the exit command?

exit causes normal process termination. What this means is, basically: it stops the script. It accepts an optional (numerical) parameter that will be the exit status of the script that called it.

This exit status enables your other scripts to use your file-exists script and is their way of knowing the file exists or not.

A simple example that puts this to use is this script (save it as file-exists-cli.sh):

#!/bin/bash

echo "Enter a filename and I will tell you if it exists or not: "
read FILE
# Run `file-exists.sh` but discard any output because we don't need it in this example
./file-exists.sh $FILE &>> /dev/null
# #? is a special variable that holds the exit status of the previous command
if [[ $? == 0 ]]; then
    echo "$FILE exists"
else
    echo "$FILE does not exist"
fi

Do the usual chmod +x file-exists-cli.sh and then run it: ./file-exists-cli.sh. You'll see something like this:

File exists (exit 0):

➜  ~ ./file-exists-cli.sh 
Enter a filename and I will tell you if it exists or not: 
booleans.py
booleans.py exists

File does not exist (exit 1):

➜  ~ ./file-exists-cli.sh
Enter a filename and I will tell you if it exists or not: 
asdf
asdf does not exist
1
  • 2
    Good, but Thou shalt quote variable references like "$1". If it has white space in it, it will break without quotes and other bad things can happen.
    – Joe
    Jun 15, 2016 at 22:33
9

In the bash shell on the command line.

if [[ -f /path/to/file ]]; then echo "Yes"; fi

This uses the bash conditional operator -f, and is checking whether the file exists and is a regular file. If you want to test for any files including directories and links then use -e.

This is a great resource for bash conditionals.

5
  • 1
    Why the double braces?
    – Byte Commander
    Jun 14, 2016 at 17:05
  • 1
    I'm not a fan of test in it's original format, I know it's overkill, but it's habit of mine.
    – Arronical
    Jun 14, 2016 at 17:07
  • 7
    @ByteCommander: double braces are just generally better than single braces. They're part of the shell syntax (instead of invoking the [ command, which usually calls a builtin.) They correct inconsistencies and dumb behaviors of the single brace version. The [ builtin remains for compatibility, but I don't see any good reason to use it in new bash code. (If you need to write POSIX-compliant scripts that can run on ancient shells or something, that's different.) Jun 14, 2016 at 19:13
  • @kundor: Mostly agreed, though, to be fair, "[ usually calls a builtin" is dubious. type -a [ for me lists "shell builtin", then "/usr/bin/[."
    – wchargin
    Jun 16, 2016 at 0:37
  • 1
    @wchargin: I'm afraid I don't understand what you're getting at. Your type results seem to corroborate what I said, no? Jun 16, 2016 at 4:27
5

The shortest command for doing what you want is:

test -e FILENAME && echo Yes

test -e will test whether the name given exists in the file system. (You can use test -f to restrict to regular files only. See man test for more.)

If the condition given evaluates to true, then test returns a successful exit status (otherwise it returns a failure status). We combine the two commands using &&, which means "execute the next command if the previous command exited with a success status". The next command simply prints Yes on standard output; in the case of an interactive shell, on the terminal.

This avoids the additional textual material of an if statement, yet gives the same result. Using && (or its opposite, ||) to tie commands together works well when only a single command is involved. If you want to do more than execute a single command in response to a single command's exit status, then using the if syntax quickly becomes much more readable.

As already pointed out in other answers, the equivalent if style construct would be:

if test -e FILENAME; then echo Yes; fi

or alternatively:

if test -e FILENAME
then
    echo Yes
fi

For these purposes, [ and test are equivalent, except that [ demands a terminating ].

0
4

To skin this type of question there's multiple ways, and here's another one : use find command with -exec flag . The path to file can be split into two parts find /etc which sets directory and -name FILENAME which specifies filename (duh!) . -maxdepth will keep find working with /etc directory only and won't descend into subdirectories

adminx@L455D:~$ find /etc -maxdepth 1  -name passwd -exec printf "YES\n" \;
YES
adminx@L455D:~$ find /etc -maxdepth 1  -name passwd1 -exec printf "YES\n" \;
adminx@L455D:~$ 

Another way , via stat :

adminx@L455D:~$ stat /etc/passwd1 &>/dev/null && echo YES
adminx@L455D:~$ stat /etc/passwd &>/dev/null && echo YES
YES

And alternatively via python :

>>> import os
>>> if os.stat('/etc/passwd'): 
...    print 'YES'
... 
YES
>>> if os.stat('/etc/passwd1'): 
...    print 'YES'
... 
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
OSError: [Errno 2] No such file or directory: '/etc/passwd1'

Or a short command line alternative :

 python -c "from os.path import exists; print 'Yes' if exists('/etc/fstab') else '' "
0

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .