How do I use find command with pipe in bash? - bash

The directory structure looks like
home
--dir1_foo
----subdirectory.....
--dir2_foo
--dir3_foo
--dir4_bar
--dir5_bar
I'm trying to use 'find' command to get directories containing specific strings first, (in this case 'foo'), then use 'find' command again to retrieve some directories matching conditions.
So, I first tried
#!/bin/bash
for dir in `find ./ -type d -name "*foo*" `;
do
for subdir in `find $dir -mindepth 2 -type d `;
do
[Do some jobs]
done
done
, and this script works fine.
Then I thought that using only one loop with pipe like below would also work, but this does not work
#!/bin/bash
for dir in `find ./ -type d -name "*foo*" | find -mindepth 2 -type d `;
do
[Do some jobs]
done
and actually this script works the same as
for dir in `find -mindepth 2 -type d`;
do
[Do some jobs]
done
, which means that the first find command is ignored..
What is the problem?

What your script is doing is not a good practice and has lot of potential pitfalls. See BashFAQ- Why you don't read lines with "for" to understand why.
You can use xargs with -0 to read null delimited files and use the another find command without needing to use the for-loop
find ./ -type d -name "*foo*" -print0 | xargs -0 -I{.} find {.} -mindepth 2 -type d
The string following -I in xargs acts like a placeholder for the input received from the previous pipeline and passes it to the next command. The -print0 option is GNU specific which is a safe option to hande filenames/directory names containing spaces or any other shell meta-characters.
So with the above command in-place, if you are interested in doing some action over the output from 2nd command, do a process-substitution syntax with the while command,
while IFS= read -r -d '' f; do
echo "$f"
# Your other actions can be done on "$f" here
done < <(find ./ -type d -name "*foo*" -print0 | xargs -0 -I{.} find {.} -mindepth 2 -type d -print0)
As far the reason why your pipelines using find won't work is that you are not reading the previous find command's output. You needed either xargs or -execdir while the latter is not an option I would recommend.

Related

Find command output to echo without variable assignment, in one line

I'm trying to write one line of code that finds all .sh files in the current directory and its subdirectories, and print them without the .sh extension (preferably without the path too).
I think I got the find command down. I tried using the output of
find . -type f -iname "*.sh" -print
as input for echo, and formatting it along these lines
echo "${find_output%.sh}"
However, I cannot get it to work in one line, without variable assigment.
I got inspiration from this answer on stackoverflow https://stackoverflow.com/a/18639136/15124805
to use this line:
echo "${$( find . -type f -iname "*.sh" -print)%.sh}"
But I get this error:
ash: ${$( find . -type f -iname "*.sh" -print)%.sh}: bad substitution
I also tried using xargs
find . -type f -iname "*.sh" -print |"${xargs%.sh}" echo
But I get a "command not found error" -probably I didn't use xargs correctly, but I'm not sure how I could improve this or if it's the right way to go.
How can I make this work?
That's the classic useless use of echo. You simply want
find . -type f -iname "*.sh" -exec basename {} .sh \;
If you have GNU find, you can also do this with -printf.
However, basename only matches .sh literally, so if you really expect extensions with different variants of capitalization, you need a different approach.
For the record, the syntax you tried to use for xargs would attempt to use the value of a variable named xargs. The correct syntax would be something like
find . -type f -iname "*.sh" -print |
xargs -n 1 sh -c 'echo "${1%.[Ss][Hh]}"' _
but that's obviously rather convoluted. In some more detail, you need sh because the parameter expansion you are trying to use is a feature of the shell, not of echo (or xargs, or etc).
(You can slightly optimize by using a loop:
find . -type f -iname "*.sh" -print |
xargs sh -c 'for f; do
echo "${f%.[Ss][Hh]}"
done' _
but this is still not robust for all file names; see also https://mywiki.wooledge.org/BashFAQ/020 for probably more than you realized you needed to know about this topic. If you have GNU find and GNU xargs, you can use find ... -print0 | xargs -r0)

Shell stop script if find command fails

Good day.
In a script of fine i have the following find command:
find -maxdepth 1 \! -type d -name "some_file_name_*" -name "*.txt" -name "*_${day_month}_*" -exec cp {} /FILES/directory1/directory2/directory3/ +
I want to know how to stop the script if the command does't find anything.
Use GNU xargs with the -r switch and a pipeline to ensure the output of find is passed to cp only if its non-empty.
find -maxdepth 1 \! -type d -name "some_file_name_*" -name "*.txt" -name "*_${day_month}_*" \
| xargs -r I{} cp "{}" /FILES/directory1/directory2/directory3/
I{} is a place-holder for the output from the find command which is passed to cp,
The flags, -r and I{} represent the following according to the man xargs page,
-r, --no-run-if-empty
If the standard input does not contain any nonblanks, do not run
the command. Normally, the command is run once even if there is
no input. This option is a GNU extension.
-I replace-str
Replace occurrences of replace-str in the initial-arguments with
names read from standard input.
You may add -exec false {} so you get a false exit status when something is found (which makes it a bit upside-down though)
if find . -name foo -exec echo ok ';' -exec false {} +
then
echo 'not found'
exit
fi
echo found
See similar question in stackexchange: How to detect whether “find” found any matches?, in particular this answer which suggests the false trick

Using cp in bash to use piped in information about files like modification date

I am trying to copy files from one directory into another from certain modification date ranges. For example, copy all files created after May 10 from dir1 to dir2. I have tried a few things but have been unsuccessful so far.
This made sense to me but cp does not take the filenames piped to it, but just executes ./* and copies all files in the directory:
find . -type f -daystart -mtime 2 | cp ./* /dir/
This almost worked, but did not copy all of the matching files, I also tried xargs -s 50000, but did not work:
find . -type f -daystart -mtime 2 | xargs -I {} cp {} /dir/
find . -type f -daystart -mtime 2 | xargs cp -t /dir/
Found this online, does not work:
cp $(find . -type f -daystart -mtime 2) /dir/
Ideas? Thanks.
Given as your actual question is about using filenames from stdin rather than metadata from stdin, this is quite straightforward:
while IFS= read -r -d '' filename; do
cp "$filename" /wherever
done < <(find . -type f -daystart -mtime 2 -print0)
Note the use of IFS= read -r -d '' and -print0 -- as NUL and / are the only two characters which can't be used in UNIX filenames, using any other character, including the newline, to delimit them is unsafe. Think about what would happen if someone (or a software bug) created a file called $'./ \n/etc/passwd'; you want to be damned sure none of your scripts try to delete or overwrite /etc/passwd when they're trying to delete or overwrite that file.
That said, you don't actually need to use a pipe at all:
find . -type f -daystart -mtime -2 -exec cp '{}' /wherever ';'
...or, if you're only trying to support GNU cp, you can use this more efficient variant:
find . -type f -daystart -mtime -2 -exec cp -t /wherever '{}' +
You don't specify why the various attempts didn't work, so I can only assume that they are the result of whitespace in the filenames.
Try using find's useful -exec action instead of using xargs:
find . -type f -daystart -mtime 2 -exec cp {} /media/alex/Extra/Music/watchfolder/ \;
find . -type f -daystart -mtime 2 \
| cpio -pdv /media/alex/Extra/Music/watchfolder/

Piping find to find

I want to pipe a find result to a new find. What I have is:
find . -iname "2010-06*" -maxdepth 1 -type d | xargs -0 find '{}' -iname "*.jpg"
Expected result: Second find receives a list of folders starting with 2010-06, second find returns a list of jpg's contained within those folders.
Actual result: "find: ./2010-06 New York\n: unknown option"
Oh darn. I have a feeling it concerns the format of the output that the second find receives as input, but my only idea was to suffix -print0 to first find, with no change whatsoever.
Any ideas?
You need 2 things. -print0, and more importantly -I{} on xargs, otherwise the {} doesn't do anything.
find . -iname "2010-06*" -maxdepth 1 -type d -print0 | xargs -0 -I{} find '{}' -iname '*.jpg'
Useless use of xargs.
find 2010-06* -iname "*.jpg"
At least Gnu-find accepts multiple paths to search in. -maxdepth and type -d is implicitly assumed.
How about
find . -iwholename "./2010-06*/*.jpg
etc?
Although you did say that you specifically want this find + pipe problem to work, its inefficient to fork an extra find command. Since you are specifying -maxdepth as 1, you are not traversing subdirectories. So just use a for loop with shell expansion.
for file in *2010-06*/*.jpg
do
echo "$file"
done
If you want to find all jpg files inside each 2010-06* folders recursively, there is also no need to use multiple finds or xargs
for directory in 2010-06*/
do
find $directory -iname "*.jpg" -type f
done
Or just
find 2006-06* -type f -iname "*.jpg"
Or even better, if you have bash 4 and above
shopt -s globstar
shopt -s nullglob
for file in 2010-06*/**/*.jpg
do
echo "$file"
done

How can I list all unique file names without their extensions in bash?

I have a task where I need to move a bunch of files from one directory to another. I need move all files with the same file name (i.e. blah.pdf, blah.txt, blah.html, etc...) at the same time, and I can move a set of these every four minutes. I had a short bash script to just move a single file at a time at these intervals, but the new name requirement is throwing me off.
My old script is:
find ./ -maxdepth 1 -type f | while read line; do mv "$line" ~/target_dir/; echo "$line"; sleep 240; done
For the new script, I basically just need to replace find ./ -maxdepth 1 -type f
with a list of unique file names without their extensions. I can then just replace do mv "$line" ~/target_dir/; with do mv "$line*" ~/target_dir/;.
So, with all of that said. What's a good way to get a unique list of files without their file names with bash script? I was thinking about using a regex to grab file names and then throwing them in a hash to get uniqueness, but I'm hoping there's an easier/better/quicker way. Ideas?
A weird-named files tolerant one-liner could be:
find . -maxdepth 1 -type f -and -iname 'blah*' -print0 | xargs -0 -I {} mv {} ~/target/dir
If the files can start with multiple prefixes, you can use logic operators in find. For example, to move blah.* and foo.*, use:
find . -maxdepth 1 -type f -and \( -iname 'blah.*' -or -iname 'foo.*' \) -print0 | xargs -0 -I {} mv {} ~/target/dir
EDIT
Updated after comment.
Here's how I'd do it:
find ./ -type f -printf '%f\n' | sed 's/\..*//' | sort | uniq | ( while read filename ; do find . -type f -iname "$filename"'*' -exec mv {} /dest/dir \; ; sleep 240; done )
Perhaps it needs some explaination:
find ./ -type f -printf '%f\n': find all files and print just their name, followed by a newline. If you don't want to look in subdirectories, this can be substituted by a simple ls;
sed 's/\..*//': strip the file extension by removing everything after the first dot. Both foo.tar ad foo.tar.gz are transformed into foo;
sort | unique: sort the filenames just found and remove duplicates;
(: open a subshell:
while read filename: read a line and put it into the $filename variable;
find . -type f -iname "$filename"'*' -exec mv {} /dest/dir \;: find in the current directory (find .) all the files (-type f) whose name starts with the value in filename (-iname "$filename"'*', this works also for files containing whitespaces in their name) and execute the mv command on each one (-exec mv {} /dest/dir \;)
sleep 240: sleep
): end of subshell.
Add -maxdepth 1 as argument to find as you see fit for your requirements.
Nevermind, I'm dumb. there's a uniq command. Duh. New working script is: find ./ -maxdepth 1 -type f | sed -e 's/.[a-zA-Z]*$//' | uniq | while read line; do mv "$line*" ~/target_dir/; echo "$line"; sleep 240; done
EDIT: Forgot close tag on code and a backslash.

Resources