Shell Programming and Scripting

Hi:

I'm writing a script that will take source / destination pathnames and metadata information from a tab-delimited input file, and then perform various directory creation, file moving and renaming, and tagging of files.

I think I have what I need to do the file manipulation in the script - the problem I am having is with the input file, which is being created in Excel on my MacBook.

The basic loop I'm using in the script is:

Code:

#!/bin/bash

while read inputline
do
	echo $inputline
done < inputfile.txt

exit 0

My input file contains: Source path/filename, Destination path/filename, Date, and Keywords, separated by tabs.

When I save the file in Mac Excel, the lines are terminated with a ^M character on all except the last line, which brings me to my problem. The script above ignores the last line in the input file. Interestingly enough, when I create the exact same file in PC Excel, the final line terminator does come through, and the script handles it fine. Unfortunately (for me), I'm using my MacBook exclusively to develop the spreadsheet, and it looks like I'll have to live with the bug.

My question is - is there a way to modify the script above, so that it will take the last line as input (without the terminator), or can the script check for existence of a final line terminator and add one if needed before executing the loop? My script will be run numerous times against numerous copies of the spreadsheet, so remembering to add it manually isn't going to be very effective.

Thanks in advance for any help!

I suspect the last line not only has no ^M, but no newline as well. Not just bash read, but many shell utilities such as awk and sed can't be depended on to read a last line which has no ending newline.

Fortunately it's easy enough to add it.

Code:

cp input /tmp/$$
printf "\n" >> /tmp/$$

while read LINE
do
        # If we added an extra blank line, ignore it
        [ -z "$LINE" ] && continue
done < /tmp/$$
rm -f /tmp/$$

Ok - trying to unscramble this...
- first lines make a temp copy of the input file and put a newline character at the end
- next section loops through the file

I'm not clear on what the line '[ -z "$LINE" ] && continue' does. Is it jumping to the 'done' line, which would skip my processing code?

Thanks!

It would, but only if the line is completely, utterly blank anyway. That's what -z means.

The idea is, if we add an extra newline to a file which doesn't need one, that won't add an extra line -- and an invalid one, at that -- to your processing.

Just tried it - looks like exactly what I needed. Thank you!!!!