Search code examples
c#compressionlz77

Decoding Z64 (ZB64) string

I'm working on breaking down ZPL label definitions generated by NiceLabel label making software. For the most part I don't have to worry about decoding the Z64 because it is just encoded graphics and I don't need to change the underlying data.

However I have a line of text that is used as a graphic by the label for some reason probably due to Font's or something.

Anyways, the Z64 or ZB64 string is created by compressing the original data using LZ77 and encoding that as Base64 and then appending a CRC at the end.

TEST STRING FULL EXAMPLE:

:Z64:eJztkDFOxDAQRb81hRsULmBtruECyRwpZYpFGLmg5AhwFKMUuYal9CtL26QwHsbe3RMguv3lz9P85wD3/CWaiZ+56OjqWA44cwKIAyfeXXL1sQ7YWqd54czltTge+VOdOQsXFp8TrLUw9KEW3+6pLU4Zk3mC0ataonSEzU8JMywGCiFcue+c8YLGvYcLF5a+68WFhbvtRs5jdmVkWolj96vgXe/it7eucT+0+gxV5N5RrdTveQpevhnxO+BEfRe0xIzc/EbUzkn3lhLSIH6DdFeu+c39Hb7c7vksfrJryB8vu6A4cxE/NjpK1/6LkJZ3+nL1gaLt3D33/Ed+AehfkrY=:6C38

TEST STRING TARGET EXAMPLE:

eJztkDFOxDAQRb81hRsULmBtruECyRwpZYpFGLmg5AhwFKMUuYal9CtL26QwHsbe3RMguv3lz9P85wD3/CWaiZ+56OjqWA44cwKIAyfeXXL1sQ7YWqd54czltTge+VOdOQsXFp8TrLUw9KEW3+6pLU4Zk3mC0ataonSEzU8JMywGCiFcue+c8YLGvYcLF5a+68WFhbvtRs5jdmVkWolj96vgXe/it7eucT+0+gxV5N5RrdTveQpevhnxO+BEfRe0xIzc/EbUzkn3lhLSIH6DdFeu+c39Hb7c7vksfrJryB8vu6A4cxE/NjpK1/6LkJZ3+nL1gaLt3D33/Ed+AehfkrY=

My Code to Decode / Decompress:

static string DecompressZb64(string compressedString)
{
    var b64 = SmartWarehouse.Shared.Utils.Parser.ConvertFromBase64(compressedString);
    var encoding = new ASCIIEncoding();
    var inBytes = Encoding.ASCII.GetBytes(b64);
    var outBytes = new byte[inBytes.Length];
    try
    {
        using (var memoryStream = new MemoryStream())
        using (var decompressionStream = new DeflateStream(memoryStream, CompressionMode.Decompress))
        {
            decompressionStream.Read(outBytes, 0, inBytes.Length);
        }

        return encoding.GetString(outBytes);
    }
    catch (Exception e)
    {
        // TODO: DOcument exception
        Console.WriteLine(e.Message);
    }

    return string.Empty;
}

Current Exception:

Block length does not match with its complement.

Stacktrace:

   at System.IO.Compression.Inflater.DecodeUncompressedBlock(Boolean& end_of_block)
   at System.IO.Compression.Inflater.Decode()
   at System.IO.Compression.Inflater.Inflate(Byte[] bytes, Int32 offset, Int32 length)
   at System.IO.Compression.DeflateStream.Read(Byte[] array, Int32 offset, Int32 count)
   at SmartWarehouse.Tools.Program.DecompressZb64(String compressedString) in C:\Users\[user_dir]\Source\Repos\Handheld.[user].[fork]\SmartWarehouse.Tools\Program.cs:line 511

UPDATE:

I was looking into this and I found this SO post it is essentially the same issue. So i did some more research and i found this article on a blog from 2007. Which discusses a workaround by skipping the first 2 bytes in the input array due to those bytes not actually being included in the RFC specification.

CODE CHANGE:

static string DecompressZb64(string compressedString)
{
    var b64 = Convert.FromBase64String(compressedString);
    var encoding = new ASCIIEncoding();
    var outBytes = new byte[b64.Length - 2];
    try
    {
        using (var memoryStream = new MemoryStream(b64))
        {
            memoryStream.ReadByte();
            memoryStream.ReadByte();
            using (var decompressionStream = new DeflateStream(memoryStream, CompressionMode.Decompress))
            {
                decompressionStream.Read(outBytes, 0, b64.Length - 2);
            }
        }

        return encoding.GetString(outBytes);
    }
    catch (Exception e)
    {
        // TODO: DOcument exception
        Console.WriteLine(e.Message);
    }

    return string.Empty;
}

This code change doesn't actually produce an exception anymore however it doesn't decompress properly and returns this result:

"\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"
Solution

  • That data looks fine to me, I got this image out of it:

    image

    It's a monochrome bitmap, one bit per pixel.

    Your code didn't read until the end of the stream, there are actually 1280 bytes of image data, which can then be decoded to the image above.