C#: How would you make a unique filename by adding a number?

I would like to create a method which takes either a filename as a string or a FileInfo and adds an incremented number to the filename if the file exists. But can't quite wrap my head around how to do this in a good way.

For example, if I have this FileInfo

var file = new FileInfo(@"C:\file.ext");

I would like the method to give me a new FileInfo with C:\file 1.ext if C:\file.ext existed, and C:\file 2.ext if C:\file 1.ext existed and so on. Something like this:

public FileInfo MakeUnique(FileInfo fileInfo)
{
    if(fileInfo == null)
        throw new ArgumentNullException("fileInfo");
    if(!fileInfo.Exists)
        return fileInfo;

    // Somehow construct new filename from the one we have, test it, 
    // then do it again if necessary.
}

public FileInfo MakeUnique(string path)
{            
    string dir = Path.GetDirectoryName(path);
    string fileName = Path.GetFileNameWithoutExtension(path);
    string fileExt = Path.GetExtension(path);
 
    for (int i = 1; ;++i) {
        if (!File.Exists(path))
            return new FileInfo(path);

        path = Path.Combine(dir, fileName + " " + i + fileExt);
    }
}

Obviously, this is vulnerable to race conditions as noted in other answers.

Lots of good advice here. I ended up using a method written by Marc in an answer to a different question. Reformatted it a tiny bit and added another method to make it a bit easier to use "from the outside". Here is the result:

private static string numberPattern = " ({0})";

public static string NextAvailableFilename(string path)
{
    // Short-cut if already available
    if (!File.Exists(path))
        return path;

    // If path has extension then insert the number pattern just before the extension and return next filename
    if (Path.HasExtension(path))
        return GetNextFilename(path.Insert(path.LastIndexOf(Path.GetExtension(path)), numberPattern));

    // Otherwise just append the pattern to the path and return next filename
    return GetNextFilename(path + numberPattern);
}

private static string GetNextFilename(string pattern)
{
    string tmp = string.Format(pattern, 1);
    if (tmp == pattern)
        throw new ArgumentException("The pattern must include an index place-holder", "pattern");

    if (!File.Exists(tmp))
        return tmp; // short-circuit if no matches

    int min = 1, max = 2; // min is inclusive, max is exclusive/untested

    while (File.Exists(string.Format(pattern, max)))
    {
        min = max;
        max *= 2;
    }

    while (max != min + 1)
    {
        int pivot = (max + min) / 2;
        if (File.Exists(string.Format(pattern, pivot)))
            min = pivot;
        else
            max = pivot;
    }

    return string.Format(pattern, max);
}

Only partially tested it so far, but will update if I find any bugs with it. (Marcs code works nicely!) If you find any problems with it, please comment or edit or something :)

Not pretty, but I've had this for a while :

private string getNextFileName(string fileName)
{
	string extension = Path.GetExtension(fileName);

	int i = 0;
	while (File.Exists(fileName))
	{
		if (i == 0)
			fileName = fileName.Replace(extension, "(" + ++i + ")" + extension);
		else
			fileName = fileName.Replace("(" + i + ")" + extension, "(" + ++i + ")" + extension);
	}

	return fileName;
}

Assuming the files already exist:

• File.txt
• File(1).txt
• File(2).txt

the call getNextFileName("File.txt") will return "File(3).txt".

Not the most efficient because it doesn't use binary search, but should be ok for small file count. And it doesn't take race condition into account...

If checking if the file exists is too hard you can always just add a date and time to the file name to make it unique:

FileName.YYYYMMDD.HHMMSS

Maybe even add milliseconds if necessary.

If the format doesn't bother you then you can call:

try{
    string tempFile=System.IO.Path.GetTempFileName();
    string file=System.IO.Path.GetFileName(tempFile);
    //use file
    System.IO.File.Delete(tempFile);
}catch(IOException ioe){
  //handle 
}catch(FileIOPermission fp){
  //handle
}

PS:- Please read more about this at msdn before using.

/// <summary>
/// Create a unique filename for the given filename
/// </summary>
/// <param name="filename">A full filename, e.g., C:\temp\myfile.tmp</param>
/// <returns>A filename like C:\temp\myfile633822247336197902.tmp</returns>
public string GetUniqueFilename(string filename)
{
    string basename = Path.Combine(Path.GetDirectoryName(filename),
                                   Path.GetFileNameWithoutExtension(filename));
    string uniquefilename = string.Format("{0}{1}{2}",
                                            basename,
                                            DateTime.Now.Ticks,
                                            Path.GetExtension(filename));
    // Thread.Sleep(1); // To really prevent collisions, but usually not needed
    return uniquefilename;
}

As DateTime.Ticks has a resolution of 100 nanoseconds, collisions are extremely unlikely. However, a Thread.Sleep(1) will ensure that, but I doubt that it's needed

Insert a new GUID into the file name.

I must throw my 2-cents in. This is how I did it and it works for my use.

    private static string IterateFileName(string fileName)
    {
        if (!File.Exists(fileName)) return fileName;

        FileInfo fi = new FileInfo(fileName);
        string ext = fi.Extension;
        string name = fi.FullName.Substring(0, fi.FullName.Length - ext.Length);

        int i = 2;
        while (File.Exists($"{name}_{i}{ext}"))
        {
            i++;
        }


        return $"{name}_{i}{ext}";
    }

The idea is to get a list of the existing files, parse out the numbers, then make the next highest one.

Note: This is vulnerable to race conditions, so if you have more than one thread creating these files, be careful.

Note 2: This is untested.

public static FileInfo GetNextUniqueFile(string path)
{
    //if the given file doesn't exist, we're done
    if(!File.Exists(path))
        return new FileInfo(path);

    //split the path into parts
    string dirName = Path.GetDirectoryName(path);
    string fileName = Path.GetFileNameWithoutExtension(path);
    string fileExt = Path.GetExtension(path);

    //get the directory
    DirectoryInfo dir = new DirectoryInfo(dir);

    //get the list of existing files for this name and extension
    var existingFiles = dir.GetFiles(Path.ChangeExtension(fileName + " *", fileExt);

    //get the number strings from the existing files
    var NumberStrings = from file in existingFiles
                        select Path.GetFileNameWithoutExtension(file.Name)
                            .Remove(0, fileName.Length /*we remove the space too*/);

    //find the highest existing number
    int highestNumber = 0;

    foreach(var numberString in NumberStrings)
    {
        int tempNum;
        if(Int32.TryParse(numberString, out tempnum) && tempNum > highestNumber)
            highestNumber = tempNum;
    }
    
    //make the new FileInfo object
    string newFileName = fileName + " " + (highestNumber + 1).ToString();
    newFileName = Path.ChangeExtension(fileName, fileExt);

    return new FileInfo(Path.Combine(dirName, newFileName));
}

Instead of poking the disk a number of times to find out if it has a particular variant of the desired file name, you could ask for the list of files that already exist and find the first gap according to your algorithm.

public static class FileInfoExtensions
{
	public static FileInfo MakeUnique(this FileInfo fileInfo)
	{
		if (fileInfo == null)
		{
			throw new ArgumentNullException("fileInfo");
		}

		string newfileName = new FileUtilities().GetNextFileName(fileInfo.FullName);
		return new FileInfo(newfileName);
	}
}

public class FileUtilities
{
	public string GetNextFileName(string fullFileName)
	{
		if (fullFileName == null)
		{
			throw new ArgumentNullException("fullFileName");
		}

		if (!File.Exists(fullFileName))
		{
			return fullFileName;
		}
		string baseFileName = Path.GetFileNameWithoutExtension(fullFileName);
		string ext = Path.GetExtension(fullFileName);

		string filePath = Path.GetDirectoryName(fullFileName);
		var numbersUsed = Directory.GetFiles(filePath, baseFileName + "*" + ext)
			.Select(x => Path.GetFileNameWithoutExtension(x).Substring(baseFileName.Length))
			.Select(x =>
					{
						int result;
						return Int32.TryParse(x, out result) ? result : 0;
					})
			.Distinct()
			.OrderBy(x => x)
			.ToList();

		var firstGap = numbersUsed
			.Select((x, i) => new { Index = i, Item = x })
			.FirstOrDefault(x => x.Index != x.Item);
		int numberToUse = firstGap != null ? firstGap.Item : numbersUsed.Count;
		return Path.Combine(filePath, baseFileName) + numberToUse + ext;
	}
}    

Here's one that decouples the numbered naming question from the check of the filesystem:

/// <summary>
/// Finds the next unused unique (numbered) filename.
/// </summary>
/// <param name="fileName">Name of the file.</param>
/// <param name="inUse">Function that will determine if the name is already in use</param>
/// <returns>The original filename if it wasn't already used, or the filename with " (n)"
/// added to the name if the original filename is already in use.</returns>
private static string NextUniqueFilename(string fileName, Func<string, bool> inUse)
{
    if (!inUse(fileName))
    {
        // this filename has not been seen before, return it unmodified
        return fileName;
    }
    // this filename is already in use, add " (n)" to the end
    var name = Path.GetFileNameWithoutExtension(fileName);
    var extension = Path.GetExtension(fileName);
    if (name == null)
    {
        throw new Exception("File name without extension returned null.");
    }
    const int max = 9999;
    for (var i = 1; i < max; i++)
    {
        var nextUniqueFilename = string.Format("{0} ({1}){2}", name, i, extension);
        if (!inUse(nextUniqueFilename))
        {
            return nextUniqueFilename;
        }
    }
    throw new Exception(string.Format("Too many files by this name. Limit: {0}", max));
}

And here's how you might call it if you are using the filesystem

var safeName = NextUniqueFilename(filename, f => File.Exists(Path.Combine(folder, f)));

    private async Task<CloudBlockBlob> CreateBlockBlob(CloudBlobContainer container,  string blobNameToCreate)
    {
        var blockBlob = container.GetBlockBlobReference(blobNameToCreate);

        var i = 1;
        while (await blockBlob.ExistsAsync())
        {
            var newBlobNameToCreate = CreateRandomFileName(blobNameToCreate,i.ToString());
            blockBlob = container.GetBlockBlobReference(newBlobNameToCreate);
            i++;
        }

        return blockBlob;
    }



    private string CreateRandomFileName(string fileNameWithExtension, string prefix=null)
    {

        int fileExtPos = fileNameWithExtension.LastIndexOf(".", StringComparison.Ordinal);

        if (fileExtPos >= 0)
        {
            var ext = fileNameWithExtension.Substring(fileExtPos, fileNameWithExtension.Length - fileExtPos);
            var fileName = fileNameWithExtension.Substring(0, fileExtPos);

            return String.Format("{0}_{1}{2}", fileName, String.IsNullOrWhiteSpace(prefix) ? new Random().Next(int.MinValue, int.MaxValue).ToString():prefix,ext);
        }

        //This means there is no Extension for the file and its fine attaching random number at the end.
        return String.Format("{0}_{1}", fileNameWithExtension, new Random().Next(int.MinValue, int.MaxValue));
    }

I use this code to create a consecutive _1,_2,_3 etc.. file name everytime a file exists in the blob storage.

This is an answer to question in this Link, but they marked it as a duplicate, so I post my answer here.

I created this proof of concept class (may contain bugs). More explanation in code comments.

using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
using System.Text.RegularExpressions;

namespace ConsoleApp
{
    class Program
    {
        static void Main( string[] args )
        {
            var testFilePaths = new List<string>
            {
                @"c:\test\file.txt",
                @"c:\test\file(1).txt",
                @"c:\test\file(2).txt",
                @"c:\TEST2\file(3).txt",
                @"c:\test\file(5).txt",
                @"c:\test\file(5)abc.txt",
                @"c:\test\file(5).avi"
            };

            // inspect in debbuger for correct values
            var withSuffix      = new DecomposedFilePath( "c:\\files\\file(13).txt");
            var withoutSuffix   = new DecomposedFilePath( "c:\\files\\file(abc).txt");
            var withExtraNumber = new DecomposedFilePath( "c:\\files\\file(34)xyz(35).txt"); // "(34)" in the middle should be ignored

            DecomposedFilePath changedSuffix = withExtraNumber.ReplaceSuffix( 1999 ); // "file(34)xyz(35).txt" -> "file(34)xyz(1999).txt"
            DecomposedFilePath removedSuffix = changedSuffix.ReplaceSuffix( null ); // "file(34)xyz(1999).txt" -> "file(34)xyz.txt"

            var testPath = new DecomposedFilePath( "c:\\test\\file.txt");
            DecomposedFilePath nextPath1 = testPath.GetFirstFreeFilePath( testFilePaths );

            // update our list
            testFilePaths.Add( nextPath1.FullFilePath );
            DecomposedFilePath nextPath2 = testPath.GetFirstFreeFilePath( testFilePaths );
        
            testFilePaths.Add( nextPath2.FullFilePath );
            DecomposedFilePath nextPath3 = testPath.GetFirstFreeFilePath( testFilePaths );
        }
    }

    public sealed class DecomposedFilePath
    {
        public DecomposedFilePath( string filePath )
        {
            FullFilePath = Path.GetFullPath( filePath );
        }

        // "c:\myfiles\file(4).txt"
        public string FullFilePath { get; }

        // "file" or "file(1)"
        public string FileNameWithoutExt => Path.GetFileNameWithoutExtension( FullFilePath );

        // "file(13)" -> "file"
        public string FileNameWithoutExtAndSuffix => FileNameWithoutExt.Substring( 0, FileNameWithoutExt.Length - Suffix.Length ); // removes suffix

        // ".txt"
        public string Extenstion => Path.GetExtension( FullFilePath );

        // "c:\myfiles"
        public string DirectoryPath => Path.GetDirectoryName( FullFilePath );

        // "file(23)" -> "23", file -> stirng.Empty
        public string Suffix
        {
            get
            {
                // we want to extract suffix from file name, e.g. "(34)" from "file(34)"
                // I am not good at regex, but I hope it will work correctly

                var regex = new Regex( @"\([0-9]+\)$" );
                Match match = regex.Match( FileNameWithoutExt );

                if (!match.Success) return string.Empty; // suffix not found

                return match.Value; // return "(number)"
            }
        }

        // tranlates suffix "(33)" to 33. If suffix is does not exist (string.empty), returns null (int?)
        public int? SuffixAsInt
        {
            get
            {
                if (Suffix == string.Empty) return null;

                string numberOnly = Suffix.Substring( 1, Suffix.Length - 2 ); // remove '(' from beginning and ')' from end

                return int.Parse( numberOnly );
            }
        }

        // e.g. input is suffix: 56 then it changes file name from "file(34)" to "file(56)"
        public DecomposedFilePath ReplaceSuffix( int? suffix ) // null - removes suffix
        {
            string strSuffix = suffix is null ? string.Empty : $"({suffix})"; // add ( and )

            string path = Path.Combine( DirectoryPath, FileNameWithoutExtAndSuffix + strSuffix + Extenstion ); // build full path

            return new DecomposedFilePath( path );
        }

        public DecomposedFilePath GetFirstFreeFilePath( IEnumerable<string> filesInDir )
        {
            var decomposed = filesInDir
                // convert all paths to our class
                .Select( x => new DecomposedFilePath( x ) )
                // pick files only with the same extensionm as our base file, ignore case
                .Where( x => string.Equals( Extenstion, x.Extenstion, StringComparison.OrdinalIgnoreCase) )
                // pick files only with the same name (ignoring suffix)
                .Where( x => string.Equals( FileNameWithoutExtAndSuffix, x.FileNameWithoutExtAndSuffix, StringComparison.OrdinalIgnoreCase) )
                // with the same directory
                .Where( x => string.Equals( DirectoryPath, x.DirectoryPath, StringComparison.OrdinalIgnoreCase) )
                .ToList(); // create copy for easier debugging

            if (decomposed.Count == 0) return this; // no name collision

            int? firstFreeSuffix = Enumerable.Range( 1, int.MaxValue) // start numbering duplicates from 1
                                  .Select( x => (int?) x) // change to int? because SuffixAsInt is of that type
                                  .Except( decomposed.Select( x => x.SuffixAsInt) ) // remove existing suffixes
                                  .First(); // get first free suffix

            return ReplaceSuffix( firstFreeSuffix );
        }

        public override string ToString() => FullFilePath;
    }
}

This is just a string operation; find the location in the filename string where you want to insert the number, and re-construct a new string with the number inserted. To make it re-usable, you might want to look for a number in that location, and parse it out into an integer, so you can increment it.

Please note that this in general this way of generating a unique filename is insecure; there are obvious race condition hazards.

There might be ready-made solutions for this in the platform, I'm not up to speed with C# so I can't help there.

Take a look at the methods in the Path class, specifically Path.GetFileNameWithoutExtension(), and Path.GetExtension().

You may even find Path.GetRandomFileName() useful!

Edit:

In the past, I've used the technique of attempting to write the file (with my desired name), and then using the above functions to create a new name if an appropriate IOException is thrown, repeating until successful.

This method will add a index to existing file if needed:

If the file exist, find the position of the last underscore. If the content after the underscore is a number, increase this number. otherwise add first index. repeat until unused file name found.

static public string AddIndexToFileNameIfNeeded(string sFileNameWithPath)
{
    string sFileNameWithIndex = sFileNameWithPath;

    while (File.Exists(sFileNameWithIndex)) // run in while scoop so if after adding an index the the file name the new file name exist, run again until find a unused file name
    { // File exist, need to add index

        string sFilePath = Path.GetDirectoryName(sFileNameWithIndex);
        string sFileName = Path.GetFileNameWithoutExtension(sFileNameWithIndex);
        string sFileExtension = Path.GetExtension(sFileNameWithIndex);

        if (sFileName.Contains('_'))
        { // Need to increase the existing index by one or add first index

            int iIndexOfUnderscore = sFileName.LastIndexOf('_');
            string sContentAfterUnderscore = sFileName.Substring(iIndexOfUnderscore + 1);

            // check if content after last underscore is a number, if so increase index by one, if not add the number _01
            int iCurrentIndex;
            bool bIsContentAfterLastUnderscoreIsNumber = int.TryParse(sContentAfterUnderscore, out iCurrentIndex);
            if (bIsContentAfterLastUnderscoreIsNumber)
            {
                iCurrentIndex++;
                string sContentBeforUnderscore = sFileName.Substring(0, iIndexOfUnderscore);

                sFileName = sContentBeforUnderscore + "_" + iCurrentIndex.ToString("000");
                sFileNameWithIndex = sFilePath + "\\" + sFileName + sFileExtension;
            }
            else
            {
                sFileNameWithIndex = sFilePath + "\\" + sFileName + "_001" + sFileExtension;
            }
        }
        else
        { // No underscore in file name. Simple add first index
            sFileNameWithIndex = sFilePath + "\\" + sFileName + "_001" + sFileExtension;
        }
    }

    return sFileNameWithIndex;
}

I did it like this:

for (int i = 0; i <= 500; i++) //I suppose the number of files will not pass 500
        {       //Checks if C:\log\log+TheNumberOfTheFile+.txt exists...
            if (System.IO.File.Exists(@"C:\log\log"+conta_logs+".txt"))
            {
                conta_logs++;//If exists, then increment the counter
            }
            else
            {              //If not, then the file is created
                var file = System.IO.File.Create(@"C:\log\log" + conta_logs + ".txt");
                break; //When the file is created we LEAVE the *for* loop
            }
        }

I think this version is not so hard like the others, and It's a straightforward answer for what the user wanted.

Hope this self iterating function may help. It works fine for me.

public string getUniqueFileName(int i, string filepath, string filename)
    {
        string path = Path.Combine(filepath, filename);
        if (System.IO.File.Exists(path))
        {
            string name = Path.GetFileNameWithoutExtension(filename);
            string ext = Path.GetExtension(filename);
            i++;
            filename = getUniqueFileName(i, filepath, name + "_" + i + ext);
        }
        return filename; 
    }

If you need just a unique file name, so, how about this?

Path.GetRandomFileName()

I ran into this problem and, since none of the other answers seemed to have solved it in the way I wanted to, I did it on my own.

static string CheckIfFileExists(string filePath)
{
    if (File.Exists(filePath))
    {
        string parentDir = Directory.GetParent(filePath).FullName;
        string fileName = new DirectoryInfo(filePath).Name;
        string extension = Path.GetExtension(fileName);
        fileName = Path.GetFileNameWithoutExtension(fileName);
        if (CheckIfFileNameHasIndex(fileName))
        {
            string strIndex = fileName[(fileName.LastIndexOf('(')+1)..fileName.LastIndexOf(')')]; //range
            int index = int.Parse(strIndex);
            index++;
            fileName = fileName.Substring(0, fileName.LastIndexOf('(')) + "(" + index + ')';
            filePath = Path.Combine(parentDir, fileName + extension);
            return CheckIfFileExists(filePath);
        }
        else
        {
            fileName = fileName + " (1)";
            filePath = Path.Combine(parentDir, fileName + extension);
            return CheckIfFileExists(filePath);
        }
    }
    return filePath;
}

//checks if filename has an index (e.g. "file(2).jpg")
static bool CheckIfFileNameHasIndex(string fileName)
{
    bool isSuccessful = false;
    if (fileName.LastIndexOf('(')!=-1 && fileName.LastIndexOf(')')!=-1)
    {
        string index = fileName[(fileName.LastIndexOf('(')+1)..fileName.LastIndexOf(')')]; //range
        int result;
        isSuccessful = int.TryParse(index, out result);
    }
    return isSuccessful;
}

The method CheckIfFileExists is recursive, so in theory it should be able to handle a potentially unlimited number of duplicates (e.g. "file (3484939).txt"). Of course, in reality, what happens is that the maximum imposed filename length of your operating system and stuff like eventually become a bottleneck.