files=`find C:/PATH/TO/DIRECTORY -name *.txt`
for file in $files; do
#need code to rename $file, by moving it into the same directory
eg. $file was found in C:/PATH/TO/DIRECTORY/2014-05-08.
how do I rename $file to back to that directory and not to C:/PATH/TO/DIRECTORY?
You can use -execdir option in find:
find C:/PATH/TO/DIRECTORY -name '*.txt' -execdir mv '{}' '{}'-new \;
As per man find:
-execdir utility [argument ...] ;
The -execdir primary is identical to the -exec primary with the exception that utility will be executed
from the directory that holds the current file.
You would be better served by using this structure:
while read fname
do
....
done < <(find ...)
Or, if you're not using bash:
find ... | while read fname
do
....
done
The problem with storing the output of find in a variable, or even doing for fname in $(find ...), is with word splitting on whitespace. The above structures still fail if you have a file name with a newline in it, since they assume that you have one file name per line, but they're better than what you have now.
An even better solution would be something like this:
find ... -print0 | xargs -0 -n1 -Ixxx somescript.sh "xxx"
But even that might have issues if filenames have quotes or other things in them.
The bottom line is that parsing arbitrary data (which filenames can be) is hard...
find C:/PATH/TO/DIRECTORY -name \*.txt | while read file
do
dir=$(dirname "$file")
base=$(basename "$file")
mv "$file" "$dir/new_file_name"
done
Related
What is the shell command for renaming all files in a directory and sub-directory (recursively)?
I would like to add an underscore to all the files ending with *scss from filename.scss to _filename.scss in all the directories and sub-directories.
I have found answers relating to this but most if not all require you to know the filename itself, and I do not want this because the filenames differ and are a lot to know by heart or even type them manually and some of them are deeply nested in directories.
Edit: I was under the impression that the bash -c bit was somehow necessary for multiple expansion of the found element; anubhava's answer proved me wrong. I am leaving that bit in the answer for now as it worked for the OP.
find . -type f -name *scss -exec bash -c 'mv $1 _$1' -- {} \;
find . -- find in current directory (recursively)
-type f -- files
-name *scss -- matching the pattern *scss
-exec -- execute for each element found
bash -c '...' -- execute command in a subshell
-- -- end option parsing
{} -- expands to the name of the element found (which becomes the positional parameter for the bash -c command)
\; -- end the -exec command
You can use -execdir option here:
find ./src/components -iname "*.scss" -execdir mv {} _{} \;
You are close to a solution:
find ./src/components -iname "*.scss" -print0 | xargs -0 -n 1 -I{} mv {} _{}
In this approach, the "loop" is executed by xargs. I prefer this solution overt the usage of the -exec in find. The syntax is clear to me.
Also, if you want to repeat the command and avoid double-adding the underscore to the already processed files, use a regexp to get only the files not yet processed:
find ./src/components -iregex ".*/[^_][^/]*\.scss" -print0 | xargs -0 -n 1 -I{} mv {} _{}
By adding the -print0/-0 options, you also avoid problems with whitespaces.
#!/bin/sh
EXTENSION='.scss'
cd YOURDIR
find . -type f | while read -r LINE; do
FILE="$( basename "$LINE" )"
case "$LINE" in
*"$EXTENSION")
DIRNAME="$( dirname "$LINE" )"
mv -v "$DIRNAME/$FILE" "$DIRNAME/_$FILE"
;;
esac
done
I want to rename a file present in several subdirectories using bash script.
my files are in folders:
./FolderA/ABCD/ABCD_Something.ctl
./FolderA/EFGH/EFGH_Something.ctl
./FolderA/WXYZ/WXYZ_Something.ctl
I want to rename all of the .ctl file with the same name (name.ctl).
I tried several command using mv or rename but didnt work.
Working from FolderA:
find . -name '*.ctl' -exec rename *.ctl name.ctl '{}' \;
or
for f in ./*/*.ctl; do mv "$f" "${f/*.ctl/name .ctl}"; done
or
for f in $(find . -type f -name '*.ctl'); do mv $f $(echo "$f" | sed 's/*.ctl/name.ctl/'); done
Can you help me using bash?
thanks
You can do this with one line with:
find . -name *.ctl -exec sh -c 'mv "$1" `dirname "$1"`/name.ctl' x {} \;
The x just allows the filename to be positional character 1 rather than 0 which (in my opinion) wrong to use as a parameter.
Try this:
find . -name '*.ctl' | while read f; do
dn=$(dirname "${f}")
# remove the echo after you sanity check the output
echo mv "${f}" "${dn}/name.ctl"
done
find should get all the files you want, dirname will get just the directory name, and mv will perform the rename. You can remove the quotes if you're sure that you'll never have spaces in the names.
So in a bashscript i've the following very simple line , but how can i chain it further to move the file ?
find . -type f -ctime -$2 -name "mylog*.log" | xargs bzip2
This works fine but i'd also like to move the file to a new directory once I am done with the bzip2.
One standard trick is to use a new script that does whatever you need. Here, I assume that ${OTHER_DIRECTORY} is an environment variable that says where to put the compressed files, but there are plenty of other (better!) ways to get that information to the script (such as specifying the directory as the first argument — as the last argument is a bad idea).
#!/bin/bash
for file in "$#"
do
bzip2 "$file"
mv "$file.bz2" "${OTHER_DIRECTORY:-/tmp}"
done
You then run that script with find:
find . -type f ctime -$2 -name "mylog*.log" -exec tinyscript.sh {} +
This is pretty effective. If you only want one mv command, you can consider something along the lines of:
bzip2 "$#"
bz2=()
for file in "$#"; do bz2+=( "$file.bz2" ) done
mv "${bz3[#]}" "${OTHER_DIRECTORY:-/tmp}"
This code works even if the path names contain spaces and other untoward characters.
One option might be something like this:
find . -type f -ctime -$2 -name "mylog*.log" -exec bzip2 {} \; -exec mv {} /path/to/new_dir/ \;
How could I move all .txt files from a folder and all included folders into a target directory .
And preferably rename them to the folder they where included in, although thats not that important. I'm not exactly familiar with bash.
To recursively move files, combine find with mv.
find src/dir/ -name '*.txt' -exec mv -t target/dir/ -- {} +
Or if on a UNIX system without GNU's version of find, such as macOS, use:
find src/dir/ -name '*.txt' -exec mv -- {} target/dir/ ';'
To rename the files when you move them it's trickier. One way is to have a loop that uses "${var//from/to}" to replace all occurrences of from with to in $var.
find src/dir/ -name '*.txt' -print0 | while IFS= read -rd $'\0' file; do
mv -- "$file" target/dir/"${file//\//_}"
done
This is ugly because from is a slash, which needs to be escaped as \/.
See also:
Unix.SE: Understanding IFS= read -r line
BashFAQ: How can I read a file (data stream, variable) line-by-line (and/or field-by-field)?
Try this:
find source -name '*.txt' | xargs -I files mv files target
This will work faster than any option with -exec, since it will not invoke a singe mv process for every file which needs to be moved.
If it's just one level:
mv *.txt */*.txt target/directory/somewhere/.
I have a lot of files whose names end with '_100.jpg'. They spread in nested folder / sub-folders. Now I want a trick to recursively copy and rename all of them to have a suffix of '_crop.jpg'. Unfortunately I'm not familiar with bash scripting so don't know the exact way to do this thing. I googled and tried the 'find' command with the '-exec' para but with no luck.
Plz help me. Thanks.
find bar -iname "*_100.jpg" -printf 'mv %p %p\n' \
| sed 's/_100\.jpg$/_crop\.jpg/' \
| while read l; do eval $l; done
if you have bash 4
shopt -s globstar
for file in **/*_100.jpg; do
echo mv "$file" "${file/_100.jpg/_crop.jpg}"
one
or using find
find . -type f -iname "*_100.jpg" | while read -r FILE
do
echo mv "${FILE}" "${FILE/_100.jpg/_crop.jpg}"
done
This uses a Perl script that you may have already on your system. It's sometimes called prename instead of rename:
find /dir/to/start -type f -iname "*_100.jpg" -exec rename 's/_100/_crop' {} \;
You can make the regexes more robust if you need to protect filenames that have "_100" repeated or in parts of the name you don't want changed.