For example I have a file called fileNames.txt. cat >fileNames returns
abet
able
abacus
How do I use fileNames to create abet, able, and abacus as .txt files? I tried using pipes with cat and touch.
Add a comment
|
3 Answers
There are numerous ways to do that, including the following (assuming GNU xargs):
sed -e '/^[[:space:]]*$/d' -e 's/$/.txt/' fileNames.txt | xargs -rd '\n' touch --
This assumes that there is only one filename per line in fileNames.txt. The sed script deletes blank lines (thanks to @rubynorails) and adds '.txt' to the end of each filename before piping them to xargs touch.
By using the -d '\n' option with xargs (a GNU extension, which tells xargs to use only newline or '\n' as the delimiter rather than any whitespace, and disables quote processing) this even copes with filenames that have spaces and other potentially-problematic characters (except for \n or newline itself). e.g. if a line contained 'abet able', without -d '\n', two files (abet and able.txt) would be created. With it, only one file abet able.txt will be created.
-
Thanks. This answer is probably the easiest to understand. sed and xargs were the two things I was missing. I had the right conceptual stuff, just not the commands to do it. xargs seems like a really powerful command.I'm a noob– I'm a noob2015-10-02 03:55:55 +00:00Commented Oct 2, 2015 at 3:55
-
-
1using xargs is also more efficient than a for or while etc loop around touch - it executes only one
touchcommand per many hundreds of filenames rather than one touch per filename.cas– cas2015-10-02 04:07:19 +00:00Commented Oct 2, 2015 at 4:07 -
Wow that's waaaaay more efficient. Thanks for the help.I'm a noob– I'm a noob2015-10-02 04:49:56 +00:00Commented Oct 2, 2015 at 4:49
You can use an input redirection in a while loop with the read command
while read line; do
touch "${line}.txt"
done < fileNames.txt
-
This is much cleaner and more portable than the accepted answer.Peschke– Peschke2015-10-03 10:00:07 +00:00Commented Oct 3, 2015 at 10:00
-
1@Peschke, that is more portable (assuming Bourne-like shells), but not really cleaner. That won't work properly if the lines contain leading or trailing space or tabs, or start with
-. That also means running onetouchcommand per file.Stéphane Chazelas– Stéphane Chazelas2017-06-28 08:31:24 +00:00Commented Jun 28, 2017 at 8:31
awk 'NF>0' fileNames.txt | while read line; do touch "${line}.txt"; done
The awk command just gets rid of the empty lines.
To the best of my knowledge, these lines are also literal, so whatever is on the line is getting written as a file name.
-
@Emmanuel Dieguez's answer should work as well (without using
awk), but on extremely large files, removing empty lines before cycling through each one in a while loop may speed things up a bit.rubynorails– rubynorails2015-10-02 03:46:57 +00:00Commented Oct 2, 2015 at 3:46