Bash - output to loop using find - bash

Question

I have a list of files in a file (files.txt):
file1.txt
file2.txt
file3.txt
When I execute:
cat files.txt | while read line; do find -name $line >> outfile.txt; done
Why don't I get an outfile with a list of those files' paths?
Whereas if I execute:
find -name file1.txt >> outfile.txt
the task is performed
Many thanks
Clive

Answer 1

Apparently you edited files.txt on a Windows box, then copied it to your Linux or Unix server without converting line endings. This is not an uncommon problem, and multiple solutions exist. :)
Windows uses CR+LF (\r\n) for newlines, whereas unix and linux use only LF (\n).
The first and possibly easiest option may be to re-copy the files with the appropriate conversion in place, if your copy program supports such a thing. If you copied the files using FTP protocol, check your client for a "type" option, which may be set to "ascii" or "bin". Depends on your client. If you're using something like scp, which transfers only binary files, then read on.
Another tried-and-true option is to use the dos2unix application which may already be installed on your unix or linux server. If it's not, you may be able to install it using your machine's package manager. (I don't know how, because you haven't mentioned what operating system you're using.) If installed, documentation for using dos2unix can be found by running man dos2unix. For example, if you want to convert all the text files matching *.txt in the current directory and all subdirectories under the current one, you might use the following:
find . -type f -name \*.txt -exec dos2unix -k -o {} \;
The options here work are as follows:
find . - tells find to search recursively, starting from the current directory.
-type f -name \*.txt restricts our search by file type and glob.
-exec runs the rest of the line up to \; on each file, with {} replaced with the filename.
dos2unix - well, you know.
-k - "keep" the timestamp on your original file.
-o - edit the "original" file rather than writing a new file.
If dos2unix isn't available, a number of other built-in tools may be able to do a similar job.
If you're running Linux, you can use GNU sed like this on just one file:
sed -i 's/\r$//' files.txt
Or to process all text files in the current directory:
for file in *.txt; do sed -i 's/\r$//' "$file"; done
Or to run this recursively, if you're using bash version 4 or higher:
shopt -s globstar
for file in **/*.txt; do sed -i 's/\r$//' "$file"; done
Another option for converting your line endings might be perl:
perl -pi -e 's/\r\n/\n/g' files.txt
You could easily make it handle multiple files either in one directory or recursively in a similar way to the options above, if you want.
One more option might be to leave the files as-is, and do your conversion as you process the files.txt in bash. For example:
while read line; do
find . -name "${line%$'\r'}"
done < files.txt > outfile.txt
This uses the shell's parameter expansion combined with bash's format expansion to "strip" the offending CR character from the end of each variable as it is read by your while loop.
Also of note:
https://en.wikipedia.org/wiki/Newline#Conversion_utilities