[SOLVED] Renaming files with bash (round 2)

acacius · 08-27-2016, 07:32 PM

I am reading the "Bash Cookbook" from O'Reilly and came across this script (pg. 123 in the book) to rename burned CD tracks so that the filenames will only be track numbers:

Code:

#!/usr/bin/env bash

for CDTRACK in *
 do
  if [[ "$CDTRACK" =~ "([[:alpha:][:blank:]]*)- ([[:digit:]]*) - (.*)$" ]]
   then
   echo Track ${BASH_REMATCH[2]} is ${BASH_REMATCH[3]}
   mv "$CDTRACK" "Track${BASH_REMATCH[2]}"
  fi
 done

However, when I create the file, and try the script, nothing happens.

Code:

maxim@debian2:~$ touch "Ludwig Van Beethoven - 04 - Coriolan Overture, Op. 62.ogg"
maxim@debian2:~$ ls
Ludwig Van Beethoven - 04 - Coriolan Overture, Op. 62.ogg
maxim@debian2:~$ bash /home/maxim/bash_scripts/test.sh
maxim@debian2:~$ ls
Ludwig Van Beethoven - 04 - Coriolan Overture, Op. 62.ogg

Why does the script do nothing?

jpollard · 08-27-2016, 07:44 PM

The problem is the spaces.

The "for CDTRACK in *" expands into "for CDTRACK in Ludwig Van Beethoven - 04 - Coriolan Overture, Op. 62.ogg",
thus every value of CDTRACK is one of "Ludwig", "Van", "Beethoven", "-", "04", "-", "Coriolan", "Overture", ",", "Op.", "62.ogg", and not the list of filenames with the spaces. The result is none of the matching succeeds.

If there are no subdirectories, you likely would have a better result using

Code:

find . -print | while read CDTRACK ; do
...
  done

Or anything that will provide the filenames one at a time ("ls -C1" comes to mind as an alternative to using find)

acacius · 08-27-2016, 08:02 PM

Ok I tried this but it still does nothing:

Code:

#!/usr/bin/env bash

ls | while read -r CDTRACK ;
 do
  if [[ "$CDTRACK" =~ "([[:alpha:][:blank:]]*)- ([[:digit:]]*) - (.*)$" ]]
   then
   echo Track ${BASH_REMATCH[2]} is ${BASH_REMATCH[3]}
   mv "$CDTRACK" "Track${BASH_REMATCH[2]}"
  fi
 done

allend · 08-27-2016, 09:52 PM

Quote:

if [[ "$CDTRACK" =~ "([[:alpha:][:blank:]]*)- ([[:digit:]]*) - (.*)$" ]]

The problem appears to be in the quoting of the regular expression.

Code:

  if [[ "$CDTRACK" =~ ([[:alpha:][:blank:]]+)"- "([[:digit:]]+)" - "(.+)$ ]]

works here. Note that I have changed from asterisk ( 0 or more occurrences ) to plus ( 1 or more occurrences ).

The 'for CDTRACK in *' construction is fine. You can check by adding an echo command to echo $CDTRACK.

grail · 08-28-2016, 05:36 AM

allend has solved the problem, so I will make a suggestion. It has been my experience that when using regular expressions in bash, which are more complicated that standard text, you are better
served to place your expression in a variable using single quotes to preserve all the special characters and then calling the variable raw, ie. without quotes.
So using your current example, my approach would be:

Code:

regex='([[:alpha:][:blank:]]+)- ([[:digit:]]+) - (.+)$'

...
if [[ "$CDTRACK" =~ $regex ]]

As you can see, this allows you to write the regex in a normal fashion without having to escape / quote bash sensitive items

allend · 08-28-2016, 08:15 AM

Nice one! Thanks grail.

PS - I really do not like '([[:alpha:][:blank:]]+)' at the head of the regex. How do you process '634-5789'? https://www.youtube.com/watch?v=3KnhJSne-3Y

ondoho · 08-28-2016, 10:35 AM

i'm sorry but what is this mysterious BASH_REMATCH array? am i missing something?

allend · 08-28-2016, 11:00 AM

From 'man bash'

Quote:

An additional binary operator, =~, is available, with the same precedence as == and !=. When it is used, the string to the right of the operator is considered an extended regular expression and matched accordingly (as in regex(3)).
...
Substrings matched by parenthesized subexpressions within the regular expression are saved in the array variable BASH_REMATCH. The element of BASH_REMATCH with index 0 is the portion of the string matching the entire regular expression. The element of BASH_REMATCH with index n is the portion of the string matching the nth parenthesized subexpression.

shabang · 08-28-2016, 03:31 PM

Quote:

Originally Posted by acacius

I am reading the "Bash Cookbook" from O'Reilly and came across this script (pg. 123 in the book) to rename burned CD tracks so that the filenames will only be track numbers:

Code:

#!/bin/bash

x=1

for i in *
do
echo mv "$i" $(printf "%02d.${i##*.}\n" $x)
(( x++ ))
done

This will rename your filenames to numbers as you wanted in your post. Your files will be rename to 01.ext, 02.ext. 03.ext,..etc. The .ext is the original file extension of the files.

**Note: If you want the track numbers to be 001.ext, 002.ext, 003.ext, etc. Replace "%02d.${i##*.}\n" with "%03d.${i##*.}\n"

If you don't want any extensions with the track numbers, replace the code with echo mv "$i" $(printf "%02d\n" $x)

Continuing on...

The quotes around the variable "$i" will handle filenames with spaces or any other odd characters.

The echo command is like a preview on how the script will execute the code. This will not do anything to your files until you remove the echo statement. If happy with the preview, then remove the echo statement and run the script again.