[SOLVED] How to handle files with spaces in their names in bash?

cgcamal · 04-01-2011, 05:26 PM

Hi all,

*I'm using Ubuntu 10.10

My issue is I can't handle the files with spaces in their name, I've donde the below script to print each file
found inside folder and subbfolders with "find".

I would like to "ls" to each file found with its complete path and with its basename too.

Code:

files=$(find . -type f)

for each in "$files"
do
ls -l "$each" # 1rst option I've tried to list with full path
ls -l "$(/bin/echo "$each")" # 2nd option I've tried to list with full path
ls -l "$(/bin/echo $(basename "$each"))" # 1nd option I've tried to list with it basename
ls -l "$(/bin/echo "$(basename "$each")")" # 2nd option I've tried to list with it basename
done

All of the "ls -l" options above I've tried, give me errors (ls: cannot access...) if there is at least one file within $files list, with spaces in the filename.

How can I list "ls -l" in both cases (with full path and with basename) when there are files with spaces in their name?

Thanks in advance,

Regards.

Nominal Animal · 04-01-2011, 05:53 PM

The safe method is to use EOS (zero) as a separator, and read them to an array:

Code:

#!/bin/bash
find . -type f -print0 | (
    while read -d "" FILE ; do FILES=("${FILES[@]}" "$FILE") ; done

    echo "Found ${#FILES[@]} files:"
    for FILE in "${FILES[@]}" ; do
        echo "'$FILE'"
    done
)

Remember to use doublequotes around the file names when using them as a parameter. For example, to list all the files, you can use

Code:

#!/bin/bash
find . -type f -print0 | (
    while read -d "" FILE ; do FILES=("${FILES[@]}" "$FILE") ; done

    ls -laF "${FILES[@]}"
)

or

Code:

#!/bin/bash
find . -type f -print0 | (
    while read -d "" FILE ; do FILES=("${FILES[@]}" "$FILE") ; done

    for FILE in "${FILES[@]}" ; do
        ls -laF "$FILE"
    done
)

which are otherwise equivalent, except the latter one uses a loop, listing each file separately, while the previous one gives the ls command all the file names in one go.

Hope this helps.

cgcamal · 04-01-2011, 08:59 PM

Quote:

Originally Posted by Nominal Animal

The safe method is to use EOS (zero) as a separator, and read them to an array:

Code:

#!/bin/bash
find . -type f -print0 | (
    while read -d "" FILE ; do FILES=("${FILES[@]}" "$FILE") ; done

    echo "Found ${#FILES[@]} files:"
    for FILE in "${FILES[@]}" ; do
        echo "'$FILE'"
    done
)

Hi Nominal_A,

Thanks for your reply. It works nice!

I understand better now why you introduced the option print0.

Code:

-print0 True; print the full file name on the standard output, followed  by a null character (instead of the newline character that '-print'  uses). This allows file names that contain newlines or other types of white space to be  correctly interpreted by programs that process the find output. This option corresponds to the '-0' option of xargs.

Now I have 2 questions:
1-)May you explain how the array is loaded with this syntax? How it works?

Code:

FILES=("${FILES[@]}" "$FILE")

2-) How can be introduce the basename option?
I've tried:

Code:

    ls -laF "$(basename ${FILES[@]})"

and

Code:

    ls -laF "${basename FILES[@]}"

But doesn't work.

Thanks in advance,

Regards.

Nominal Animal · 04-01-2011, 09:45 PM

Quote:

Originally Posted by cgcamal

1-)May you explain how the array is loaded with this syntax? How it works?

Code:

FILES=("${FILES[@]}" "$FILE")

The statement replaces the list in variable FILES with another list. The new list contains each and every one of the old list items, plus one new one, $FILE. For Bash, "${list[@]}" expands to each of the list elements as a separate token. Quotes are important, since Bash uses them to understand which ones are supposed to be separate tokens.

Quote:

Originally Posted by cgcamal

2-) How can be introduce the basename option?

You will need to loop thorough the FILES array, and apply basename separately to each one:

Code:

for FILE in "${FILES[@]}" ; do
    BASE="`basename "$FILE"`"

    ls -laF "$BASE"
done

If you want, you can also create a parallel array containing only the base names:

Code:

BASENAMES=()
for FILE in "${FILES[@]}" ; do
    BASENAME=("${BASENAMES[@]}" "`basename "$FILE"`")
done

after which you can run any command and supply all the base names to it:

Code:

ls -laF "${BASENAMES[@]}"

If you have multiple parallel arrays like FILES and BASENAMES above, and you want to loop thorough them, you'll want to use this:

Code:

for ((I=0; I<${#FILES[@]}; I++)) ; do
    FILE="${FILES[I]}"
    BASE="${BASENAMES[I]}"

    # Do something with $FILE and $BASE ... 
done

Note that ${#FILES[@]} evaluates to the number of items in FILES, and in Bash arrays, indexes start from zero.

cgcamal · 04-01-2011, 11:09 PM

Hi NA,

Very, very, very helpful.

Really thanks for your kindly and clear explanation.

I'll try to practice this concepts asap to understand better.

Really thanks

.

Regards.

grail · 04-02-2011, 02:57 AM

I would also add that you could have it all done with your find:

Code:

find . -type f -ls -printf "basename is %f\n"

Also, you can append to a bash array using += notation as well:

Code:

FILES+=("$FILE")

Nominal Animal · 04-02-2011, 05:42 AM

Quote:

Originally Posted by grail

Also, you can append to a bash array using += notation as well:

Code:

FILES+=("$FILE")

I think that only works in Bash 3 and newer. I frequently use systems with an older version of Bash, so I always try to make my scripts compatible.

Unfortunately, I cannot find reliable documentation for older versions of Bash -- even Waybackmachine is glitching right now --, so I'm not sure. If you find a good source for previous bash manuals, I'd appreciate it.