[ccoss] Re: seeing how many files are in a directory tree --a solution

  • From: Bill <rivetwa@xxxxxxxxxxxxxx>
  • To: ccoss@xxxxxxxxxxxxx
  • Date: Fri, 28 Nov 2003 10:58:31 -0500

Christopher Paulin wrote:

>Bill wrote:
>
>  
>
>>right...that's what I get for not trying it out for rea....
>>
>>What I actually do is more like this:
>>
>>(cd src/path; find  . -type f -exec md5sum {} \; ) > src.md5sum
>>(cd dest/path; find . -type f -exec md5sum {} \; ) > dest.md5sum
>>diff src.md5sum dest.md5sum
>>
>>    
>>
><--cut uneeded stuff-->
>

>p450:/mnt/mandrake/home/chris_temp # (cd ../chris ; find  . -type f 
>-exec md5sum {} \; ) > src.md5sum
>.................takes some time to complete...............
>p450:/mnt/mandrake/home/chris_temp # exit
>exit
>chris@p450:/mnt/mandrake/home/chris_temp> more src.md5sum
>c0cdfa2dbe9f4c977b0a71fb08ddff0e  ./src/cdrtools-2.00.3.tar.gz
><--snip-->
>

>You get relative paths instead of absolute paths. The paths and md5 sums 
>can both be compared.
>  
>
No...the "." is the relative path part and is identical in source and 
destination find...it does not introduce a diff..If the filesystem you 
use does not maintain case, then you do of course have a valid point 
about potential case-changes...."man diff" shows "diff -i " will address 
this directly.

but I suspect we have exhausted this topic. Other weaknesses are that 
empty directories are not verified, other types of things that ar 
non-files (links, pipes, etc) are ignored, but then again, each can be 
addressed by more sophisticated means, but most of the time it's not 
necessary.

>Quoted from below:
>
>Also copying to a Windows partition [from a different file system] changes the 
>case (lower- and upper-case) of some files.
>
>...except that would create a difference.
>
>Christopher Paulin
>
>  
>
>>But I also usually do it the first (wrong) way first, before remembering 
>>this detail!
>>
>>Cheers,
>>
>>Bill
>>
>>Christopher Paulin wrote:
>>
>> 
>>    
>>
<--snip-->

_______________________________________________
CCOSS mailing list
ccoss@xxxxxxxxxxxxx
CCOSS mailing list page: //www.freelists.org/cgi-bin/list?list_id=3594
CCOSS Web page: http://www.ccoss.org

Other related posts: