I want to use the output of find command with one of my scripts, in one command.
My script accepts:
some-script --input file1.txt file2.txt file3.txt --output final.txt
I want to replace file1.txt file2.txt file3.txt with a `find command.
some-script --input `find ./txts/*/*.txt` --output final.txt
But this breaks, as there are new lines in the output of find command.
Find can output a space-separated list of results without relying on other commands:
find . -name someFile -type f -printf "%p "
The above will find all files named "someFile" in the current directory or any sub-directories, and then print the results without line breaks.
For example, given this directory tree:
.../dir1/someFile.txt
.../dir1/dir2/someFile.txt
Running find . -name '*.txt' -type f -printf "%p " from dir1 will output:
someFile.txt ./dir2/someFile.txt
The arguments, options, and "actions" (i.e. -printf) passed to the above find command can be modified as follows:
-type f option, find will report file system entries of all types named someFile. Further, find's man page lists other parameters that can be given to -type that will cause find to search for directories only, or executable files only, or links only, etc...-printf "%p " to -printf "%f " to print file names sans path, or again, see find's man page for other format specifications that find's printf action accepts.. to some other directory path to search other directory trees.This should be adaptable to whatever shell script language you are using..
Example:
find . -type f | grep "\.cc" > ~/lolz.tmp
tr '\n' ' ' < ~/lolz.tmp
Your example:
find ./txts/*/*.txt > ~/lolz2.tmp
tr '\n' ' ' < ~/lolz2.tmp
Why dont you use the ls command instead of find command.
e.g.
some-script --input ls ./txts/*/*.txt --output final.txt
ls output the files as ./txt/file1.txt ./txt/one/file2.txt .. etc
EDIT ---
You could use the find and replace the newline with space as follows :
some-script --input find ./txts/*/*.txt | tr '\n' ' ' --output final.txt
