UNIX for Dummies Questions & Answers

need help....just wanna copy all files with no extensions but with tree-informations to one dir

first I make a new directory with all foldertrees...like

find / -name "*" -type d >a1
sed -e 's/^/new/g' <a1 >b1
mkdir -p `cat a1`

then I wanna copy all specific files with no extensions(!)..like

cd /temp
find / -name "*" -type f >a2
grep -v "\.TMP$" a2 >b2
grep -v "\.Z$" b2 >a2
grep -v "\.gz$" a2 >b2
.....and so on, then...

sed -e 's/^/new/g' <b2 >c2
mv c2 /
cd
for i in `cat b2`
do
cp -pf $i `cat c2`
done
....
this loop isn't working.

[no such directory....but the directory is there!]

also it is possible to make easily....
cp /dasdöadk/fsldfjsdfö/sdfsdf (is a file!) /dasddsss/reww/sdfs (also a file!) ...but the loop can't it .

but cp -pf $i /temp works - but without the subfolders.....



how to solve this ?

--------------------------

to say it more easier....just like copy files by a loop, for example...users/123/abc (is a file) to new/users/123/abc (also file)

Probably the easiest way to do this would be to cp -r the source to the destination then rm -r *.* the destination to remove everything with an extension.

I thought of a real solution........

by the way I wanna get the loop working. not the files - the loop is what I'm interesting in.

in file a there are f.e. the files:

users/tmp/1212
users/tmp/232121
users/hhsfs/234q2q3
ada/qeqweq/rwerwe3

and now I just wanna transfer them to like in file b:

new/users/tmp/1212
new/users/tmp/232121
new/users/hhsfs/234q2q3
new/ada/qeqweq/rwerwe3

...that's all.

Here is one to find files with no extension in the current directory

Code:

#!/bin/ksh
for file in *
do
    file_ext=${file#*.*}
    if [ $file_ext = $file ]  
    then
        echo "no extension = $file"
    else
        echo "filename = $file extension= $file_ext"
    fi
done
exit

nice ....but I've already found them. just only copy them....

If I had a directory, say
/tmp/foo

And within /tmp/foo I had some files, some with extensions, some without.

Within /tmp/foo, I've got a directory /tmp/foo/new. I want to copy all files without an extension from /tmp/foo to /tmp/foo/new, creating directories as I go.

I'd probably do something along the lines of

Code:

$ cd /tmp/foo
$ DEST=./new
$ find . -print | grep -v "^[.].*[.].*$" | grep -v "^[.]$" |\
> grep -v "$DEST" | cpio -pdv $DEST

Sticking to the loop requirement does, IMHO, make the whole thing unecessarily complicated, and involves far too many temp files.

Cheers
ZB

find . -print | grep -v "^[.].*[.].*$" | grep -v "^[.]$" |\

...there you take off all extensions, what about the [] and why print (..never used..) ? why is there a backslash at the end of the line?

> grep -v "$DEST" | cpio -pdv $DEST

...just don't understand really. why grepping out $DEST (./new)and what does cpio do - is it not like a compression method?

..the idea with | is very nice. but what about the subfolder information? f.e. if the files are like...

users/abc/ff (file)
users/gef/geh (file)
fun/much/124 (file)

- is there a chance of

new/users/abc/ff
new/users/gef/geh
new/fun/much/124

- not like....

new/ff
new/geh
new/124?