regex - for loop in bash go though files with two specific extensions

Question

I want the following loop to go through m4a files AND webm files. At the moment I use two diffrent loops, the other one just replaces all the m4a from this loop. Also the files that ffmpeg outputs should remove a m4a extension if it was a m4a file, and a webm extension if it was a webm file. And replace that with mp3. (As it does here with m4a). I have no Idea how to do it, I think it has something to do with regex expressions, but I have no idea how really use them nor have I ever found a good tutorial/documentation, so if you have one, please link it aswell.

for i in *.m4a ; do
        echo "Converting file $converted / $numfiles : $i"
        ffmpeg -hide_banner -loglevel fatal -i "$i" "./mp3/${i/.m4a}.mp3"
        mv "$i" ./done
        converted=$((converted + 1))
done

Answer 1

Try the following:

for i in *.m4a *.webm; do
  echo "Converting file $converted / $numfiles : $i"
  ffmpeg -hide_banner -loglevel fatal -i "$i" "./mp3/${i%.*}.mp3"
  mv "$i" ./done
  converted=$((converted + 1))
done

• You can use for with multiple patterns (globs), as demonstrated here: *.m4a *.webm will expand to a single list of tokens that for iterates over.

  • You may want to add shopt -s nullglob before the loop so that it isn't entered with the unexpanded patterns in the event that there are no matching files.

• ${i%.*}.mp3 uses parameter expansion - specifically, % for shortest-suffix removal - to strip any existing extension from the filename, and then appends .mp3.

Note that the techniques above use patterns, not regular expressions. While distantly related, there are fundamental differences; patterns are simpler, but much more limited; see pattern matching.

^{P.S.: You can simplify converted=$((converted + 1)) to (( ++converted )).}