C# read/write in case sensitive path

Question

I have a problem.

I save json from web, in json files on my computer, and the name of this json file, is the web adress of the json. For that, I get the web json into string, and then I append it in a file, with File.AppendAllText(path, content) After some time, i also need to read json from this file with File.ReadAllText(path)

My problem is sometimes, two json have a very similar name, for example :

*com/doc/BACr and *com/doc/BAcr

Problem, the path given in the methods of the class File are note case sensitive, and I end writing twice in the same file, corrupting it.

I've found on the internet solutions for the same problem for the method File.Exists(path), but nothing to replace the methods I use to read or write.

Any of you know a setting, or even another method that would be case sensitive on the path ?

Thank you

Edit : I'm obviously working on windows :(

Edit bis : I can't change the filename, because in some others json, there is reference to web path, and when I play again my local jsons, if the filename is modified, it won't be found. It's the reason I need both write and read with case sensitive path.

Answer 1

You need something that makes your files unique and in the same time something that allows you to rebuild this uniqueness when you want to read back these files.

Suppose that your couple of files is named "BAcr" and "BACr". You can get the HashCode of these two strings and you will get two different values

string file1 = "BAcr";
int file1Hash = file1.GetHashCode(); //742971449
string file2 = "BACr";
int file1Hash = file2.GetHashCode(); //-681949991

Now if you concatenate this hashcode to your filename you will get two different files and you will be able to recalculate the same hashcode for the same input filename

string newFile1 = $"{file1}.{file1Hash}";
string newFile2 = $"{file2}.{file2Hash}";

you will save your data in these two recalculated filenames and when you need to reload them you use the same trick to get the filename used to save the data starting from the same input "BAcr" or "BACr".

But string.GetHashCode doesn't guarantee uniqueness in its results so, still using the same general idea Jeroen Mostert uses this method to get an unique code from the input value

string unique1 = string.Join("", file1.Select(c => char.IsUpper(c) ? "1" : "0"))
string newFileName1 = $"{file1}.{unique1}";

Answer 2

Windows paths are indeed case insensitive, so you cannot have these filenames.

One solution would be to change the filename if it already exists... For example;

if (File.Exists(fileNameToSaveTo)){
    // Note: Your example file names did not have an extension,
    //       but if they do, you will need to first extract that then add it back on
    fileNameToSaveTo = fileNameToSaveTo + "1";
}

If using this solution, you would have to also update whatever identifier your program uses to read back from the file at a later date... as you have not posted any code I cannot guess as to what form this takes, but hopefully you get the idea?

Edit: Upon re-reading your question... it appears you use AppendAllText... In this case, this should not 'corrupt' the file as you suggest, but should simply add the contents to the end of the file? Is this not what you observe?

Edit2: After reading comments Iomed - you could use Convert.ToBase64String on the filename in your write before writing the file, the use Convert.FromBase64String on the filename in your read function before reading the file. This will allow the filename to be different based on the capitalization.

Another alternative would be to parse the JSON (the new one AND the existing file) and add the objects to an array, then write that to the file instead, avoiding your 'corruption' issue?

Answer 3

given paths: PathA, patha

for two files, use base64 trick:

string PathToFile(string url) =>  System.Convert.ToBase64String(Encoding.UTF8.GetBytes(url));

so:

Console.WriteLine(PathToFile("pathA")); //cGF0aEE=
Console.WriteLine(PathToFile("patha")); //cGF0aGE=