Noodlesoft Forums

I have a rule to rename PDF files via `sed` and an external text file of fules.

I'm using this embedded script:

Code: Select all

mv -n "$1" "`echo $1 | sed -E -f ~/sedFileThisRename.txt`"

This works great for most files, but there are a few folders where it does NOT work and i cannot figure out why.
By not work, I mean it either throws up an error in the log ("[Error] Shell script failed: Error processing shell script on file ...") or the file simply does not get renamed.

The same command DOES work when I run it manually in Terminal, e.g., with $x as the file:

Code: Select all

mv "$x" "`echo $x | sed -E -f ~/sedFileThisRename.txt`"

Why would it work in Terminal but not in Hazel?

Is $1 the full path or just the filename?


UPDATE: OK, pretty sure the issue is that $1 is the full path, not just the filename.
So do I need to cd into the folder, then mv just the file?
What's the best way to do that?

Pretty sure I need to use $(dirname "$1") and $(basename "$1") ... but how do I use cd with that? I'm getting lost with quotes/backticks/etc.

OK, I got it, thanks for listening. Leaving this here in case someone else can use it...

Code: Select all

cd "$(dirname "$1")";
b=$(basename "$1");
mv -n "$b" "`echo $b | sed -E -f ~/sedFileThisRename.txt`"


I suppose this could be done on ONE line by combining the dirname/basename ... but that's beyond me.

This thing is throwing up errors again...

hazelworker[78420] [Error] Shell script failed: Error processing shell script on file ...

Can anyone help?

You can turn on debug mode to see any script output. Add echo commands to output different phases of your script to isolate where things are going wrong.