I don't write bash-scripts, so i can't give you a complete solution, but a few suggestions:

1) If you use a wildcard expansion like "*.flac" and then rename the files matched by that wildcard the results may or may not get obfuscated. Rename the *.flac-files to something intermediate like "*.flac.processed" and in a last step (after the for .... *.flac... do.....done-loop has finished) rename them again.

2) Maybe this is not posing a problem here, but generally it is better not to use the "for file in * ; do ....done", but a pipeline instead:

Code:

find . -prune -name "*" -type f -print | while read file ; do
   ....
done

The reason is: wildcards are expanded by the shell. The shell tries to execute the line "for file in * ; do" and replaces the asterisk with a list of filenames matching it. Only then the command will be executed. If the wildcard now matches a very big list of files chances are you hit the limitation of the input line length (in POSIX 4k characters) and get an error. This will not be the case with the find command and the pipeline.

3) You write a file to /tmp without checking if this is possible. Usually it is, but it is a risk you are taking if you do not check that first. Two problems could arise: /tmp is full and you (your user account) is not allowed to write to there.

Either check first if you can indeed write or work in a pipeline instead of an intermediate file:

Code:

newname="$(metaflac myfile.flac --export-tags-to=- | sed -n 's/^TITLE=//p;q')"

Hope this helps.
bakunin