File names with spaces

File names with spaces

[Theo. Sean Schulze]

Hello,

I am trying to write a bash shell script that will translate spaces in file names into underline characters.  This is the script as I have it now:


for file in `ls`
do
	echo $file
	newfile=`ls ${file} | tr '[:space:]' '[_*]'`
	echo File is named ${file}
	echo The new file is named ${newfile}
	# [[ -s $newfile ]] || (mv $file $newfile)
	sleep 2
done

The lines that begin with echo and the sleep line are for debugging.  What they have shown me is that the $file is getting set to the first word in the file name on the first iteration, the second word on the second interation, etc.  (The file names look like "001 of 150 files", "002 of 150 files", etc.)  So, on the first iteration, $file is egual to "001", on the second iteration $file is equal to "of", etc.  Yet, if I go to the directory and issue `ls`, the filenames are shown as one would expect with the whole four word filename on one line.

Can anyone give me a hint on how to fix this so that the whole filename is loaded into $file?

TIA,
Sean

-- 
Theo. Sean Schulze
tschulze@teamfinders.org
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: File names with spaces

[Brian Jackson]

You can try to adapt this example from The Advanced Bash Scripting Guide:
http://www.tldp.org/LDP/abs/html/moreadv.html#EX57

It deletes the file, but it shouldn't be too hard to adapt to your needs.

--Brian

On Monday 17 February 2003 01:11 pm, Theo. Sean Schulze wrote:
> Hello,
>
> I am trying to write a bash shell script that will translate spaces in file
> names into underline characters.  This is the script as I have it now:
>
>
> for file in `ls`
> do
> 	echo $file
> 	newfile=`ls ${file} | tr '[:space:]' '[_*]'`
> 	echo File is named ${file}
> 	echo The new file is named ${newfile}
> 	# [[ -s $newfile ]] || (mv $file $newfile)
> 	sleep 2
> done
>
> The lines that begin with echo and the sleep line are for debugging.  What
> they have shown me is that the $file is getting set to the first word in
> the file name on the first iteration, the second word on the second
> interation, etc.  (The file names look like "001 of 150 files", "002 of 150
> files", etc.)  So, on the first iteration, $file is egual to "001", on the
> second iteration $file is equal to "of", etc.  Yet, if I go to the
> directory and issue `ls`, the filenames are shown as one would expect with
> the whole four word filename on one line.
>
> Can anyone give me a hint on how to fix this so that the whole filename is
> loaded into $file?
>
> TIA,
> Sean

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: File names with spaces

[robin]

> I am trying to write a bash shell script that
> will translate spaces in file names into
> underline characters. This is the script as I
> have it now: 
>
> for file in `ls` 
> do 
> echo $file 
> newfile=`ls ${file} | tr '[:space:]' '[_*]'` 
> echo File is named ${file} 
> echo The new file is named ${newfile} 
> # [[ -s $newfile ]] || (mv $file $newfile) 
> sleep 2 
> done 

A solution should look like this:

for file in `ls -1`; do
  newfile=`echo "$file" | sed 's/ /_/'`
  echo "File is named ${file}"
  echo "The new file is named ${newfile}"

  mv "$file" "$newfile"
done

IHMO in the main-loop it is better to choose "ls -1", so the field
separator is \n and there's only one filename in each line.
The next <big> thing is to put the filename into quotations. Now a
filename, even with spaces, will be interpreted as one word.

Hope it helps,
Robin
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: File names with spaces

[Jim Reimer]

robin@robind.de wrote:
> 
> A solution should look like this:
> 
> for file in `ls -1`; do
>   newfile=`echo "$file" | sed 's/ /_/'`
>   echo "File is named ${file}"
>   echo "The new file is named ${newfile}"
> 
>   mv "$file" "$newfile"
> done
> 
> IHMO in the main-loop it is better to choose "ls -1", so the field
> separator is \n and there's only one filename in each line.
> The next <big> thing is to put the filename into quotations. Now a
> filename, even with spaces, will be interpreted as one word.

Robin, that still doesn't work right - try it and see:

$echo > file\ 001
$echo > file\ 002
$ ./test.sh
File is named file
The new file is named file
File is named 001
The new file is named 001
File is named file
The new file is named file
File is named 002
The new file is named 002
$

The previously referenced Bash Scripting Guide has the answer.
Change the for statement to read:
for file in *; do
and it will work.

$ ./test.sh
File is named file 001
The new file is named file_001
File is named file 002
The new file is named file_002
$

-jdr-






-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: File names with spaces

[Theo. Sean Schulze]

Thanks, that did help, although it didn't solve the problem.  I now recognize that the problem is in assigning the variable.  Both my version with ls and the version with find in the example give the expected results when printing to the console, but they both fail when used to assign a string including spaces to a variable.

I need to find a way to maintain the integrity of the string as I assign it to the file variable.  I tried `echo (ls -1)` and `echo "(ls -1)"`, but neither works.  Changing the parentheses to brackets doesn't help either.

Cheers,
Sean

On Mon, Feb 17, 2003 at 02:34:16PM -0600, Brian Jackson hunted and pecked out:
> You can try to adapt this example from The Advanced Bash Scripting Guide:
> http://www.tldp.org/LDP/abs/html/moreadv.html#EX57
> 
> It deletes the file, but it shouldn't be too hard to adapt to your needs.
> 
> --Brian
> 
> On Monday 17 February 2003 01:11 pm, Theo. Sean Schulze wrote:
> > Hello,
> >
> > I am trying to write a bash shell script that will translate spaces in file
> > names into underline characters.  This is the script as I have it now:
> >
> >
> > for file in `ls`
> > do
> > 	echo $file
> > 	newfile=`ls ${file} | tr '[:space:]' '[_*]'`
> > 	echo File is named ${file}
> > 	echo The new file is named ${newfile}
> > 	# [[ -s $newfile ]] || (mv $file $newfile)
> > 	sleep 2
> > done
> >
> > The lines that begin with echo and the sleep line are for debugging.  What
> > they have shown me is that the $file is getting set to the first word in
> > the file name on the first iteration, the second word on the second
> > interation, etc.  (The file names look like "001 of 150 files", "002 of 150
> > files", etc.)  So, on the first iteration, $file is egual to "001", on the
> > second iteration $file is equal to "of", etc.  Yet, if I go to the
> > directory and issue `ls`, the filenames are shown as one would expect with
> > the whole four word filename on one line.
> >
> > Can anyone give me a hint on how to fix this so that the whole filename is
> > loaded into $file?
> >
> > TIA,
> > Sean
> 

-- 
Theo. Sean Schulze
tschulze@teamfinders.org
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: File names with spaces

[J.]

On Tue, 18 Feb 2003, Theo. Sean Schulze wrote:

> Thanks, that did help, although it didn't solve the problem.  
> I now recognize that the problem is in assigning the variable.  
> Both my version with ls and the version with find in the example give
> the expected results when printing to the console, 
> but they both fail when used to assign a string including spaces to a
> variable. I need to find a way to maintain the integrity of the string
> as I assign it to the file variable.  I tried `echo (ls -1)` and `echo
> "(ls -1)"`, but neither works.  Changing the parentheses to brackets
> doesn't help either.
> 
> Cheers,
> Sean

Assigning variables does not happen in the `ls -1' statement. This only
generates strings. You need a command that read's the variables correctly.
Use `read', as in:

ls -1 | while read file ; do echo "$file" ; done 

or

find . -type f | while read file ; do
 echo "$file"
done

or more exotic, array version.

IFS=$'\n' eval 'lines=( $(cat < $file) )'
# now you have a array of lines.... which can be proccessed further..

These all where tested and work just fine.

G00d lUcK

J.

-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs

Re: File names with spaces

[Mike Castle]

In article <20030218093623.1FB578D8FE@basicbox3.server-home.net>,
 <robin@robind.de> wrote:
>for file in `ls -1`; do
>  newfile=`echo "$file" | sed 's/ /_/'`
>  echo "File is named ${file}"
>  echo "The new file is named ${newfile}"
>
>  mv "$file" "$newfile"
>done

>
>IHMO in the main-loop it is better to choose "ls -1", so the field
>separator is \n and there's only one filename in each line.

ls should detect that stdout is not a terminal and fall back to -1
automatically.  Granted, it can't hurt, but should be unnecessary.

However, a few comments.

First, why use ls at all?  Might as well use:
for file in *; do

Of course, since you are only interested in files with spaces in them,
while not limit to that in the first place?

mcastle@dl-mcastle[03:37pm]~/foo(776) for file in *; do echo $file; done
a b
bar
mcastle@dl-mcastle[03:37pm]~/foo(777) for file in *\ *; do echo $file; done
a b

Of course, if you have a lot of files, this technique simply doesn't work
as you'll overflow your max command line length.

You could go back to:
ls | while read; do

But you're processing every file again.

Another alternative might be something like:

find -name '* *' -maxdepth 1

I think maxdepth is gnu find specific, so keep that in mind (I don't have
access to any non-gnu systems to test).

>The next <big> thing is to put the filename into quotations. Now a
>filename, even with spaces, will be interpreted as one word.

Not just quotes, but prepend ./ too, in case any of your file names look
like "- -"

mcastle@dl-mcastle[03:45pm]~/foo(793) find
.
./bar
./a b
./blah
./blah/c d
./ns
./- -
mcastle@dl-mcastle[03:45pm]~/foo(794) ./ns
`././a b' -> `././a_b'
`././- -' -> `././-_-'
mcastle@dl-mcastle[03:45pm]~/foo(795) cat ns
#!/bin/bash

find -name '* *' -maxdepth 1 | while read name; do
        newname=$(echo $name | tr ' ' '_')
        mv -iv "./$name" ./$newname
done


Heh... forgot that find will already add in the ./ ... clean up as
appropriate.

mrc

-- 
     Mike Castle      dalgoda@ix.netcom.com      www.netcom.com/~dalgoda/
    We are all of us living in the shadow of Manhattan.  -- Watchmen
fatal ("You are in a maze of twisty compiler features, all different"); -- gcc
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.linux-learn.org/faqs