Visual Basic for Applications/String Hashing in VBA

Summary

 * The VBA code below generates the digests for the MD5, SHA1, SHA2-256, SHA2-384, and SHA2-512 hashes; in this case for strings.
 * A hash is an output string that resembles a pseudo random sequence, and is essentially unique for any string that is used as its starting value.  Hashes cannot be easily cracked to find the string that was used in their making and they are very sensitive to input change.   That is to say, just a change in one character at the start will produce a completely different output.   Hashes can be used as the basis of pseudo random character tables, and although not purely random, such methods can produce output quality that is at least as good as the in-built Rnd function of VBA.
 * The use of a hash allows programmers to avoid the embedding of password strings in their code.
 * The memory space occupied by an application can be read with special utilities, so passwords might be found in code, then used in a normal user login. Instead, the hash of the password is listed in code, and the password is hashed for comparison only during a logon. This avoids access to the application via the conventional user route, since any hash that is found could not be reverse engineered to obtain the value needed at the user interface. This method assumes that the code cannot be run by the intruder at any location other than the logon device, and that they are unable to change the memory contents.
 * If a hacker can change the memory contents, then a common exploit is to change the hash in memory for one of their own; one that corresponds to a password that they can use at the user logon interface. The counter action against this attack is for all of the logon files to be encrypted with the user's officially issued password.  Then, even if the hash is changed, the files needed for the logon attempt cannot be decrypted for use.
 * Hashes can also be made from entire files, and the code for doing so differs only slightly from the string hashing versions given below.  The main difference in file hashing is that the file is first turned into a string before using conventional techniques.   Code is given elsewhere in this series for file hashing.   String hashes will produce an output even when the empty string is used as a starting point, unlike for file hashing where an empty text file can raise errors.
 * This VBA code is not specific for any one application, so it will work in any of say, MS Word, MS Excel, or MS Access.  These code versions include options for base-64 output or hex.

Notes on the Code
IMPORTANT. It was found that the hash routines errored in a Windows 10, 64 bit Office setup. However, subsequent checking revealed the solution. The Windows platform must have intalled the Net Framework 3.5 (includes .Net 2 and .Net 3), this older version, and not only the Net Framework 4.8 Advanced Services that was enabled in Turn Windows Features on and off. When it was selected there, the routines worked perfectly.

The VBA code below generates the digests for the MD5, SHA1, SHA2-256, SHA2-384, and SHA2-512 hashes, for strings, in either of the hex or base-64 output formats. These codings each make use of MS Office's built-in functions, and provide consistent results. It has been noted that original implementations elsewhere for the same digests can differ widely in their outputs. Only one example has been given with a seed or salt parameter (StrToSHA512Salt), and it should be noted that the HMACSHA512 class output differs from the SHA*Managed class hashes given in the remainder. The Managed classes give the best widely reported results. Note the VBA references required for correct operation. A reminder of these is given in some procedure headings.

In each case, coders can find the unmodified hash values in the bytes array and at that point they are in 8-bit bytes, that is, the numbers that represent the ASCI code as it applies to a full eight-bit, 256 character set. The code that follows the filling of the bytes array in each case decides which version of the ASCI character set to deliver. For a hex set of characters, 0-9, and A to F, the total bit set is broken into double the number of four-bit bytes, then returned for use. For the base-64 set, lower case letters, upper case letters, and integers mainly, six bit characters are made for output. These two sets are the most useful here, since they consist of commonly used characters. The 128 and 256 ASCI sets are too full of both exotic and non-printing characters to be useful. For each hash version its bit count is a constant, so the length of its output will vary according to the chosen type.

If your data is in ANSI, you will get different results between Excel/ACCESS and SQL Server when using T-SQL HASHBYTES function for characters code over 128. To solve those differences use StrConv instead of .GetBytes_4