Re: [Duplicity-talk] File / Directory names with spaces

[Peter Valdemar Mørch (Lists)]

Richard Scott richard-at-pointb.co.uk |Lists/Send to lists| wrote:
> --- code start ---
> find / -type f -name .duplicity-ignore -print0 | while read -d $'\0' file
> do
>   find / | grep "^$(dirname "$file")" >> /var/tmp/exclude.list
> done
> --- code end ---

Yikes!

Am I reading this shell snippet correctly?:

It searches the entire file system for files called .duplicity-ignore, 
and for each file it finds, it searches the entire file system *again*? 
So if you have 45 .duplicity-ignore files, you end up searching the 
entire file system 46 times? Lets at least hope you don't have any slow 
network attached storage.

Wouldn't it really help to avoid that nested "find /"? I haven't tried 
it, but would this work instead?:

--- code start ---
# Delete any pre-existing file for good measure so it doesn't
# just grow and grow and grow
rm -f /var/tmp/exclude.list

find / -type f -name .duplicity-ignore -print0 | while read -d $'\0' file
do
  echo $(dirname "$file") >> /var/tmp/exclude.list
done
--- code end ---

Peter
--
Peter Valdemar Mørch
http://www.morch.com