bash script to rename (renumber) files

comiconomenclaturist · 01-27-2010, 06:43 AM

I'm trying to write a script to process some images and rename them, or more specifically, renumber them so that pg_0001.png becomes pg_0.png, pg_0002.png becomes pg_1.png, etc. I've looked at the rename command and sed, but I'm not really very familiar with these. It should also be part of a bash script that I've written for the processing of these files - this is what I have so far:

Code:

#!/bin/bash
pdftk 2012\ theatre\ Sep\ 2009.pdf burst
mogrify -format png -resize 1200 -negate pg_00**.pdf
rm pg_00**.pdf

So I just need the the next line to renumber the png's - any ideas?

neonsignal · 01-27-2010, 07:15 AM

You might be able to use rename for this, eg:

Code:

rename 's/pg_([0-9]*)\.png/$1/; $_="pg_".($_-1).".png"' *.png

The first part of the rename strips off the non-numeric part; the second part subtracts 1 and regenerates the filename.

I'm assuming that the numbers are sequential and start at 1.

==LATER EDIT==

Slight improvement so that filenames without a number don't get mangled:

Code:

rename 'if (m/[0-9]+/) {$_ = "pg_".($&-1).".png";}' *.png

pixellany · 01-27-2010, 07:24 AM

You say that you looked at rename and SED, but you are not familiar with them----Did you look at the man pages or any tutorials? The hardest part of this is actually changing the number

A brute force way using sed--NOT TESTED:
In a folder containing the only files to be renamed:

Code:

for fil in *; do
     newnum = $(($(echo $fil|sed 's/[^0-9]//g')-1))
     newname = pg_${newnum}.png
     mv $fil $newname
done

Test this (or any other code) on some duplicate files in a separate folder.

comiconomenclaturist · 01-27-2010, 07:44 AM

Quote:

Originally Posted by neonsignal

Code:

rename 's/pg_([0-9]*)\.png/$1/; $_="pg_".($_-1).".png"' *.png

This works great for me, thanks. I'm learning some of these commands still, but this was a bit beyond me!

David the H. · 01-27-2010, 07:58 AM

According to the man page, you can specify the format in which pdftk outputs it's files.

Quote:

burst

Splits a single, input PDF document into individual pages. Also creates a report named doc_data.txt which is the same as the output from dump_data. If the output section is omitted, then PDF pages are named: pg_%04d.pdf, e.g.: pg_0001.pdf, pg_0002.pdf, etc. To name these pages yourself, supply a printf-styled format string via the output section. For example, if you want pages named: page_01.pdf, page_02.pdf, etc., pass output page_%02d.pdf to pdftk.

Edit: By the way, if the above doesn't do it for you, here's a way to rename the files that only uses bash built-ins. No need to call on external tools like sed.

Code:

i=1
for f in pg_*.pdf; do

   i=$(printf %02d $i)   #zero-pad "$i", if wanted

   mv "$f" "${f/%_*.pdf/_$i.pdf}"  #replace the orginal file ending with "$i.pdf"
   let i++     #increment "$i" for the next file

done

dr_smit · 07-22-2010, 09:24 AM

I have scanned a book with Xsane--
the pages have been scanned and stored in sets of 50's in folder named "50odd" "50even"
"100odd" "100even" "150odd" "150even" storing first 50 odd & even, next 50 odd & even, next 50 odd & even pages respectively as image-0001.pnm, image-0002.pnm and so on in each...

Now I wish to
1. know that in future scans can I have Xsane autonumber scan as odd numbers and even numbers seperately as it scans odd numbered pages first and then even numbered pages.

2. do number all files in "50odd" folder as page 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25 and in "50even" as 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24 and in "100odd" as 27, 29, 31, 33,35, 37, 39, 41, 43, 45, 47, 49 and in "100even" as 26, 28,30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50 and so on..

3. finally add all pages into a pdf book from page 1 to 150 page.