Related
What command checks if a directory exists or not within a Bash shell script?
To check if a directory exists:
if [ -d "$DIRECTORY" ]; then
echo "$DIRECTORY does exist."
fi
To check if a directory does not exist:
if [ ! -d "$DIRECTORY" ]; then
echo "$DIRECTORY does not exist."
fi
However, as Jon Ericson points out, subsequent commands may not work as intended if you do not take into account that a symbolic link to a directory will also pass this check.
E.g. running this:
ln -s "$ACTUAL_DIR" "$SYMLINK"
if [ -d "$SYMLINK" ]; then
rmdir "$SYMLINK"
fi
Will produce the error message:
rmdir: failed to remove `symlink': Not a directory
So symbolic links may have to be treated differently, if subsequent commands expect directories:
if [ -d "$LINK_OR_DIR" ]; then
if [ -L "$LINK_OR_DIR" ]; then
# It is a symlink!
# Symbolic link specific commands go here.
rm "$LINK_OR_DIR"
else
# It's a directory!
# Directory command goes here.
rmdir "$LINK_OR_DIR"
fi
fi
Take particular note of the double-quotes used to wrap the variables. The reason for this is explained by 8jean in another answer.
If the variables contain spaces or other unusual characters it will probably cause the script to fail.
Always wrap variables in double quotes when referencing them in a Bash script.
if [ -d "$DIRECTORY" ]; then
# Will enter here if $DIRECTORY exists, even if it contains spaces
fi
Kids these days put spaces and lots of other funny characters in their directory names. (Spaces! Back in my day, we didn't have no fancy spaces!)
One day, one of those kids will run your script with $DIRECTORY set to "My M0viez" and your script will blow up. You don't want that. So use double quotes.
Note the -d test can produce some surprising results:
$ ln -s tmp/ t
$ if [ -d t ]; then rmdir t; fi
rmdir: directory "t": Path component not a directory
File under: "When is a directory not a directory?" The answer: "When it's a symlink to a directory." A slightly more thorough test:
if [ -d t ]; then
if [ -L t ]; then
rm t
else
rmdir t
fi
fi
You can find more information in the Bash manual on Bash conditional expressions and the [ builtin command and the [[ compound commmand.
I find the double-bracket version of test makes writing logic tests more natural:
if [[ -d "${DIRECTORY}" && ! -L "${DIRECTORY}" ]] ; then
echo "It's a bona-fide directory"
fi
Shorter form:
# if $DIR is a directory, then print yes
[ -d "$DIR" ] && echo "Yes"
A simple script to test if a directory or file is present or not:
if [ -d /home/ram/dir ] # For file "if [ -f /home/rama/file ]"
then
echo "dir present"
else
echo "dir not present"
fi
A simple script to check whether the directory is present or not:
mkdir tempdir # If you want to check file use touch instead of mkdir
ret=$?
if [ "$ret" == "0" ]
then
echo "dir present"
else
echo "dir not present"
fi
The above scripts will check if the directory is present or not
$? if the last command is a success it returns "0", else a non-zero value.
Suppose tempdir is already present. Then mkdir tempdir will give an error like below:
mkdir: cannot create directory ‘tempdir’: File exists
To check if a directory exists you can use a simple if structure like this:
if [ -d directory/path to a directory ] ; then
# Things to do
else #if needed #also: elif [new condition]
# Things to do
fi
You can also do it in the negative:
if [ ! -d directory/path to a directory ] ; then
# Things to do when not an existing directory
Note: Be careful. Leave empty spaces on either side of both opening and closing braces.
With the same syntax you can use:
-e: any kind of archive
-f: file
-h: symbolic link
-r: readable file
-w: writable file
-x: executable file
-s: file size greater than zero
You can use test -d (see man test).
-d file True if file exists and is a directory.
For example:
test -d "/etc" && echo Exists || echo Does not exist
Note: The test command is same as conditional expression [ (see: man [), so it's portable across shell scripts.
[ - This is a synonym for the test builtin, but the last argument must, be a literal ], to match the opening [.
For possible options or further help, check:
help [
help test
man test or man [
Or for something completely useless:
[ -d . ] || echo "No"
Here's a very pragmatic idiom:
(cd $dir) || return # Is this a directory,
# and do we have access?
I typically wrap it in a function:
can_use_as_dir() {
(cd ${1:?pathname expected}) || return
}
Or:
assert_dir_access() {
(cd ${1:?pathname expected}) || exit
}
The nice thing about this approach is that I do not have to think of a good error message.
cd will give me a standard one line message to standard error already. It will also give more information than I will be able to provide. By performing the cd inside a subshell ( ... ), the command does not affect the current directory of the caller. If the directory exists, this subshell and the function are just a no-op.
Next is the argument that we pass to cd: ${1:?pathname expected}. This is a more elaborate form of parameter substitution which is explained in more detail below.
Tl;dr: If the string passed into this function is empty, we again exit from the subshell ( ... ) and return from the function with the given error message.
Quoting from the ksh93 man page:
${parameter:?word}
If parameter is set and is non-null then substitute its value;
otherwise, print word and exit from the shell (if not interactive).
If word is omitted then a standard message is printed.
and
If the colon : is omitted from the above expressions, then the
shell only checks whether parameter is set or not.
The phrasing here is peculiar to the shell documentation, as word may refer to any reasonable string, including whitespace.
In this particular case, I know that the standard error message 1: parameter not set is not sufficient, so I zoom in on the type of value that we expect here - the pathname of a directory.
A philosophical note:
The shell is not an object oriented language, so the message says pathname, not directory. At this level, I'd rather keep it simple - the arguments to a function are just strings.
if [ -d "$Directory" -a -w "$Directory" ]
then
#Statements
fi
The above code checks if the directory exists and if it is writable.
More features using find
Check existence of the folder within sub-directories:
found=`find -type d -name "myDirectory"`
if [ -n "$found" ]
then
# The variable 'found' contains the full path where "myDirectory" is.
# It may contain several lines if there are several folders named "myDirectory".
fi
Check existence of one or several folders based on a pattern within the current directory:
found=`find -maxdepth 1 -type d -name "my*"`
if [ -n "$found" ]
then
# The variable 'found' contains the full path where folders "my*" have been found.
fi
Both combinations. In the following example, it checks the existence of the folder in the current directory:
found=`find -maxdepth 1 -type d -name "myDirectory"`
if [ -n "$found" ]
then
# The variable 'found' is not empty => "myDirectory"` exists.
fi
DIRECTORY=/tmp
if [ -d "$DIRECTORY" ]; then
echo "Exists"
fi
Try online
Actually, you should use several tools to get a bulletproof approach:
DIR_PATH=`readlink -f "${the_stuff_you_test}"` # Get rid of symlinks and get abs path
if [[ -d "${DIR_PATH}" ]] ; Then # Now you're testing
echo "It's a dir";
fi
There isn't any need to worry about spaces and special characters as long as you use "${}".
Note that [[]] is not as portable as [], but since most people work with modern versions of Bash (since after all, most people don't even work with command line :-p), the benefit is greater than the trouble.
Have you considered just doing whatever you want to do in the if rather than looking before you leap?
I.e., if you want to check for the existence of a directory before you enter it, try just doing this:
if pushd /path/you/want/to/enter; then
# Commands you want to run in this directory
popd
fi
If the path you give to pushd exists, you'll enter it and it'll exit with 0, which means the then portion of the statement will execute. If it doesn't exist, nothing will happen (other than some output saying the directory doesn't exist, which is probably a helpful side-effect anyways for debugging).
It seems better than this, which requires repeating yourself:
if [ -d /path/you/want/to/enter ]; then
pushd /path/you/want/to/enter
# Commands you want to run in this directory
popd
fi
The same thing works with cd, mv, rm, etc... if you try them on files that don't exist, they'll exit with an error and print a message saying it doesn't exist, and your then block will be skipped. If you try them on files that do exist, the command will execute and exit with a status of 0, allowing your then block to execute.
[[ -d "$DIR" && ! -L "$DIR" ]] && echo "It's a directory and not a symbolic link"
N.B: Quoting variables is a good practice.
Explanation:
-d: check if it's a directory
-L: check if it's a symbolic link
To check more than one directory use this code:
if [ -d "$DIRECTORY1" ] && [ -d "$DIRECTORY2" ] then
# Things to do
fi
Check if the directory exists, else make one:
[ -d "$DIRECTORY" ] || mkdir $DIRECTORY
[ -d ~/Desktop/TEMPORAL/ ] && echo "DIRECTORY EXISTS" || echo "DIRECTORY DOES NOT EXIST"
Using the -e check will check for files and this includes directories.
if [ -e ${FILE_PATH_AND_NAME} ]
then
echo "The file or directory exists."
fi
This answer wrapped up as a shell script
Examples
$ is_dir ~
YES
$ is_dir /tmp
YES
$ is_dir ~/bin
YES
$ mkdir '/tmp/test me'
$ is_dir '/tmp/test me'
YES
$ is_dir /asdf/asdf
NO
# Example of calling it in another script
DIR=~/mydata
if [ $(is_dir $DIR) == "NO" ]
then
echo "Folder doesnt exist: $DIR";
exit;
fi
is_dir
function show_help()
{
IT=$(CAT <<EOF
usage: DIR
output: YES or NO, depending on whether or not the directory exists.
)
echo "$IT"
exit
}
if [ "$1" == "help" ]
then
show_help
fi
if [ -z "$1" ]
then
show_help
fi
DIR=$1
if [ -d $DIR ]; then
echo "YES";
exit;
fi
echo "NO";
As per Jonathan's comment:
If you want to create the directory and it does not exist yet, then the simplest technique is to use mkdir -p which creates the directory — and any missing directories up the path — and does not fail if the directory already exists, so you can do it all at once with:
mkdir -p /some/directory/you/want/to/exist || exit 1
if [ -d "$DIRECTORY" ]; then
# Will enter here if $DIRECTORY exists
fi
This is not completely true...
If you want to go to that directory, you also need to have the execute rights on the directory. Maybe you need to have write rights as well.
Therefore:
if [ -d "$DIRECTORY" ] && [ -x "$DIRECTORY" ] ; then
# ... to go to that directory (even if DIRECTORY is a link)
cd $DIRECTORY
pwd
fi
if [ -d "$DIRECTORY" ] && [ -w "$DIRECTORY" ] ; then
# ... to go to that directory and write something there (even if DIRECTORY is a link)
cd $DIRECTORY
touch foobar
fi
In kind of a ternary form,
[ -d "$directory" ] && echo "exist" || echo "not exist"
And with test:
test -d "$directory" && echo "exist" || echo "not exist"
file="foo"
if [[ -e "$file" ]]; then echo "File Exists"; fi;
The ls command in conjunction with -l (long listing) option returns attributes information about files and directories.
In particular the first character of ls -l output it is usually a d or a - (dash). In case of a d the one listed is a directory for sure.
The following command in just one line will tell you if the given ISDIR variable contains a path to a directory or not:
[[ $(ls -ld "$ISDIR" | cut -c1) == 'd' ]] &&
echo "YES, $ISDIR is a directory." ||
echo "Sorry, $ISDIR is not a directory"
Practical usage:
[claudio#nowhere ~]$ ISDIR="$HOME/Music"
[claudio#nowhere ~]$ ls -ld "$ISDIR"
drwxr-xr-x. 2 claudio claudio 4096 Aug 23 00:02 /home/claudio/Music
[claudio#nowhere ~]$ [[ $(ls -ld "$ISDIR" | cut -c1) == 'd' ]] &&
echo "YES, $ISDIR is a directory." ||
echo "Sorry, $ISDIR is not a directory"
YES, /home/claudio/Music is a directory.
[claudio#nowhere ~]$ touch "empty file.txt"
[claudio#nowhere ~]$ ISDIR="$HOME/empty file.txt"
[claudio#nowhere ~]$ [[ $(ls -ld "$ISDIR" | cut -c1) == 'd' ]] &&
echo "YES, $ISDIR is a directory." ||
echo "Sorry, $ISDIR is not a directoy"
Sorry, /home/claudio/empty file.txt is not a directory
There are great solutions out there, but ultimately every script will fail if you're not in the right directory. So code like this:
if [ -d "$LINK_OR_DIR" ]; then
if [ -L "$LINK_OR_DIR" ]; then
# It is a symlink!
# Symbolic link specific commands go here
rm "$LINK_OR_DIR"
else
# It's a directory!
# Directory command goes here
rmdir "$LINK_OR_DIR"
fi
fi
will execute successfully only if at the moment of execution you're in a directory that has a subdirectory that you happen to check for.
I understand the initial question like this: to verify if a directory exists irrespective of the user's position in the file system. So using the command 'find' might do the trick:
dir=" "
echo "Input directory name to search for:"
read dir
find $HOME -name $dir -type d
This solution is good because it allows the use of wildcards, a useful feature when searching for files/directories. The only problem is that, if the searched directory doesn't exist, the 'find' command will print nothing to standard output (not an elegant solution for my taste) and will have nonetheless a zero exit. Maybe someone could improve on this.
The below find can be used,
find . -type d -name dirname -prune -print
One Liner:
[[ -d $Directory ]] && echo true
(1)
[ -d Piyush_Drv1 ] && echo ""Exists"" || echo "Not Exists"
(2)
[ `find . -type d -name Piyush_Drv1 -print | wc -l` -eq 1 ] && echo Exists || echo "Not Exists"
(3)
[[ -d run_dir && ! -L run_dir ]] && echo Exists || echo "Not Exists"
If an issue is found with one of the approaches provided above:
With the ls command; the cases when a directory does not exists - an error message is shown
[[ `ls -ld SAMPLE_DIR| grep ^d | wc -l` -eq 1 ]] && echo exists || not exists
-ksh: not: not found [No such file or directory]
I use the following pipe as variable in a script:
match=$( ls | grep -i "$search")
this is then used in an if statement:
if [ "$match" ]; then
echo "matches found"
else
echo "no matches found"
fi
what would be an alternative if I did not want to use find?
ShellCheck recommends:
ls /directory/target_file_pattern
but I do not get the syntax right to get the same result.
also I want no output when there are no matches for the if statement to work.
If you just want to tell if there exist any matches with bash you could use the builtin compgen like so:
if compgen -G 'glob_pattern_like_your_grep' >/dev/null; then
echo "matches found"
else
echo "no matches found"
fi
if you want to operate on the files that are matched, find is usually the right tool for the job:
find . -name 'glob_pattern_like_your_grep' -exec 'your command to operate on each file that matches'
the key though is that you have to use glob patterns, not regex type patterns.
If your find supports it, you might be able to match a regex like
find . -regex 'pattern'
and use that in either the if or with -exec
Use find:
match="$(find . -path "*${search}*" -printf "." | wc -c)"
$match will contain the number of matches. You can check it like this:
if [ "${match}" -gt 0 ] ; then
echo "${match} files found"
else
echo "No files found"
fi
I have a script that queries the Twitter API for several queries, and then writes the raw data to a file with the query in the name, plus a timestamp. I'd like to have a script that, given the list of query strings (regexs?) and for all files in a folder, if one of the query strings is a substring in that file, move it to a specific folder. Right now I have just a script with just a few dozen mv commands, but I'd like a simpler and more maintainable version. Here's an example of what I'm doing now:
mv /home/nick/TwitterSearchToDatabase/queries_for_amita/*femin*/home/nick/TwitterSearchToDatabase/queries_for_amita/feminism
mv /home/nick/TwitterSearchToDatabase/queries_for_amita/*patriarchy* /home/nick/TwitterSearchToDatabase/queries_for_amita/feminism
mv /home/nick/TwitterSearchToDatabase/queries_for_amita/*yesallwomen* /home/nick/TwitterSearchToDatabase/queries_for_amita/feminism
mv /home/nick/TwitterSearchToDatabase/queries_for_amita/*womanpower* /home/nick/TwitterSearchToDatabase/queries_for_amita/feminism
I would use a for loop:
for i in femin patriarchy yesallwomen womanpower; do
mv /home/nick/TwitterSearchToDatabase/queries_for_amita/*$i* /home/nick/TwitterSearchToDatabase/queries_for_amita/feminism
done
That way the list is in the first line so it is easy to amend.
I would isolate data (the words to be moved to feminism) and code.
When you have more keywords (feminism and so), you can make files with keywords and check these keywordfiles for the files you are considering to move.
With ${fromdir} where the files come from, ${todir} where you want them and ${keyfiledir} with the keywords, you get something like
for keyfile in ${keyfiledir}/*; do
key="${keyfile##*/}"
find $from -type f | sed 's#.*/##' | while read -r file; do
echo "${file}" | grep -q -f "${keyfiledir}"/"${key}" && mv "${from}"/"${file}" "${to}"/"${key}"
done
done
How does that work? I tested the solution above with the following script.
from=fromdir
to=todir
keyfiledir=keyfiledir
rm -rf ${from} ${to} ${keyfiledir}
mkdir ${from} ${to} ${keyfiledir}
mkdir ${to}/feminism ${to}/so
touch ${from}/yesallwomen ${from}/women ${from}/some_femin ${from}/"help move"
cat <<# > ${keyfiledir}/feminism
femin
patriarchy
yesallwomen
womanpower
#
touch ${from}/yesallwomen ${from}/women ${from}/some_femin
cat <<# > ${keyfiledir}/so
stack
exchange
help
#
test ! -d "${from}" && echo " Wrong dir ${from}" && exit 1
test ! -d "${to}" && echo " Wrong dir ${to}" && exit 1
test ! -d "${keyfiledir}" && echo " Wrong dir ${keyfiledir}" && exit 1
for keyfile in ${keyfiledir}/*; do
key="${keyfile##*/}"
find $from -type f | sed 's#.*/##' | while read -r file; do
echo "${file}" | grep -q -f "${keyfiledir}"/"${key}" && mv "${from}"/"${file}" "${to}"/"${key}"
done
done
echo "Not moved"
ls ${from}
echo "Moved"
ls -R ${to}
A simple combination of mv and egrep should suffice. egrep can take a pattern list from a file (and then you get to use full regexp syntax, not just glob syntax.) Make sure to exclude the name of the target folder.
cd /home/nick/TwitterSearchToDatabase/queries_for_amita
mv $(ls | egrep -f patterns.txt | grep -v '^feminism$') feminism
What's the best way to check if two paths are equal in Bash? For example, given the directory structure
~/
Desktop/
Downloads/ (symlink to ~/Downloads)
Downloads/
photo.png
and assuming that the current directory is the home directory, all of the following would be equivalent:
./ and ~
~/Desktop and /home/you/Desktop
./Downloads and ~/Desktop/Downloads
./Downloads/photo.png and ~/Downloads/photo.png
Is there a native Bash way to do this?
Bash's test commands have a -ef operator for this purpose
if [[ ./ -ef ~ ]]; then ...
if [[ ~/Desktop -ef /home/you/Desktop ]]; then ...
etc...
$ help test | grep -e -ef
FILE1 -ef FILE2 True if file1 is a hard link to file2.
You can use realpath. For example:
realpath ~/file.txt
Result: /home/testing/file.txt
realpath ./file.txt
Result: /home/testing/file.txt
Also take a look at a similar answer here: bash/fish command to print absolute path to a file
Native bash way:
pwd -P returns the physical directory irrespective of symlinks.
cd "$dir1"
real1=$(pwd -P)
cd "$dir2"
real2=$(pwd -P)
# compare real1 with real2
Another way is to use cd -P, which will follow symlinks but leave you in the physical directory:
cd -P "$dir1"
real1=$(pwd)
cd -P "$dir2"
real2=$(pwd)
# compare real1 with real2
If you have coreutils 8.15 or later installed, you have a realpath command that fully normalizes a path. It removes any . and .. components, makes the path absolute, and resolves all symlinks. Thus:
if [ "$(realpath "$path1")" = "$(realpath "$path2")" ]; then
echo "Same!"
fi
Methods based on resolving symlinks fail when there are other factors involved. For example, bind mounts. (Like mount --bind /raid0/var-cache /var/cache
Using find -samefile is a good bet. That will compare filesystem and inode number.
-samefile is a GNU extension. Even on Linux, busybox find probably won't have it. GNU userspace and Linux kernel often go together, but you can have either without the other, and this question is only tagged Linux and bash.)
# params: two paths. returns true if they both refer to the same file
samepath() {
# test if find prints anything
[[ -s "$(find -L "$1" -samefile "$2")" ]] # as the last command inside the {}, its exit status is the function return value
}
e.g. on my system:
$ find /var/tmp/EXP/cache/apt/archives/bash_4.3-14ubuntu1_amd64.deb -samefile /var/cache/apt/archives/bash_4.3-14ubuntu1_amd64.deb
/var/tmp/EXP/cache/apt/archives/bash_4.3-14ubuntu1_amd64.deb
$ stat {/var/tmp/EXP,/var}/cache/apt/archives/bash_4.3-14ubuntu1_amd64.deb
File: ‘/var/tmp/EXP/cache/apt/archives/bash_4.3-14ubuntu1_amd64.deb’
...
Device: 97ch/2428d Inode: 2147747863 Links: 1
...
File: ‘/var/cache/apt/archives/bash_4.3-14ubuntu1_amd64.deb’
Device: 97ch/2428d Inode: 2147747863 Links: 1
You can use find -L for cases where you want to follow symlinks in the final path component:
$ ln -s /var/cache/apt/archives/bash_4.3-14ubuntu1_amd64.deb foo
$ find /var/cache/apt/archives/bash_4.3-14ubuntu1_amd64.deb -samefile foo
$ find -L /var/cache/apt/archives/bash_4.3-14ubuntu1_amd64.deb -samefile foo
/var/cache/apt/archives/bash_4.3-14ubuntu1_amd64.deb
Obviously this works for paths which refer to directories or any type of file, not just regular files. They all have inode numbers.
usage:
$ samepath /var/cache/apt/ /var/tmp/EXP/cache/apt/ && echo true
true
$ ln -sf /var/cache/apt foobar
$ samepath foobar /var/tmp/EXP/cache/apt/ && echo true
true
samepath /var/tmp/EXP/cache/apt/ foobar && echo true
true
samepath foobar && echo true # doesn't return true when find has an error, since the find output is empty.
find: `': No such file or directory
So find -L dereferences symlinks for -samefile, as well as for the list of paths. So either or both can be symlinks.
Is there an easy way I can print the full path of file.txt ?
file.txt = /nfs/an/disks/jj/home/dir/file.txt
The <command>
dir> <command> file.txt
should print
/nfs/an/disks/jj/home/dir/file.txt
Use readlink:
readlink -f file.txt
I suppose you are using Linux.
I found a utility called realpath in coreutils 8.15.
realpath -s file.txt
/data/ail_data/transformed_binaries/coreutils/test_folder_realpath/file.txt
Since the question is about how to get the full/absolute path of a file and not about how to get the target of symlinks, use -s or --no-symlinks which means don't expand symlinks.
As per #styrofoam-fly and #arch-standton comments, realpath alone doesn't check for file existence, to solve this add the e argument: realpath -e file
The following usually does the trick:
echo "$(cd "$(dirname "$1")" && pwd -P)/$(basename "$1")"
I know there's an easier way that this, but darned if I can find it...
jcomeau#intrepid:~$ python -c 'import os; print(os.path.abspath("cat.wav"))'
/home/jcomeau/cat.wav
jcomeau#intrepid:~$ ls $PWD/cat.wav
/home/jcomeau/cat.wav
On Windows:
Holding Shift and right clicking on a file in Windows Explorer gives you an option called Copy as Path.
This will copy the full path of the file to clipboard.
On Linux:
You can use the command realpath yourfile to get the full path of a file as suggested by others.
find $PWD -type f | grep "filename"
or
find $PWD -type f -name "*filename*"
If you are in the same directory as the file:
ls "`pwd`/file.txt"
Replace file.txt with your target filename.
I know that this is an old question now, but just to add to the information here:
The Linux command which can be used to find the filepath of a command file, i.e.
$ which ls
/bin/ls
There are some caveats to this; please see https://www.cyberciti.biz/faq/how-do-i-find-the-path-to-a-command-file/.
You could use the fpn (full path name) script:
% pwd
/Users/adamatan/bins/scripts/fpn
% ls
LICENSE README.md fpn.py
% fpn *
/Users/adamatan/bins/scripts/fpn/LICENSE
/Users/adamatan/bins/scripts/fpn/README.md
/Users/adamatan/bins/scripts/fpn/fpn.py
fpn is not a standard Linux package, but it's a free and open github project and you could set it up in a minute.
Works on Mac, Linux, *nix:
This will give you a quoted csv of all files in the current dir:
ls | xargs -I {} echo "$(pwd -P)/{}" | xargs | sed 's/ /","/g'
The output of this can be easily copied into a python list or any similar data structure.
echo $(cd $(dirname "$1") && pwd -P)/$(basename "$1")
This is explanation of what is going on at #ZeRemz's answer:
This script get relative path as argument "$1"
Then we get dirname part of that path (you can pass either dir or file to this script): dirname "$1"
Then we cd "$(dirname "$1") into this relative dir
&& pwd -P and get absolute path for it. -P option will avoid all symlinks
After that we append basename to absolute path: $(basename "$1")
As final step we echo it
You may use this function. If the file name is given without relative path, then it is assumed to be present in the current working directory:
abspath() { old=`pwd`;new=$(dirname "$1");if [ "$new" != "." ]; then cd $new; fi;file=`pwd`/$(basename "$1");cd $old;echo $file; }
Usage:
$ abspath file.txt
/I/am/in/present/dir/file.txt
Usage with relative path:
$ abspath ../../some/dir/some-file.txt
/I/am/in/some/dir/some-file.txt
With spaces in file name:
$ abspath "../../some/dir/another file.txt"
/I/am/in/some/dir/another file.txt
You can save this in your shell.rc or just put in console
function absolute_path { echo "$PWD/$1"; }
alias ap="absolute_path"
example:
ap somefile.txt
will output
/home/user/somefile.txt
I was surprised no one mentioned located.
If you have the locate package installed, you don't even need to be in the directory with the file of interest.
Say I am looking for the full pathname of a setenv.sh script. This is how to find it.
$ locate setenv.sh
/home/davis/progs/devpost_aws_disaster_response/python/setenv.sh
/home/davis/progs/devpost_aws_disaster_response/webapp/setenv.sh
/home/davis/progs/eb_testy/setenv.sh
Note, it finds three scripts in this case, but if I wanted just one I
would do this:
$ locate *testy*setenv.sh
/home/davis/progs/eb_testy/setenv.sh
This solution uses commands that exist on Ubuntu 22.04, but generally exist on most other Linux distributions, unless they are just to hardcore for s'mores.
The shortest way to get the full path of a file on Linux or Mac is to use the ls command and the PWD environment variable.
<0.o> touch afile
<0.o> pwd
/adir
<0.o> ls $PWD/afile
/adir/afile
You can do the same thing with a directory variable of your own, say d.
<0.o> touch afile
<0.o> d=/adir
<0.o> ls $d/afile
/adir/afile
Notice that without flags ls <FILE> and echo <FILE> are equivalent (for valid names of files in the current directory), so if you're using echo for that, you can use ls instead if you want.
If the situation is reversed, so that you have the full path and want the filename, just use the basename command.
<0.o> touch afile
<0.o> basename $PWD/afile
afile
In a similar scenario, I'm launching a cshell script from some other location. For setting the correct absolute path of the script so that it runs in the designated directory only, I'm using the following code:
set script_dir = `pwd`/`dirname $0`
$0 stores the exact string how the script was executed.
For e.g. if the script was launched like this: $> ../../test/test.csh,
$script_dir will contain /home/abc/sandbox/v1/../../test
For Mac OS X, I replaced the utilities that come with the operating system and replaced them with a newer version of coreutils. This allows you to access tools like readlink -f (for absolute path to files) and realpath (absolute path to directories) on your Mac.
The Homebrew version appends a 'G' (for GNU Tools) in front of the command name -- so the equivalents become greadlink -f FILE and grealpath DIRECTORY.
Instructions for how to install the coreutils/GNU Tools on Mac OS X through Homebrew can be found in this StackExchange arcticle.
NB: The readlink -f and realpath commands should work out of the box for non-Mac Unix users.
I like many of the answers already given, but I have found this really useful, especially within a script to get the full path of a file, including following symlinks and relative references such as . and ..
dirname `readlink -e relative/path/to/file`
Which will return the full path of the file from the root path onwards.
This can be used in a script so that the script knows which path it is running from, which is useful in a repository clone which could be located anywhere on a machine.
basePath=`dirname \`readlink -e $0\``
I can then use the ${basePath} variable in my scripts to directly reference other scripts.
Hope this helps,
Dave
This worked pretty well for me. It doesn't rely on the file system (a pro/con depending on need) so it'll be fast; and, it should be portable to most any *NIX. It does assume the passed string is indeed relative to the PWD and not some other directory.
function abspath () {
echo $1 | awk '\
# Root parent directory refs to the PWD for replacement below
/^\.\.\// { sub("^", "./") } \
# Replace the symbolic PWD refs with the absolute PWD \
/^\.\// { sub("^\.", ENVIRON["PWD"])} \
# Print absolute paths \
/^\// {print} \'
}
This is naive, but I had to make it to be POSIX compliant. Requires permission to cd into the file's directory.
#!/bin/sh
if [ ${#} = 0 ]; then
echo "Error: 0 args. need 1" >&2
exit 1
fi
if [ -d ${1} ]; then
# Directory
base=$( cd ${1}; echo ${PWD##*/} )
dir=$( cd ${1}; echo ${PWD%${base}} )
if [ ${dir} = / ]; then
parentPath=${dir}
else
parentPath=${dir%/}
fi
if [ -z ${base} ] || [ -z ${parentPath} ]; then
if [ -n ${1} ]; then
fullPath=$( cd ${1}; echo ${PWD} )
else
echo "Error: unsupported scenario 1" >&2
exit 1
fi
fi
elif [ ${1%/*} = ${1} ]; then
if [ -f ./${1} ]; then
# File in current directory
base=$( echo ${1##*/} )
parentPath=$( echo ${PWD} )
else
echo "Error: unsupported scenario 2" >&2
exit 1
fi
elif [ -f ${1} ] && [ -d ${1%/*} ]; then
# File in directory
base=$( echo ${1##*/} )
parentPath=$( cd ${1%/*}; echo ${PWD} )
else
echo "Error: not file or directory" >&2
exit 1
fi
if [ ${parentPath} = / ]; then
fullPath=${fullPath:-${parentPath}${base}}
fi
fullPath=${fullPath:-${parentPath}/${base}}
if [ ! -e ${fullPath} ]; then
echo "Error: does not exist" >&2
exit 1
fi
echo ${fullPath}
This works with both Linux and Mac OSX:
echo $(pwd)$/$(ls file.txt)
find / -samefile file.txt -print
Will find all the links to the file with the same inode number as file.txt
adding a -xdev flag will avoid find to cross device boundaries ("mount points"). (But this will probably cause nothing to be found if the find does not start at a directory on the same device as file.txt)
Do note that find can report multiple paths for a single filesystem object, because an Inode can be linked by more than one directory entry, possibly even using different names. For instance:
find /bin -samefile /bin/gunzip -ls
Will output:
12845178 4 -rwxr-xr-x 2 root root 2251 feb 9 2012 /bin/uncompress
12845178 4 -rwxr-xr-x 2 root root 2251 feb 9 2012 /bin/gunzip
Usually:
find `pwd` | grep <filename>
Alternatively, just for the current folder:
find `pwd` -maxdepth 1 | grep <filename>
This will work for both file and folder:
getAbsolutePath(){
[[ -d $1 ]] && { cd "$1"; echo "$(pwd -P)"; } ||
{ cd "$(dirname "$1")" || exit 1; echo "$(pwd -P)/$(basename "$1")"; }
}
Another Linux utility, that does this job:
fname <file>
For Mac OS, if you just want to get the path of a file in the finder, control click the file, and scroll down to "Services" at the bottom. You get many choices, including "copy path" and "copy full path". Clicking on one of these puts the path on the clipboard.
fp () {
PHYS_DIR=`pwd -P`
RESULT=$PHYS_DIR/$1
echo $RESULT | pbcopy
echo $RESULT
}
Copies the text to your clipboard and displays the text on the terminal window.
:)
(I copied some of the code from another stack overflow answer but cannot find that answer anymore)
In Mac OSX, do the following steps:
cd into the directory of the target file.
Type either of the following terminal commands.
Terminal
ls "`pwd`/file.txt"
echo $(pwd)/file.txt
Replace file.txt with your actual file name.
Press Enter
you#you:~/test$ ls
file
you#you:~/test$ path="`pwd`/`ls`"
you#you:~/test$ echo $path
/home/you/test/file
Beside "readlink -f" , another commonly used command:
$find /the/long/path/but/I/can/use/TAB/to/auto/it/to/ -name myfile
/the/long/path/but/I/can/use/TAB/to/auto/it/to/myfile
$
This also give the full path and file name at console
Off-topic: This method just gives relative links, not absolute. The readlink -f command is the right one.