Bash: cant' read out strings with spaces after looping through array of strings

Question

I am using a loop to read the contents of an array, which contains all of the directories and files in the directory hierarchy called 'music' (contents are strings from the previous output of 'find' command). The idea is to separate the full directory path of each array element in "directory_contents" into substrings according to genre, artist, and title. Since my music directory is sorted first by genre, then by artist, then by title, I am grabbing each relevant item using awk where the delimiter "/" shows up. For example, if the directory looks like this after using find "./Electronic/Squarepusher/My Red Hot Car.aif", I will separate "Electronic", "Squarepusher", and "My Red Hot Car", then store them each in separate arrays for genre, artist, and title. Later I will sort these data, then pipe the sorted output into another utility to print all the directory contents in a nice looking table (haven't done this yet). For now, I am just trying to view the results of the string separation with echo statements, and for the most part it seems to work. However, I can't seem to extract substrings which contain spaces, which isn't good:

-->./Hip-Hop/OutKast/title1.aif<--
Genre:
Hip-Hop
Artist:
OutKast
Title:
title1

-->./Hip-Hop/OutKast/title2.aif<--
Genre:
Hip-Hop
Artist:
OutKast
Title:
title2

-->./Hip-Hop/OutKast/title3.aif<--
Genre:
Hip-Hop
Artist:
OutKast
Title:
title3

-->./Jazz/John<--
Genre:
Jazz
Artist:
John
Title:


-->Coltrane/title1.aif<--
Genre:
title1.aif
Artist:

Title:

As you can see, when the loop reads in the string "John Coltrane", it is treating the space as a delimiter, and treating everything after "John" as a different filename. I tried looking for a solution in the bash manual under the section "Arrays" as well as other posts here, but couldn't find a solution that worked for my specific problem (sorry). If anyone has ideas, they would be greatly appreciated. The problematic code appears below, in the for loop (I didn't post the whole script because it is pretty lengthy, but let me if it is needed):

#more before this...

#declare variables                                                                                                      
declare -a genre_list                                                                                                   
declare -a title_list                                                                                                   
declare -a artist_list                                                                                                  
declare -a directory_contents                                                                                           


#populate directory with contents                                                                                       
cd $directory                                                                                                           
directory_contents=$(find .  -mindepth 1 -type f)                                                                       
cd ..                                                                                                                   


for music_file in ${directory_contents[*]}; do                                                                          
    if [[ $DEBUG = "true" ]] ; then                                                                                     
        echo "-->$music_file<--"                                                                                        
    fi                                                                                                                  

    echo "Genre:"                                                                                                       
    echo $music_file | awk -F"/" '{print $2}'                                                              
    echo "Artist:"                                                                                                      
    echo $music_file | awk -F"/" '{print $3}'                                                               
    echo "Title:"                                                                                                       
    echo $music_file | awk -F"/" '{print $4}' | awk -F"." '{print $1}'                                     
    echo ""                                                                                                             
done

ssapkota · Accepted Answer · 2011-05-26 05:59:58Z

2

Why don't you simply do it in single line:

cd $directory && \
find .  -mindepth 3 -maxdepth 3 -type f | \ 
awk -F'/' '{split($4,A,".aif"); printf "Genre: %s\nArtist: %s\nTitle: %s\n\n",$2,$3,A[1];}'

Update: (removed the .aif from the Title part)

edited May 26, 2011 at 5:59

answered May 24, 2011 at 21:35

ssapkota

3,30022 silver badges31 bronze badges

Sign up to request clarification or add additional context in comments.

7 Comments

dtg Over a year ago

This one is much more elegant than my solution. However, I will need to modify the last argument $4 further by filtering out everything after the "." in the string to grab the title just by itself (not "title.aif"). Not sure if this would work in this configuration? Thanks though!

ssapkota Over a year ago

@Dylan, The ".aif" is also removed now. :)

Pianosaurus Over a year ago

File names can also contain newlines, so you should probably also use -print0 and RS=$'\0'. However, only gawk supports this. Other awk implementations don't accept null-characters as record separators (AFAIK). By the way, why -mindepth 1? Why not -mindepth 3 -maxdepth 3?

ssapkota Over a year ago

@Pianosaurus, Dylan is having problem with spaces, he doesn't seem to have newline in the filenames, so there is no point in worrying about it. Thanks for pointing out on the depth issue. (I will update accordingly.)

Pianosaurus Over a year ago

@ssapkota, I ALWAYS worry about these things. Probably because I rarely write scripts only I use. =)

|

Roland Illig · Accepted Answer · 2011-05-24 21:39:48Z

1

If you can, you should use Perl for this task:

#! /usr/bin/perl

foreach my $fname (<*/*/*>) {
  next unless -f $fname;
  next unless $fname =~ m"^([^/]+)/([^/]+)/([^/.]+)\.\w+$";
  my ($genre, $artist, $title) = ($1, $2, $3);
  printf("Genre: %s\nArtist: %s\nTitle: %s\n\n", $genre, $artist, $title);
}

It is faster and simpler than the shell, and it is immune against whitespace in file names.

answered May 24, 2011 at 21:39

Roland Illig

41.9k12 gold badges91 silver badges127 bronze badges

1 Comment

dtg Over a year ago

Thanks for the suggestion, although I haven't done any Perl programming before and the requirement for this assignment is to use Bash.

clt60 · Accepted Answer · 2011-05-24 21:54:09Z

1

MUSICDIR="/some/path"
cd "$MUSICDIR"

find  . -type f -print | (IFS=/;while read dot genre artist title
do
    echo =$genre= =$artist= =$title=
done)

edited May 24, 2011 at 21:54

answered May 24, 2011 at 21:47

clt60

64.3k17 gold badges114 silver badges206 bronze badges

2 Comments

dtg Over a year ago

What does the "dot" paramter mean in the while loop?

Pianosaurus Over a year ago

If you run the find command manually, you will see that all the file names begin with "./". The "dot" variable will simply contain ".".

Jay · Accepted Answer · 2011-05-24 21:27:14Z

0

You might try setting the delimiter that bash uses. Perhaps to a newline character?

IFS=\n

answered May 24, 2011 at 21:27

Jay

14.5k5 gold badges47 silver badges74 bronze badges

1 Comment

dtg Over a year ago

Thanks, this one worked out with a few modifications; I needed to use IFS=$'\n' instead. Works perfectly!

Collectives™ on Stack Overflow

Bash: cant' read out strings with spaces after looping through array of strings

4 Answers 4

7 Comments

1 Comment

2 Comments

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

7 Comments

1 Comment

2 Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related