C# Method like Base64String, but only alphanumeric (no plus or slash)

Question

is there any C# method that works similar to Convert.ToBase64String but doesn't generate anything except alphanumeric output?

Thanks!

Solution 1

You're probably looking at using something like Base32 encoding then. There is a Base32 encoder/decoder for C# here by Michael Giagnocavo. It uses a combination of capitalized letters and numbers.

There's also a related discussion on StackOverflow here.

EDIT: And if by any chance this is for URL-safe related Base64 encoding, just do Base64 and replace "+" with "-" and "/" with "_". But I'm guessing, you may not want it for that.

Solution 2

The answers are a bit outdated now. For the benefit of future searchers: The best way to handle this now in C# is:

byte[] b; // fill your byte array somehow
string s = System.Web.HttpServerUtility.UrlTokenEncode(b);

This returns a Base64-encoded string that is URL-safe (which is what you said you were really after in the comments to your question).

You can then decode it again using, you guessed it:

byte[] b = System.Web.HttpServerUtility.UrlTokenDecode(s);

Solution 3

A common variant of base-64 (for use on query-string) is to use '-' and '_' in place of '+' and '/'. Perhaps a bit of Replace(...) at each end would do the job?

Solution 4

you can replace + or slash with some predefined string if possible.

Solution 5

I was searching for a such like encoder and I used https://github.com/renmengye/base62-csharp/

It's an encoder/decoder in "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz" character space range.

I just wrapped its methods with

System.Text.Encoding.UTF8.GetBytes(plainText); and System.Text.Encoding.UTF8.GetString(decoded);

calls to allow it to work easily with strings.

View more solutions

Comments

[Smashery]

So you're really after base 62 encoding?

[Xiaofu]

Can you tell us your specific reason for not wanting the plus or slash from Base64, just out of interest?

[Alex]

I'm using the encoded string in a URL afterwards, and the random slashes are messing with my URL routing, while the + is disallowed by my IIS7 settings (the one where you can double encode with +, forgot the name).

[pettys]

base 62 encoding - yes, please! (See @AFract answer below)

[Alex]

I thought about doing that, but if I happen to hit that exact string during regular base64 encoding then hell breaks loose. Plus it would be really just a hack. :(

[TheVillageIdiot]

@Alex it is not a hack. It is used in communications. I really not remember the exact scheme (that's why such short answer) but it is possible to handle the occurrence of pattern in base64 string gracefully. I've read about it in AST's "Computer Network" and Stalling's "Data and Computer Communications" during my masters.

[Alex]

I thought Base64 would encode by itself to "-" and "_" as well. I guess it doesn't??

[Xiaofu]

The method provided by the .NET libraries doesn't, no.

[Alex]

Got it. Didn't know about "-" and "_" by the way - that works I guess. Thanks so much though! :)

[Marc Gravell]

No; the "classic" base-64 set is "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456‌​789+/"

[Xiaofu]

There's a brief description on URL-safe Base64 here: en.wikipedia.org/wiki/Base64#URL_applications

[Hele]

Always ends with a 0 for some reason.

[Mason G. Zhwiti]

@Hele are you saying the code doesn't work? I.e. I don't understand why it matters if your encoded strings end in 0.

[Mason G. Zhwiti]

@Hele What does base64 encoding have to do with generating random strings?

[Hele]

Just saying it's a waste. Why is it there?

[Mason G. Zhwiti]

@Hele My guess is whatever you're encoding simply has the trait that it ends with zeros when encoded in base64. If you try different strings with different lengths, you may get a different result.

[pettys]

Yes! This is what I'm looking for. Not for URL's, though - just representing a string of bytes in an easy copy/pasteable way.