SQL Server GUID sort algorithm. Why?

Question

Problem with UniqueIdentifiers

We have an existing database which uses uniqueidentifiers extensively (unfortunately!) both as primary keys and some

Answer 1

Necromancing.
A special service for those that find that the accepted answer a bit vague.
The code speaks for itselfs, the magical parts are

System.Guid g
g.ToByteArray();
int[] m_byteOrder = new int[16] // 16 Bytes = 128 Bit 
    {10, 11, 12, 13, 14, 15, 8, 9, 6, 7, 4, 5, 0, 1, 2, 3};


public int Compare(Guid x, Guid y)
{
    byte byte1, byte2;

    //Swap to the correct order to be compared
    for (int i = 0; i < NUM_BYTES_IN_GUID; i++)
    {
        byte1 = x.ToByteArray()[m_byteOrder[i]];
        byte2 = y.ToByteArray()[m_byteOrder[i]];
        if (byte1 != byte2)
            return (byte1 < byte2) ? (int)EComparison.LT : (int)EComparison.GT;
    } // Next i 

    return (int)EComparison.EQ;
}

Full code:

namespace BlueMine.Data
{


    public class SqlGuid
        : System.IComparable
        , System.IComparable<SqlGuid>
        , System.Collections.Generic.IComparer<SqlGuid>
        , System.IEquatable<SqlGuid>
    {
        private const int NUM_BYTES_IN_GUID = 16;

        // Comparison orders.
        private static readonly int[] m_byteOrder = new int[16] // 16 Bytes = 128 Bit 
        {10, 11, 12, 13, 14, 15, 8, 9, 6, 7, 4, 5, 0, 1, 2, 3};

        private byte[] m_bytes; // the SqlGuid is null if m_value is null


        public SqlGuid(byte[] guidBytes)
        {
            if (guidBytes == null || guidBytes.Length != NUM_BYTES_IN_GUID)
                throw new System.ArgumentException("Invalid array size");

            m_bytes = new byte[NUM_BYTES_IN_GUID];
            guidBytes.CopyTo(m_bytes, 0);
        }


        public SqlGuid(System.Guid g)
        {
            m_bytes = g.ToByteArray();
        }


        public byte[] ToByteArray()
        {
            byte[] ret = new byte[NUM_BYTES_IN_GUID];
            m_bytes.CopyTo(ret, 0);
            return ret;
        }

        int CompareTo(object obj)
        {
            if (obj == null)
                return 1; // https://msdn.microsoft.com/en-us/library/system.icomparable.compareto(v=vs.110).aspx

            System.Type t = obj.GetType();

            if (object.ReferenceEquals(t, typeof(System.DBNull)))
                return 1;

            if (object.ReferenceEquals(t, typeof(SqlGuid)))
            {
                SqlGuid ui = (SqlGuid)obj;
                return this.Compare(this, ui);
            } // End if (object.ReferenceEquals(t, typeof(UInt128)))

            return 1;
        } // End Function CompareTo(object obj)


        int System.IComparable.CompareTo(object obj)
        {
            return this.CompareTo(obj);
        }


        int CompareTo(SqlGuid other)
        {
            return this.Compare(this, other);
        }


        int System.IComparable<SqlGuid>.CompareTo(SqlGuid other)
        {
            return this.Compare(this, other);
        }


        enum EComparison : int
        {
            LT = -1, // itemA precedes itemB in the sort order.
            EQ = 0, // itemA occurs in the same position as itemB in the sort order.
            GT = 1 // itemA follows itemB in the sort order.
        }


        public int Compare(SqlGuid x, SqlGuid y)
        {
            byte byte1, byte2;

            //Swap to the correct order to be compared
            for (int i = 0; i < NUM_BYTES_IN_GUID; i++)
            {
                byte1 = x.m_bytes[m_byteOrder[i]];
                byte2 = y.m_bytes[m_byteOrder[i]];
                if (byte1 != byte2)
                    return (byte1 < byte2) ? (int)EComparison.LT : (int)EComparison.GT;
            } // Next i 

            return (int)EComparison.EQ;
        }


        int System.Collections.Generic.IComparer<SqlGuid>.Compare(SqlGuid x, SqlGuid y)
        {
            return this.Compare(x, y);
        }


        public bool Equals(SqlGuid other)
        {
            return Compare(this, other) == 0;
        }


        bool System.IEquatable<SqlGuid>.Equals(SqlGuid other)
        {
            return this.Equals(other);
        }


    }


}

Answer 2

Here's a different approach. The GUID is simply shuffled around ready for a normal string comparison like it occurs in SQL Server. This is Javascript but it is very easy to convert to any language.

function guidForComparison(guid) {
  /*
  character positions:  
            11111111112222222222333333
  012345678901234567890123456789012345

  00000000-0000-0000-0000-000000000000

  byte positions:  
                          111111111111
  00112233 4455 6677 8899 001122334455
  */
  return guid.substr(24, 12) + 
         guid.substr(19, 4) + 
         guid.substr(16, 2) + 
         guid.substr(14, 2) + 
         guid.substr(11, 2) + 
         guid.substr(9, 2) + 
         guid.substr(6, 2) +
         guid.substr(4, 2) +
         guid.substr(2, 2) +
         guid.substr(0, 2);
};

Answer 3

The algorithm is documented by the SQL Server guys here: How are GUIDs compared in SQL Server 2005? I Quote here here (since it's an old article that may be gone forever in a few years)

In general, equality comparisons make a lot of sense with uniqueidentifier values. However, if you find yourself needing general ordering, then you might be looking at the wrong data type and should consider various integer types instead.

If, after careful thought, you decide to order on a uniqueidentifier column, you might be surprised by what you get back.

Given these two uniqueidentifier values:

@g1= '55666BEE-B3A0-4BF5-81A7-86FF976E763F' @g2 = '8DD5BCA5-6ABE-4F73-B4B7-393AE6BBB849'

Many people think that @g1 is less than @g2, since '55666BEE' is certainly smaller than '8DD5BCA5'. However, this is not how SQL Server 2005 compares uniqueidentifier values.

The comparison is made by looking at byte "groups" right-to-left, and left-to-right within a byte "group". A byte group is what is delimited by the '-' character. More technically, we look at bytes {10 to 15} first, then {8-9}, then {6-7}, then {4-5}, and lastly {0 to 3}.

In this specific example, we would start by comparing '86FF976E763F' with '393AE6BBB849'. Immediately we see that @g2 is indeed greater than @g1.

Note that in .NET languages, Guid values have a different default sort order than in SQL Server. If you find the need to order an array or list of Guid using SQL Server comparison semantics, you can use an array or list of SqlGuid instead, which implements IComparable in a way which is consistent with SQL Server semantics.

Plus, the sort follows byte groups endianness (see here: Globally unique identifier). The groups 10-15 and 8-9 are stored as big endian (corresponding to the Data4 in the wikipedia article), so they are compared as big endian. Other groups are compared using little endian.