I am attempting to create a script that takes all the files in a directory "in", converts them to pdfs in an "out" directory, and creates a log file of the same name also in the "out" directory.

The script needs to be able to cope with space and control characters in the target file names.

the script I am using at the moment is:

for file in `ls in`
do
echo $file >> "out/$file.log"
date >> "out/$file.log"
# then create pdf files...
done

This creates the log files, but will wrap $file at any spaces.

I have tried using ls -Q, but this creates errors. I have also tried "`ls in`", but this creates a very long file name.

Any suggestions...

Dan
(and restricting file names is not an option...)

instead of saying

for file in `ls in`

instead say

for file in in/*

the difference is that 'ls in' would provide a list of text strings but the in/* would then be shell substitution so the filenames don't get distorted.

hope that helps.
jpbarto

Thanks for that...

The only problem that I can see with that is that it returns the results as being:

in/Site Report 1.prn
in/Site Report 2.prn
...etc

which makes it a bit more difficult to shift things across to the out directory. For the moment I will need to use a couple of "cd" commands in the script... not really a problem, but if there's a better solution it would be appreciated.

Dan

Moving to Programming



Cool

You could either seperate the name from the dir using:
file = `echo $file | cut -d"\/" -f2`

Or go back to the ls command and set the field seperator to a newline before the "for" command:
IFS="
"

If you use find you won't have the leading dir problem in/Site Report 1.prn.

for file in `find in/*`

Also if you use double quotes around "$file" you wont have the broken filename problem.