UNIX for Dummies Questions & Answers

Hi all,

I need a big help in a problem. I have a lot of files with the name "title" on a lot of folders. Each files have one single line and I need to copy all these lines to a new file, named "all_titles". For exemple, if I have

Code:

folder1
 title
  content1

folder2
 title
  content2

I want

Code:

all_title
 content1
 content2

in differents lines. One year ago I did this using the following:

Code:

find . -name title -exec cat > all_title {} \;

Before I had a SGI IRIX, and works fine, but now I have a WindowsXP with CygWin and I get

all_title
content1content2

It´s essential to me to get each one "content" in a single line on the new file "all_title".

Thank you for your help.

Daniel Uchoa.

How exactly are the files encoded? In particular, do they have DOS or Unix line endings? I'm speculating you are actually getting the same result as previously, but you are viewing it with software which doesn't understand the Unix line endings.

On Unix, each line is terminated with a line feed. On DOS, each line is terminated with the sequence carriage return, line feed.

If you need to add carriage returns, see if you have unix2dos or a similar utility. Perhaps you would like to convert the source files once and for all, and not need it in your scripts after that.

By the by, if the files are all in a subdirectory directly under the current directory, a normal wildcard should suffice.

Code:

cat */title >all_title

Maybe you could use sed to add an empty line to each input file. This will produce one completely empty line if some of your files already have a proper line terminator.

Code:

sed 'a\' */title >all_title

Thank you era, sed works fine to solve my problem.

solution:

Code:

find . -name title -exec sed 'a\' > all_title {} \;