I got a tar file, after extracting, there are many files naming like
a
b\c
d\e\f
g\h
I want to correct their name into files in sub-directories like
a
b/c
d/e/f
g/h
I face a problem when a variable contains backslash, it will change the original file name. I want to write a script to rename them.
Parameter expansion is the way to go. You have everything you need in bash, no need to use external tools like find.
$ touch a\\b c\\d\\e
$ ls -l
total 0
-rw-r--r-- 1 ghoti staff 0 11 Jun 23:13 a\b
-rw-r--r-- 1 ghoti staff 0 11 Jun 23:13 c\d\e
$ for file in *\\*; do
> target="${file//\\//}"; mkdir -p "${target%/*}"; mv -v "$file" "$target"; done
a\b -> a/b
c\d\e -> c/d/e
The for loop breaks out as follows:
for file in *\\*; do - select all files whose names contain backslashes
target="${file//\\//}"; - swap backslashes for forward slashes
mkdir -p "${target%/*}"; - create the target directory by stripping the filename from $target
mv -v "$file" "$target"; - move the file to its new home
done - end the loop.
The only tricky bit here I think is the second line: ${file//\\//} is an expression of ${var//pattern/replacement}, where the pattern is an escaped backslash (\\) and the replacement is a single forward slash.
Have a look at man bash and search for "Parameter Expansion" to learn more about this.
Alternately, if you really want to use find, you can still take advantage of bash's Parameter Expansion:
find . -name '*\\*' -type f \
-exec bash -c 't="${0//\\//}"; mkdir -p "${t%/*}"; mv -v "$0" "$t"' {} \;
This uses find to identify each file and process it with an -exec option that basically does the same thing as the for loop above. One significant difference here is that find will traverse subdirectories (limited by the -maxdepth option), so ... be careful.
Renaming a file with backslashes is simple: mv 'a\b' 'newname' (just quote it), but you'll need more than that.
You need to:
find all files with a backslash (e.g. a\b\c)
split path from filename (e.g. a\b from c)
create a complete path (e.g. a/b, dir b under dir a)
move the old file under a new name, under a created path (e.g. rename a\b\c to file named c in dir a/b)
Something like this:
#!/bin/bash
find . -name '*\\*' | while read f; do
base="${f%\\*}"
file="${f##*\\}"
path="${base//\\//}"
mkdir -p "$path"
mv "$f" "$path/$file"
done
(Edit: correct handling of filenames with spaces)
Related
I'm trying to learn bash on Linux, just for fun. I thought it would be pretty useful to have a .sh that would group together similar files. For example, let's say we have the directory
/home/docs/
Inside the directory we have /mathdocs/, /codingdocs/, etc.
Inside those sub-directories we have doc.txt, in all of them. Same name for all the files on the subdirectories.
Let's say I want to group them together, and I want to move all the files to /home/allthedocs/ and rename them after the directories they were in. (mathdocs.txt, codingdocs.txt, etc.)
How could I do that?
I've tried to create a script based on the ls and cp commmands, but I don't know how I can take the name of the directories to rename the files in it after I moved them. I guess it has to be some sort of iterative sentence (for X on Y directories) but I don't know how to do it.
You can move and rename your file in one shot with mv, with a loop that grabs all your files through a glob:
#!/bin/bash
dest_dir=/home/allthedocs
cd /home/docs
for file in */doc.txt; do
[[ -f "$file" ]] || continue # skip if not a regular file
dir="${file%/*}" # get the dir name from path
mv "$file" "$dest_dir/$dir.txt"
done
See this post for more info:
Copying files from multiple directories into a single destination directory
Here is a one liner solution that treats whitespaces in filenames, just as #codeforester 's solution does with the glob.
Note that white spaces are treated with the "-print0" option passed to "find", the internal field separator (IFS) in while loop and the wrapping of file3 variable with quotes.
The parameter substitution from file2 into file3 gets rid of the leading "./".
The parameter substition inside the move command turns the path into a filename (run under /home/docs/):
find . -maxdepth 2 -mindepth 2 -print0 | while IFS= read -r -d '' file; \
do file2=$(printf '%s\n' "$file"); file3=${file2#*\/*}; \
mv "$file2" ../allsil/"${file3//\//}"; done
I have 2 folders: folder_a & folder_b. In each of these folders there are a bunch of files. I am trying to use sed to move all of these files out of these folders and into my current working directory I am currently in.
My folder structure looks like this:
mytest:
a:
1.txt
2.txt
3.txt
b:
4.txt
5.txt
The command I am trying to use is:
find . -type d ! -iname '*.*' # find all folders other than root
| sed -r 's/.*/&\/*/' # add '/*' to each of the arguments
| sed -r 'p;s/.*/./' # output: a/* . b/* .
| xargs -n 2 mv # should be creating two commands: 'mv a/* .' and 'mv b/* .'
Unfortunately I get an error:
mv: cannot stat './aaa/*': No such file or directory
I also get the same error when I try this other strategy (using ls instead of mv):
for dir in */; do
ls $dir;
done;
Even if I use sed to replace the spaces in each directory name with '\ ', or surround the directory names with quotes I get the same error.
I'm not sure if these 2 examples are related in my misunderstanding of bash but they both seem to demonstrate my ignorance of how bash translates the output from one command into the input of another command.
Can anyone shed some light on this?
Update: Completely rewritten.
As #EtanReisner and #melpomene have noted, mv */* . or, more specifically, mv a/* b/* . is the most straightforward solution, but you state that this is in part a learning exercise, so the remainder of the answer shows an efficient find-based solution and explains the problem with the original command.
An efficient find-based solution
Generally, if feasible, it's best and most efficient to let find itself do the work, without involving additional tools; find's -exec action is like a built-in xargs, with {} representing the path at hand (with terminator \;) / all paths (with +):
find . -type f -exec echo mv -t . {} +
To be safe, his will just print the mv commands that would be executed; remove the echo to actually execute them.
This will execute a single[1] mv command to which all matching files are passed, and -t . moves them all to the current dir.
[1] If the resulting command line is too long (which is unlikely), it is split up into multiple commands, just as with xargs.
Operating on files (-type f) bypasses the need for globbing, as find will then enumerate all files for you (it also bypasses the need to exclude . explicitly).
Note that this solution works on entire subtrees, not just (immediate) subdirectories.
It's tempting to consider turning on Bash 4's globstar option and using mv */** ., but that won't work, because it will attempt to move directories as well, not just the files in them.
A caveat re -exec with +: it only works if {} - the placeholder for all paths - is the token immediately before the +.
Since you're on Linux, we can satisfy this condition by specifying the target folder for mv with option -t before the {}; on BSD-based systems such as OSX, you could not do that, because mv doesn't support -t there, so you'd have to use terminator \;, which means that mv is called once for every path, which is obviously much slower.
Why your command didn't work:
As #EtanReisner points out in a comment, xargs invokes the command specified without (implicitly) involving a shell, so globbing won't work; you can verify this with the following command:
echo '*' | xargs echo # -> '*' - NO globbing
If we leave the globbing issue aside, additional work would have been necessary to make your xargs command work correctly with folder names with embedded spaces (or other shell metacharacters):
find . -mindepth 1 -type d |
sed -r "s/.*/'&'\/* ./" | # -> '<input-path>'/* . (including single-quotes)
xargs -n 2 echo mv # NOTE: still won't work due to lack of globbing
Note how the (combined) sed command now produces a single output line '<input-path>'/* ., with the input path enclosed in embedded single-quotes, which is required for xargs to recognize <input-path> as a single argument, even if it contains embedded spaces.
(If your filenames contain single-quotes, you'd have to do more work; also note that since now all arguments for a given dir. are on a single line, you could use xargs -L 1 ....)
Also note how -mindepth 1 (only process paths at the subdirectory level or below) is used to skip processing of . itself.
The only way to make globbing happen is to get the shell involved:
find . -mindepth 1 -type d |
sed -r "s/.*/'&'\/* ./" | # -> '<input-path>'/* . (including single-quotes)
xargs -I {} sh -c 'echo mv {}' # works, but is inefficient
Note the use of xargs' -I option to treat each input line as its own argument ({} is a self-chosen placeholder for the input).
sh -c invokes the (default) shell to execute the resulting command, at which globbing does happen.
However, overall, this is quite inefficient:
A pipeline with 3 segments is used.
A shell instance is invoked for every input path, which in turn calls the mv utility.
Compare this to the efficient find-only solution above, which (typically) creates only 2 processes in total.
I want to copy recursively files to its parent folder for a specific file extension. For example:
./folderA/folder1/*.txt to ./folderA/*.txt
./folderB/folder2/*.txt to ./folderB/*.txt
etc.
I checked cp and find commands but couldn't get it working.
I suspect that while you say copy, you actually mean to move the files up to their respective parent directories. It can be done easily using find:
$ find . -name '*.txt' -type f -execdir mv -n '{}' ../ \;
The above command recurses into the current directory . and then applies the following cascade of conditionals to each item found:
-name '*.txt' will filter out only files that have the .txt extension
-type f will filter out only regular files (eg, not directories that – for whatever reason – happen to have a name ending in .txt)
-execdir mv -n '{}' ../ \; executes the command mv -n '{}' ../ in the containing directory where the {} is a placeholder for the matched file's name and the single quotes are needed to stop the shell from interpreting the curly braces. The ; terminates the command and again has to be escaped from the shell interpreting it.
I have passed the -n flag to the mv program to avoid accidentally overwriting an existing file.
The above command will transform the following file system tree
dir1/
dir11/
file3.txt
file4.txt
dir12/
file2.txt
dir2/
dir21/
file6.dat
dir22/
dir221/
dir221/file8.txt
file7.txt
file5.txt
dir3/
file9.dat
file1.txt
into this one:
dir1/
dir11/
dir12/
file3.txt
file4.txt
dir2/
dir21/
file6.dat
dir22/
dir221/
file8.txt
file7.txt
dir3/
file9.dat
file2.txt
file5.txt
To get rid of the empty directories, run
$ find . -type d -empty -delete
Again, this command will traverse the current directory . and then apply the following:
-type d this time filters out only directories
-empty filters out only those that are empty
-delete deletes them.
Fine print: -execdir is not specified by POSIX, though major implementations (at least the GNU and BSD one) support it. If you need strict POSIX compliance, you'll have to make do with the less safe -exec which would need additional thought to be applied correctly in this case.
Finally, please try your commands in a test directory with dummy files, not your actual data. Especially with the -delete option of find, you can loose all your data quicker than you might imaging. Read the man page and, if that is not enough, the reference manual of find. Never blindly copy shell commands from random strangers posted on the internet if you don't understand them.
$cp ./folderA/folder1/*.txt ./folderA
Try this commnad
Run something like this from the root(ish) directory:
#! /bin/bash
BASE_DIR=./
new_dir() {
LOC_DIR=`pwd`
for i in "${LOC_DIR}"/*; do
[[ -f "${i}" ]] && cp "${i}" ../
[[ -d "${i}" ]] && cd "${i}" && new_dir
cd ..
done
return 0
}
new_dir
This will search each directory. When a file is encountered, it copies the file up a directory. When a directory is found, it will move down into the directory and start the process over again. I think it'll work for you.
Good luck.
I would like to know whether rm can remove all files within a directory (but not the subfolders or files within the subfolders)?
I know some people use:
rm -f /direcname/*.*
but this assumes the filename has an extension which not all do (I want all files - with or without an extension to be removed).
Although find allows you to delete files using -exec rm {} \; you can use
find /direcname -maxdepth 1 -type f -delete
and it is faster. Using -delete implies the -depth option, which means process directory contents before directory.
find /direcname -maxdepth 1 -type f -exec rm {} \;
Explanation:
find searches for files and directories within /direcname
-maxdepth restricts it to looking for files and directories that are direct children of /direcname
-type f restricts the search to files
-exec rm {} \; runs the command rm {} for each file (after substituting the file's path in place of {}).
I would like to know whether rm can remove all files within a directory (but not the subfolders or files within the subfolders)?
That's easy:
$ rm folder/*
Without the -r, the rm command won't touch sub-directories or the files they contain. This will only remove the files in folder and not the sub-directories or their files.
You will see errors telling you that folder/foo is a directory can cannot be removed, but that's actually okay with you. If you want to eliminate these messages, just redirect STDERR:
$ rm folder/* 2> /dev/null
By the way, the exit status of the rm command may not be zero, so you can't check rm for errors. If that's important, you'll have to loop:
$ for file in *
> do
> [[ -f $file ]] && rm $file
> [[ $? -ne 0 ]] && echo "Error in removing file '$file'"
> done
This should work in BASH even if the file names have spaces in them.
You can use
find /direcname -maxdepth 1 -type f -exec rm -f {} \;
A shell solution (without the non-standard find -maxdepth) would be
for file in .* *; do
test -f "$file" && rm "$file"
done
Some shells, notably zsh and perhaps bash version 4 (but not version 3), have a syntax to do that.
With zsh you might just type
rm /dir/path/*(.)
and if you would want to remove any file whose name starts with foo, recursively in subdirectories, you could do
rm /dir/path/**/foo*(.)
the double star feature is (with IMHO better interactive completion) in my opinion enough to switch to zsh for interactive shells. YMMV
The dot in parenthesis suffix indicates that you want only files (not symlinks or directories) to be expanded by the zsh shell.
Unix isn't DOS. There is no special "extension" field in a file name. Any characters after a dot are just part of the name and are called the suffix. There can be more than one suffix, for example.tar.gz. The shell glob character * matches across the . character; it is oblivious to suffixes. So the MS-DOS *.* is just * In Unix.
Almost. * does not match files which start with a .. Objects named with a leading dot are, by convention, "hidden". They do not show up in ls either unless you specify -a.
(This means that the . and .. directory entries for "self" and "parent" are considered hidden.)
To match hidden entries also, use .*
The rm command does not remove directories (when not operated recursively with -r).
Try rm <directory> and see. Even if the directory is empty, it will refuse.
So, the way you remove all (non-hidden) files, pipes, devices, sockets and symbolic links from a directory (but leave the subdirectories alone) is in fact:
rm /path/to/directory/*
to also remove the hidden ones which start with .:
rm /path/to/directory/{*,.*}
This syntax is brace expansion. Brace expansion is not pattern matching; it is just a short-hand for generating multiple arguments, in this case:
rm /path/to/directory/* /path/to/directory/.*
this expansion takes place first first and then globbing takes place to generate the names to be deleted.
Note that various solutions posted here have various issues:
find /path/to/directory -type f -delete
# -delete is not Unix standard; GNU find extension
# without -maxdepth 1 this will recurse over all files
# -maxdepth is also a GNU extension
# -type f finds only files; so this neglects to delete symlinks, fifos, etc.
The GNU find solutions have the benefit that they work even if the number of directory entries to be deleted is huge: too large to pass in a single call to rm. Another benefit is that the built-in -delete does not have issues with passing funny path names to an external command.
The portable workaround for the problem of too many directory entries is to list the entries with ls and pipe to xargs:
( cd /path/to/directory ; ls -a | xargs rm -- )
The parentheses mean "do these commands in a sub-process"; this way the effect of the cd is forgotten, which is useful in scripting. ls -a includes the hidden files.
We now include a -- after rm which means "this is the last option; everything else is a non-option argument". This guards us against directory entries whose names are indistinguishable from options. What if a file is called -rf and ends up the first argument? Then you have rm -rf ... which will blow off subdirectories.
The easiest way to do this is to use:
rm *
In order to remove directories, you must specify the option -r
rm -r
so your directories and anything contained in them will not be removed by using
rm *
per the man page for rm, its purpose is to remove files, which is why this works
I have a folder with more than 5000 images, all with JPG extension.
What i want to do, is to add recursively the "thumb_" prefix to all images.
I found a similar question: Rename Files and Directories (Add Prefix) but i only want to add the prefix to files with the JPG extension.
One of possibly solutions:
find . -name '*.jpg' -printf "'%p' '%h/thumb_%f'\n" | xargs -n2 echo mv
Principe: find all needed files, and prepare arguments for the standard mv command.
Notes:
arguments for the mv are surrounded by ' for allowing spaces in filenames.
The drawback is: this will not works with filenames what are containing ' apostrophe itself, like many mp3 files. If you need moving more strange filenames check bellow.
the above command is for dry run (only shows the mv commands with args). For real work remove the echo pretending mv.
ANY filename renaming. In the shell you need a delimiter. The problem is, than the filename (stored in a shell variable) usually can contain the delimiter itself, so:
mv $file $newfile #will fail, if the filename contains space, TAB or newline
mv "$file" "$newfile" #will fail, if the any of the filenames contains "
the correct solution are either:
prepare a filename with a proper escaping
use a scripting language what easuly understands ANY filename
Preparing the correct escaping in bash is possible with it's internal printf and %q formatting directive = print quoted. But this solution is long and boring.
IMHO, the easiest way is using perl and zero padded print0, like next.
find . -name \*.jpg -print0 | perl -MFile::Basename -0nle 'rename $_, dirname($_)."/thumb_".basename($_)'
The above using perl's power to mungle the filenames and finally renames the files.
Beware of filenames with spaces in (the for ... in ... expression trips over those), and be aware that the result of a find . ... will always start with ./ (and hence try to give you names like thumb_./file.JPG which isn't quite correct).
This is therefore not a trivial thing to get right under all circumstances. The expression I've found to work correctly (with spaces, subdirs and all that) is:
find . -iname \*.JPG -exec bash -c 'mv "$1" "`echo $1 | sed \"s/\(.*\)\//\1\/thumb/\"`"' -- '{}' \;
Even that can fall foul of certain names (with quotes in) ...
In OS X 10.8.5, find does not have the -printf option. The port that contained rename seemed to depend upon a WebkitGTK development package that was taking hours to install.
This one line, recursive file rename script worked for me:
find . -iname "*.jpg" -print | while read name; do cur_dir=$(dirname "$name"); cur_file=$(basename "$name"); mv "$name" "$cur_dir/thumb_$cur_file"; done
I was actually renaming CakePHP view files with an 'admin_' prefix, to move them all to an admin section.
You can use that same answer, just use *.jpg, instead of just *.
for file in *.JPG; do mv $file thumb_$file; done
if it's multiple directory levels under the current one:
for file in $(find . -name '*.JPG'); do mv $file $(dirname $file)/thumb_$(basename $file); done
proof:
jcomeau#intrepid:/tmp$ mkdir test test/a test/a/b test/a/b/c
jcomeau#intrepid:/tmp$ touch test/a/A.JPG test/a/b/B.JPG test/a/b/c/C.JPG
jcomeau#intrepid:/tmp$ cd test
jcomeau#intrepid:/tmp/test$ for file in $(find . -name '*.JPG'); do mv $file $(dirname $file)/thumb_$(basename $file); done
jcomeau#intrepid:/tmp/test$ find .
.
./a
./a/b
./a/b/thumb_B.JPG
./a/b/c
./a/b/c/thumb_C.JPG
./a/thumb_A.JPG
jcomeau#intrepid:/tmp/test$
Use rename for this:
rename 's/(\w{1})\.JPG$/thumb_$1\.JPG/' `find . -type f -name *.JPG`
For only jpg files in current folder
for f in `ls *.jpg` ; do mv "$f" "PRE_$f" ; done