I want to replace for example all instances of "123" with "321" contained within all .txt files in a folder (recursively).
I thought of doing this
sed -i 's/123/321/g' | find . -name \*.txt
but before possibly screwing all my files I would like to ask if it will work.
You have the sed and the find back to front. With GNU sed and the -i option, you could use:
find . -name '*.txt' -type f -exec sed -i s/123/321/g {} +
The find finds files with extension .txt and runs the sed -i command on groups of them (that's the + at the end; it's standard in POSIX 2008, but not all versions of find necessarily support it). In this example substitution, there's no danger of misinterpretation of the s/123/321/g command so I've not enclosed it in quotes. However, for simplicity and general safety, it is probably better to enclose the sed script in single quotes whenever possible.
You could also use xargs (and again using GNU extensions -print0 to find and -0 and -r to xargs):
find . -name '*.txt' -type f -print0 | xargs -0 -r sed -i 's/123/321/g'
The -r means 'do not run if there are no arguments' (so the find doesn't find anything). The -print0 and -0 work in tandem, generating file names ending with the C null byte '\0' instead of a newline, and avoiding misinterpretation of file names containing newlines, blanks and so on.
Note that before running the script on the real data, you can and should test it. Make a dummy directory (I usually call it junk), copy some sample files into the junk directory, change directory into the junk directory, and test your script on those files. Since they're copies, there's no harm done if something goes wrong. And you can simply remove everything in the directory afterwards: rm -fr junk should never cause you anguish.
Related
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'm trying to count the total lines in the files within a directory. To do this I am trying to use a combination of find and wc. However, when I run find . -exec wc -l {}\;, I recieve the error find: missing argument to -exec. I can't see any apparent issues, any ideas?
You simply need a space between {} and \;
find . -exec wc -l {} \;
Note that if there are any sub-directories from the current location, wc will generate an error message for each of them that looks something like that:
wc: ./subdir: Is a directory
To avoid that problem, you may want to tell find to restrict the search to files :
find . -type f -exec wc -l {} \;
Another note: good idea using the -exec option . Too many times people pipe commands together thinking to get the same result, for instance here it would be :
find . -type f | xargs wc -l
The problem with piping commands in such a manner is that it breaks if any files has spaces in it. For instance here if a file name was "a b" , wc would receive "a" and then "b" separately and you would obviously get 2 error messages: a: no such file and b: no such file.
Unless you know for a fact that your file names never have any spaces in them (or non-printable characters), if you do need to pipe commands together, you need to tell all the tools you are piping together to use the NULL character (\0) as a separator instead of a space. So the previous command would become:
find . -type f -print0 | xargs -0 wc -l
With version 4.0 or later of bash, you don't need your find command at all:
shopt -s globstar
wc -l **/*
There's no simple way to skip directories, which as pointed out by Gui Rava you might want to do, unless you can differentiate files and directories by name alone. For example, maybe directories never have . in their name, while all the files have at least one extension:
wc -l **/*.*
I want to find the files not containing a specific string (in a directory and its sub-directories) and remove those files. How I can do this?
The following will work:
find . -type f -print0 | xargs --null grep -Z -L 'my string' | xargs --null rm
This will firstly use find to print the names of all the files in the current directory and any subdirectories. These names are printed with a null terminator rather than the usual newline separator (try piping the output to od -c to see the effect of the -print0 argument.
Then the --null parameter to xargs tells it to accept null-terminated inputs. xargs will then call grep on a list of filenames.
The -Z argument to grep works like the -print0 argument to find, so grep will print out its results null-terminated (which is why the final call to xargs needs a --null option too). The -L argument to grep causes grep to print the filenames of those files on its command line (that xargs has added) which don't match the regular expression:
my string
If you want simple matching without regular expression magic then add the -F option. If you want more powerful regular expressions then give a -E argument. It's a good habit to use single quotes rather than double quotes as this protects you against any shell magic being applied to the string (such as variable substitution)
Finally you call xargs again to get rid of all the files that you've found with the previous calls.
The problem with calling grep directly from the find command with the -exec argument is that grep then gets invoked once per file rather than once for a whole batch of files as xargs does. This is much faster if you have lots of files. Also don't be tempted to do stuff like:
rm $(some command that produces lots of filenames)
It's always better to pass it to xargs as this knows the maximum command-line limits and will call rm multiple times each time with as many arguments as it can.
Note that this solution would have been simpler without the need to cope with files containing white space and new lines.
Alternatively
grep -r -L -Z 'my string' . | xargs --null rm
will work too (and is shorter). The -r argument to grep causes it to read all files in the directory and recursively descend into any subdirectories). Use the find ... approach if you want to do some other tests on the files as well (such as age or permissions).
Note that any of the single letter arguments, with a single dash introducer, can be grouped together (for instance as -rLZ). But note also that find does not use the same conventions and has multi-letter arguments introduced with a single dash. This is for historical reasons and hasn't ever been fixed because it would have broken too many scripts.
GNU grep and bash.
grep -rLZ "$str" . | while IFS= read -rd '' x; do rm "$x"; done
Use a find solution if portability is needed. This is slightly faster.
EDIT: This is how you SHOULD NOT do this! Reason is given here. Thanks to #ormaaj for pointing it out!
find . -type f | grep -v "exclude string" | xargs rm
Note: grep pattern will match against full file path from current directory (see find . -type f output)
One possibility is
find . -type f '!' -exec grep -q "my string" {} \; -exec echo rm {} \;
You can remove the echo if the output of this preview looks correct.
The equivalent with -delete is
find . -type f '!' -exec grep -q "user_id" {} \; -delete
but then you don't get the nice preview option.
To remove files not containing a specific string:
Bash:
To use them, enable the extglob shell option as follows:
shopt -s extglob
And just remove all files that don't have the string "fix":
rm !(*fix*)
If you want to don't delete all the files that don't have the names "fix" and "class":
rm !(*fix*|*class*)
Zsh:
To use them, enable the extended glob zsh shell option as follows:
setopt extended_glob
Remove all files that don't have the string, in this example "fix":
rm -- ^*fix*
If you want to don't delete all the files that don't have the names "fix" and "class":
rm -- ^(*fix*|*class*)
It's possible to use it for extensions, you only need to change the regex: (.zip) , (.doc), etc.
Here are the sources:
https://www.tecmint.com/delete-all-files-in-directory-except-one-few-file-extensions/
https://codeday.me/es/qa/20190819/1296122.html
I can think of a few ways to approach this. Here's one: find and grep to generate a list of files with no match, and then xargs rm them.
find yourdir -type f -exec grep -F -L 'yourstring' '{}' + | xargs -d '\n' rm
This assumes GNU tools (grep -L and xargs -d are non-portable) and of course no filenames with newlines in them. It has the advantage of not running grep and rm once per file, so it'll be reasonably fast. I recommend testing it with "echo" in place of "rm" just to make sure it picks the right files before you unleash the destruction.
This worked for me, you can remove the -f if you're okay with deleting directories.
myString="keepThis"
for x in `find ./`
do if [[ -f $x && ! $x =~ $myString ]]
then rm $x
fi
done
Another solution (although not as fast). The top solution didn't work in my case because the string I needed to use in place of 'my string' has special characters.
find -type f ! -name "*my string*" -exec rm {} \; -print
How do I convert all EOL (dos->unix) of all files in a directory and sub-directories recursively without dos2unix? (I do not have it and cannot install it.)
Is there a way to do it using tr -d '\r' and pipes? If so, how?
For all files in current directory you can do it with a Perl one-liner: perl -pi -e 's/\r\n/\n/g' * (stolen from here)
EDIT: And with a small modification you can do subdirectory recursion:
find | xargs perl -pi -e 's/\r\n/\n/g'
You can use sed's -i flag to change the files in-place:
find . -type f -exec sed -i 's/\x0d//g' {} \+
If I were you, I would keep the files around to make sure the operation went okay. Then you can delete the temporary files when you get done. This can be done like so:
find . -type f -exec sed -i'.OLD' 's/\x0d//g' {} \+
find . -type f -name '*.OLD' -delete
Do you have sane file names and directory names without spaces, etc in them?
If so, it is not too hard. If you've got to deal with arbitrary names containing newlines and spaces, etc, then you have to work harder than this.
tmp=${TMPDIR:-/tmp}/crlf.$$
trap "rm -f $tmp.?; exit 1" 0 1 2 3 13 15
find . -type f -print |
while read name
do
tr -d '\015' < $name > $tmp.1
mv $tmp.1 $name
done
rm -f $tmp.?
trap 0
exit 0
The trap stuff ensures you don't get temporary files left around. There other tricks you can pull, with more random names for your temporary file names. You don't normally need them unless you work in a hostile environment.
You can also use the editor in batch mode.
find . -type f -exec bash -c 'echo -ne "%s/\\\r//\nx\n" | ex "{}" ' \;
If \r isn't followed by \n (maybe the case in files of Tim Pote):
deleting \r (using tr -d) may remove newlines
replacing \r with \n may not cause double / triple newlines
Maybe Tim Pote could verify the points above for the files he mentioned.
This removes carriage returns from all files in the current directory and all subdirectories, and should work on most Unix-like OSs:
grep -lIUre '\r' | xargs sed -i 's/\r//'
If its done in widows:
try to run the command in git bash:
$ find | xargs perl -pi -e 's/\r\n/\n/g'
It can show some Can't do inplace edit: type a message so ignore it
So, How Do I find and copy all files,
*.a
that are in,
~/DIR{1,2,3,...}
to
~/tmp/foo?
Assumed you meant recursively copy everything of type .a from some source location.
Haven't verified yet, but this should do that.
find <root-of-search> -type f -name '*.a' -exec cp {} /tmp/foo \;
replace with the top of wherever you want to search from. You might have to throw quotes around *.a, and you might have to replace escape the ending semicolon by putting it in single quotes rather than back-slashing it.
In a bash shell:
cp ~/DIR*/*.a ~/tmp/foo
find ~/DIR{1,2,...} -name *.a print0 | xargs -i -0 cp '{}' ~/tmp/foo