hi all,
i have a problem with my script. hope someone can help me.

i have a many files that named with space characters, like "test test.mp3"

in bash when i want copy this file bash help me to complete the name like this:
"cp ./test\ test.mp3 /tmp" (i know that "\" character means that bash dont see the next character and for this example bash will ignore space character after test)

here is my problem, that i want to use variable for this in my script to copy this type of named files (i mean with space characters) automatically.

when i use this type these happen:

[root@localhost ~]# C=$(grep -wi "MP3" /tmp/sync | awk -F: '{print $1}'|head -n3|tail -n1)
[root@localhost ~]# echo $C
/media/ftp/Payam test .mp3
[root@localhost ~]# cp $C .
cp: cannot stat `/media/ftp/Payam': No such file or directory
cp: cannot stat `test': No such file or directory
cp: cannot stat `.mp3': No such file or directory
[root@localhost ~]#

now how can i solve my problem?? i used some method like:

[root@localhost ~]# C=$(grep -wi "MP3" /tmp/sync | awk -F: '{print $1}'|head -n3|tail -n1|sed 's/ /\\ /g')
[root@localhost ~]# echo $C
/media/ftp/Payam\ test\ .mp3
[root@localhost ~]# cp $C .
cp: cannot stat `/media/ftp/Payam\\': Invalid argument
cp: cannot stat `test\\': No such file or directory
cp: cannot stat `.mp3': No such file or directory
[root@localhost ~]#

but nothing happened good.
please help me solve this problem.

tanx.

Place the variable in double quotes ("") so it will retain the white space.

Also, when displaying code, please use [code][/code] tags so it is easily readable.

It's vital in scripting to understand how the shell handles arguments and whitespace. See here:

http://mywiki.wooledge.org/Arguments
http://mywiki.wooledge.org/WordSplitting
http://mywiki.wooledge.org/Quotes

As grail said, always quote your variables, unless you know what you're doing.


Also, never try to insert syntactical quotes or backslashes into the variable. It won't work. Due to the order in which the shell parses the line, once a character is stored in a variable, it loses all special meaning, with the exception of whitespace* and globbing patterns (as explained above).


(*Or more accurately, the characters currently set in the IFS shell variable, space-tab-newline by default.)


Edit: BTW; your text extraction command can certainly be greatly simplified. awk is a full text processing language and can do everything that grep/head/tail can do all on its own.

Unfortunately though, without knowing exactly what /tmp/sync contains or what you want from it, I can't give you a complete solution, so I have to leave the grep in for now:

Show us the file and what you want, and we can reduce it to awk alone.

Here are a few useful awk references:
http://www.grymoire.com/Unix/Awk.html
http://www.gnu.org/software/gawk/man...ode/index.html
http://www.pement.org/awk/awk1line.txt
http://www.catonmat.net/blog/awk-one...ined-part-one/

1 members found this post helpful.