For a project I have been working on for quite some time now, I ended up handling thousands of files and as I initially named them poorly, I had to rename a bunch of them… Basically, I needed to do two kind of file name manipulations.

1. problem 1 is kind of trivial : replacing the same part of the file names by another, e.g. renaming res_old1_001.txt, res_old1_002.txt, …, res_old1_010.txt to res_new2_001.txt, res_new2_002.txt, …, res_new2_010.txt.

2. problem 2 is somewhat trickier as the parts to be replaced vary from a file to another and so are the replacing parts. This happens when you have numbers in your file names and want to increase or decrease those numbers by a fixed amount, e.g. renaming res_simu1_001.txt, res_simu1_002.txt, …, res_simu1_010.txt to res_simu1_021.txt, res_simu1_022.txt, …, res_simu1_030.txt (i.e., adding 20 to the last part of the file name).

In what follows I show how I solved these issues with mv, rename and , but keep in mind that there are various ways to do this in many (if not all) programming languages.

Using bash commands to solve problem 1

On Unix systems, the command mv allows you to rename and relocate a file pretty easily¹ : all you have to do is to give the current path to the file to be renames and the new one, like so: mv path1/name1.txt path2/name2.txt. In order to rename a bunch files, you can write a small bash script. For instance, the following lines solve problem 1:

1
2
3
4

for i in {001..10}
do
mv '/path/to/folder1/res_old1_'$i'.txt' '/path/to/folder2/res_new2_'$i'.txt'
done

Another option is to use rename, which is a command line tool based on a perl package. What I like about it is that you can rename file in a very similar fashion that you replace character strings with the stream editor sed.

Below I install rename and use it to solve problem 1:

1
2

$ apt-get install rename
$ rename 's/old1/new2/' /path/to/folder1/res_old1_*.txt

For more examples, see this post/tutorial on computerhope.

Using R to solve problem 2

Below, I use to solve problem 2. I would like to stress that I am using simply because I am more confortable doing this manipulation with R. Obviously there are ways of accomplishing this with many programming language and I believe that this is totally doable with rename. That being said, this is also a good opportunity to exemplify how to use can to rename files.

First, let’s list the in /path/to/folder1/

1
2
3
4
5
6
7
8

R> fls <- list.files("/path/to/folder1/", recursive = TRUE, full.names = TRUE)
R> head(fls)
[1] "/path/to/folder1/res_simu1_001.txt"
[2] "/path/to/folder1/res_simu1_002.txt"
[3] "/path/to/folder1/res_simu1_003.txt"
[4] "/path/to/folder1/res_simu1_004.txt"
[5] "/path/to/folder1/res_simu1_005.txt"
[6] "/path/to/folder1/res_simu1_006.txt"

Note that I purposely changed the original long path to /path/to/folder1/ for the sake of clarity. So now what I am looking for is to rename res_simu1_01.txt to res_simu1_21.txt, res_simu1_02.txt to res_simu1_22.txt and so forth. To do so, I need to extract the number add 20 and then rename the file, and I wrote the following function that does the job!

1
2
3
4
5
6
7
8
9

R> myrename <- function(x, offset) {
R>   # this is one way to extract the number identifying 
R>   # this would work even if there are numbers in the name of your files
R>   val <- sub("^.*_([0-9]{3})\\.txt$", "\\1", x)
R>   # add 20 and get a character string correclty formatted
R>   rep <- sprintf("%03d", as.numeric(val) + offset)
R>   # rename files
R>   file.rename(x, sub("[0-9]{3}", rep, x))
R> }

where x is a file name and offset allows one to chose the number to be added. So, in order to solve problem 2, I can use apply() as follows :

1

R> lapply(fls, myrename, 20)

And I can actually add any number, including negative ones. That said, with this solution it may be important to ensure that numbers obtained after addition remain within [0; 1000], otherwise you may not obtain what you were looking for. Even more important, you should check whether new files names are existing file name, cause this could be quite problematic. For instance, in the example above, if I were to use lapply(fls, myrename, 5), then res_simu1_01.txt would be renamed res_simu1_06.txt and as res_simu1_01.txt is the first file to be renamed, res_simu1_06.txt would be overwritten. One way to avoid this is to change the order in which files are handled, e.g.

1
2

R> lapply(fls[6:10], myrename, 5)
R> lapply(fls[1:5], myrename, 5)

Here, myrename() was designed specifically to solve problem 2 but it would be relatively easy to make it more general, that is writing a few extra lines to handle a wider range of patterns and to check for potential problems while renaming (e.g. the one I just described).