I'm working on modernizing Rosetta Code's infrastructure. Starting with communications. Please accept this time-limited open invite to RC's Slack.. --Michael Mol (talk) 20:59, 30 May 2020 (UTC)


From Rosetta Code
Vidir is a draft programming task. It is not yet considered ready to be promoted as a complete task, for reasons that should be found in its talk page.

Recreate the vidir utility in your favorite programming language.

Vidir is a bulk file renaming utility included in the moreutils package available for Linux and Linux-like operating systems.

Basic operation involves:

  • Reading the file names from a directory into a text file.
  • Editing the text file with a third party editor to make bulk changes to file names (modify,delete).
  • Using the edited text file to then apply the desired changes to the file system.

In general, the filenames are typically written to the text file, one to a line with an identifying label (a unique integer in the originals case). The text file is edited as desired; file names changed, files moved, files removed, files added. Then the edited text file is used to apply the changes to the file system.

It is not required to support multiple operating systems. Make a note of which one(s) are supported.

There are many possible ways to enhance the utility but that is the basic functionality.

Some possible enhancements:

  • Operate on directories other than the current directory.
  • Recursive directory search.
  • File input filtering.
  • File/Directory permissions detection.
  • Allow creation of files / directories.
  • Verbose operation.
  • Text editor selectable from the command line.
  • First party text editing.
  • Support multiple operating systems

At a minimum support file name modification and file removal in the current directory. Add as many enhancements as desired.

See also


###sysinclude dir.uh
###define SEPARATOR 9
#g argc 3 < { "." }{ 2 argv } sto mypath
@mypath 'd !istrue { ."The given directory doesn't exist! Exited.\n" end }
6 '- pidstr sto pidstring
@mypath getdir { ."Cannot load the dir! Aborted.\n" end } sto mydir
mypath per // If the end is not a '/' char, then add to...
10 mem sto aktfilename
10 mem sto neworiginalname
tömb @mydir dirlist
filtered tömb SEPARATOR [ #s 0 [-] "R" != 0 [-] "D" != #g & { dropline }{ 2 dropfield 2 dropfield
"/tmp/furordiredit_tempfile-" sto myfilename
@pidstring sum myfilename
".txt" sum myfilename
filtered @myfilename listtofile // print the list into a temporary file
"EDITOR" env sto mycommand
" " sum mycommand
@myfilename sum mycommand
@mycommand shell // execute the command
editedlist @myfilename filetolist
filtered SEPARATOR externalloop: ![ // Loop for the original list
[-?] !{ dropline } // If empty lines occured...
zero foundflag editedlist SEPARATOR ![ // Loop for the edited list
// searching for the originaltype and originalnumber:
0 [-] 0 [-]§externalloop != { dropline }
1 [-] 1 [-]§externalloop != { dropline }
2 [-] 2 [-]§externalloop != {
2 [-] sbr §aktfilenamecreate
neworiginalname @mypath = 2 [-]§externalloop sum neworiginalname
@neworiginalname @aktfilename rename
."Renamed  : " @neworiginalname print ." ==> " @aktfilename printnl
one foundflag [>]
] // end for the editedlist-loop
@foundflag !{ // No file or dir. In this case it must be deleted:
0 [-]§externalloop "D" == { // directory
2 [-] sbr §aktfilenamecreate
@aktfilename rmdir
."Deleted  : " @aktfilename printnl
0 [-]§externalloop "R" == { // regular file
2 [-] sbr §aktfilenamecreate
@aktfilename removefile
."Deleted  : " @aktfilename printnl
] // End of the loop for the original list
@myfilename removefile
aktfilenamecreate: aktfilename @mypath = sum aktfilename rts
{ „mydir” }
{ „tömb” }
{ „mypath” }
{ „filtered” }
{ „pidstring” }
{ „myfilename” }
{ „mycommand” }
{ „editedlist” }
{ „foundflag” }
{ „aktfilename” }
{ „neworiginalname” }


Should work on Windows and Linux. I have commented out all the actually destructive statements.

-- demo/rosetta/vidir.exw
string directory = ".",
editor = iff(platform()=WINDOWS?"notepad":"gedit"),
workfile = "vidir.txt"
-- filter = ""
--bool recursive = false,
-- overwrite = false,
-- verbose = false -- ,... etc
procedure process_command_line()
sequence cl = command_line()[3..$]
if length(cl) then
-- I assume you can figure out how to deal with eg
-- {`-d`,`C:\Users\Pete\Documents`,`-e`,`notepad++`}
 ?{"command line arguments (ignored)",cl}
end if
end procedure
sequence d = dir(directory)
if d=-1 then
crash("no such directory")
end if
d = d[3..$] -- (drop "." and "..")
integer fn = open(workfile,"w")
for i=1 to length(d) do
printf(fn,"%d: %s\n",{i,d[i][D_NAME]})
end for
{} = system_exec(editor&" "&workfile)
object lines = get_text(workfile,GT_LF_STRIPPED)
integer last = 0
if lines=-1 then
crash("error reading edited file")
end if
for i=1 to length(lines) do
sequence r = scanf(lines[i],"%d: %s")
if r={} then
crash("error parsing line")
end if
{{integer line, string name}} = r
for last=last+1 to line-1 do
if not file_exists(d[last][D_NAME]) then ?9/0 end if
-- if not delete_file(d[last][D_NAME]) then
-- crash("error deleting file")
-- end if
end for
string prev = d[line][D_NAME]
if prev!=name then
if not file_exists(prev) then ?9/0 end if
if file_exists(name) then ?9/0 end if
-- if name[2]=':' and name[1]!=prev[1] then
-- if not move_file(prev,name) then
-- crash(error moving file")
-- end if
-- elsif not rename_file(prev,name) then
-- crash("error renaming file")
-- end if
end if
last = line
end for
{} = wait_key()


Works with: Rakudo version 2020.05

This is a fairly faithful recreation of the vidir utility with some extra "enhancements". It is set up to work seamlessly with Linux, OSX and most unix-like operating systems. Probably could be adapted for Windows too but won't work there as is.

Takes several positional and/or named command line parameters:


  • path, string, optional, defaults to the present directory (relative './').
  • filter, string, optional, defaults to none. A regex assertion you can use to filter the returned file names. E.G. \.txt$ (filenames ending with .txt extension)


  • -r or --recurse, flag, optional, default False. Recurse into nested directories and process those files as well.
  • -v or --verbose, flag, optional, default False. Be chatty about what is going on when making changes.
  • --e=whatever or --editor=whatever, string, optional, defaults the default text editor. Pass in a command name to specify a specific editor: (E.G. --editor=vim)

Will get a list of all file names from the specified director(y|ies), put them in a temporary file and open the text file with the default (or specified) text editor.

Edit the file names or directory names in the text editor: add, remove, modify file names, then save the text file; all of the modifications made to the text file will be applied to the file system.

To edit a file name: Just edit the file name, extension, whatever. Warning! There is nothing preventing you from using characters in the file name which will make it difficult to open, modify or delete the file. You can screw yourself quite easily if you make an effort. With great power comes great responsibility. The foot cannon is primed and loaded.

If you add a forward solidus (directory separator: /) to a filename, it will treat anything between separators as a directory name and WILL CREATE non-existent directories.

To move a file: Edit the directory path. The file will be moved to the new path. Non-existing directories will be created.

To delete a file: Delete the entire line (file name and identifier) in the text file.

To add a file: Add a new line with a unique integer (need not be in any order), one or more tabs, then the new file path/name. May be a relative or absolute path.

Notice that all of the above operations will fail to apply if you lack sufficient permissions for the affected files or directories.

use Sort::Naturally;
use File::Temp;
my %*SUB-MAIN-OPTS = :named-anywhere;
unit sub MAIN (
Str $path = '.', #= default $path
Str $filter = '', #= default file filter
Bool :r(:$recurse)= False, #= recursion flag
Bool :v(:$verbose)= False, #= verbose mode
Str :e(:$editor) = $*DISTRO ~~ /'Darwin'/ ?? "open" !! "xdg-open"; #= default editor
my $dir = $path;
# fix up path if necessary
$dir ~= '/' unless $dir.substr(*-1) eq '/';
# check that path is reachable
die "Can not find directory $dir" unless $dir.IO.d;
my @files;
# get files from that path
getdir( $dir, $filter );
@files.= sort( &naturally );
# set up a temp file and file handle
my ($filename, $filehandle) = tempfile :suffix('.vidir');
# write the filenames to the tempfile
@files.kv.map: { $filehandle.printf("%s\t%s\n", $^k, $^v) };
# flush the buffer to make sure all of the filenames have been written
# editor command
my $command = "$editor $filename";
# start text editor; suppress STDERR, some editors complain about open files being deleted
shell("$command 2> /dev/null");
react {
# watch for file changes
whenever IO::Notification.watch-path($filename) {
# allow a short interval for the file to finish writing
sleep .1;
# read in changed file
my %changes = $filename.IO.lines.map( { my ($k, $v) = .split(/\t+/); "{$k.trim}" => $v} );
# walk the filenames and make the desired changes
for ^@files -> $k {
if %changes{"$k"}:exists {
# name has changed, rename the file on disk
if (%changes{"$k"}) ne @files[$k] {
# check to see that the desired directory exists
checkdir %changes{"$k"};
# notify and do it
say "Renaming: {@files[$k]} to " ~ %changes{"$k"} if $verbose;
rename @files[$k], %changes{"$k"} orelse .die;
else {
# name is gone, delete the file
# notify and do it
say "Deleting: {@files[$k]}" if $verbose;
@files[$k].unlink orelse .die;
for %changes.kv -> $k, $v {
# a new name is added, add an empty file with that name
# check to see that the desired directory exists
checkdir $v;
# notify and do it
say "Adding: $v" if $verbose;
shell("touch $v") orelse .die;
# clean up when done
# watch for CTRL-C, cleanup and exit
whenever signal(SIGINT) {
print "\b\b";
# get the files in a specified directory matching the filter parameter
sub files ( $dir, $filter = '' ) {
if $filter.chars {
$dir.IO.dir.grep( *.f ).grep( *.basename.contains(/<$filter>/) );
} else {
$dir.IO.dir.grep( *.f );
# get the files in the present directory and recurse if desired
sub getdir ($dir, $filter) {
if $recurse {
@files.append: files($dir, $filter);
getdir( $_, $filter ) for $dir.IO.dir.grep( *.d );
} else {
@files = files($dir, $filter);
# check for existence of a directory and create it if not found
sub checkdir ($dir) {
unless $dir.IO.dirname.IO.e {
# if not, create it
my @path = $dir.IO.dirname.split('/');
for 1 .. @path {
my $thispath = @path[^$_].join('/');
unless $thispath.IO.e {
say "Creating new directory $thispath" if $verbose;


Library: Wren-fmt
Library: Wren-ioutil

Just a basic solution.

Although any directory can be used, only files (not sub-directories) can be processed and filtering is not supported.

As Wren-cli cannot currently run external processes such as text editors, we instead present the user with each file in turn and ask them to confirm whether it is to be left unchanged, the name is to be changed or the file is to be deleted. The maximum number of files to be processed has been limited to 999 because of this.

The amended file is then saved to disk and the changes processed, echoing name changes/deletions to the terminal if verbose is 'on'.

This has been written to work on Linux but should work on other platforms with minor changes.

import "os" for Process
import "io" for Directory, File
import "./fmt" for Fmt
import "./ioutil" for FileUtil, Input
var args = Process.arguments
if (args.count > 2) {
System.print("Too many arguments - maximum is two.")
var verbose = false
var path = ""
if (args.count == 0) {
path = "./" // current directory
} else if (args.count == 1) {
if (args[0] == "--v" || args[0] == "--verbose") {
verbose = true
path = "./"
} else {
path = args[0]
} else if (args[0] == "--v" || args[0] == "--verbose") {
verbose = true
path = args[1]
} else {
System.print("First argument is invalid, must be --v(erbose)")
if (!Directory.exists(path)) {
System.print("Unable to find directory : %(path)")
// ignore sub-directories and other special files
if (!path.endsWith("/")) path = path + "/"
var fileNames = Directory.list(path).where { |f| File.exists(path + f) }.toList
if (fileNames.count == 0) {
System.print("There are no files in directory: %(path)")
} else if (fileNames.count > 999) {
System.print("There are too many files to process - maximum is 999.")
var origNames = [] // keep a list of the original file names and their index numbers
var ix = 1
File.create("vidir.txt") { |f|
for (fileName in fileNames) {
var ixs = Fmt.dz(3, ix) // 3 digits, zero filled
f.writeBytes(Fmt.swrite("$s $s\n", ixs, fileName))
origNames.add([ixs, fileName])
ix = ix + 1
// create a new file with amended details
File.create("vidir2.txt") { |f2|
var lines = FileUtil.readLines("vidir.txt")
for (line in lines) {
if (line == "") continue // get rid of any extraneous blank lines
var action = Input.option(" (p)ass, (a)mend, (d)elete ? : ", "padPAD")
if (action == "p" || action == "P") {
f2.writeBytes(line + "\n")
if (action == "d" || action == "D") {
var name = Input.text(" Enter amended file name : ", 1)
f2.writeBytes(line[0..3] + name + "\n")
// change vidir2.txt to vidir.txt overwriting original file
FileUtil.move("vidir2.txt", "vidir.txt", true)
// process by first creating a map of the new names by index number
var newNames = {}
var lines = FileUtil.readLines("vidir.txt")
for (line in lines) {
if (line == "") continue
var split = line.split(" ")
newNames[split[0]] = split[1]
// now iterate through the origNames list and pass/amend/delete as appropriate
if (verbose) System.print("The following changes are being made:")
for (origName in origNames) {
var ixs = origName[0]
var old = origName[1]
var new = newNames[ixs]
if (new == null) { // file to be deleted }
File.delete(path + old)
if (verbose) System.print(" Deleting '%(old)'")
} else if (new != old) { // file name to be changed
FileUtil.move(path + old, path + new, true)
if (verbose) System.print(" Changing '%(old)' to '%(new)'")
if (verbose) System.print("All changes have now been processed.")

For testing purposes I've created a sub-directory called 'test' off the current directory and placed four files in it: a.txt, b.txt, c.txt and d.txt

$ wren vidir.wren --v test
001 a.txt
  (p)ass, (a)mend, (d)elete ? : p
002 b.txt
  (p)ass, (a)mend, (d)elete ? : a
  Enter amended file name : bb.txt
003 c.txt
  (p)ass, (a)mend, (d)elete ? : d
004 d.txt
  (p)ass, (a)mend, (d)elete ? : a
  Enter amended file name : dd.txt

The following changes are being made:
  Changing 'b.txt' to 'bb.txt'
  Deleting 'c.txt'
  Changing 'd.txt' to 'dd.txt'
All changes have now been processed.

$ cat vidir.txt
001 a.txt
002 bb.txt
004 dd.txt

$ dir test
a.txt  bb.txt  dd.txt