Bashscripts.org

I have a script that reads pathnames (one in each line and may contain space) from a file and feed all the pathnames to an external program such as `find'. How can I make it (the 2nd `find') work as if it was fed with "[email protected]"? Thanks!



Run the above script with two arguments: dir1 dir2\ with\ space

I haven't examined this in detail but generally these issues can be solved by quoting everything correctly, for example;



would be better written as...



which should eliminate issues with spaces

To make things clear, I just want the following script to do `find dir1 dir2\ with\ space -type f` after reading the file "foo".

Try this!



Hope it helps!
-J

Thanks. actually, I've been doing this. But what I need is to pipe its output for further processing without saving to a temporary file.

find $list -type f | process

Thanks in advance for any idea.

Right on, that's not a problem. My example doesn't put anything into a temporary file. It's taking input from "foo" which I borrowed from your example. I'm just generating $list with 'read' instead of 'cat'. You should be able to pipe the output to another script just fine, but you still might run into problems with the spaces. you may be able to get around that by using the exec switch to find:



Haven't tested that, but it should give you a general idea. Does that help?

-J

I'd do a tricky one

Insert script to filename:


For the find!


the trick is the IFS="\n", IFS states what the delimiting character is. In this case a newline is the delimiter for for-loops. Normally (and i think most of you encountered this before) a whitespace character will delimit the for-loop, so a path like /home/fredrik eriksson/ would itterate /home/fredrik and eriksson/.

Hope this made any sense

Best regards
Fredrik Eriksson

Makes sense to me! I try to avoid messing with IFS, but it would certainly work in this situation!