Re: long index strings

I can't speak to the requirements of writing a plugin, nor can I even
admit to knowing what an MD5 hash is. I have a thought that a plug may
not be necessary.

Both FM7 (Developer) and FM8 (Advanced) allow you to create what are
called Custom Functions, which may very well solve your problem.

CFs can be anything from a simple concatenation to an exceedingly
complicated formula, including recursions.

For an idea of what's possible, take a look at some of the stuff people
have contributed to the CF list at
http://www.briandunning.com/filemaker-custom-functions/list.php.

Regarding multi-field relationships, yes, FM7/8 can do that, to an
extent. First, there;s the flexibility of being able to use traditional
logical operators (>, <, etc., etc.) in the matches (not just equals)
but also the ability to use more than one match in a relationship. The
hitch is that it is always ThisMatch AND ThatMatch. We don't have
ThisMatch OR ThatMatch yet, though that can be worked around using
multiple table occurrences.

Matt

max deodorizer wrote:

Thank you, Remi-Noel and "FP".

I'm sure breaking a long string into 20 byte segments would work,
once the string was sanitized for word separator characters, and if
the string length were limited. But it's not what I meant by
"hashing". The problem is that I don't want to limit the string
size. Using multiple sets of 3 segments would be painful because it
would kill the relational approach in FMP6. Are multi-field relations
possible in FMP8?

What I was hoping for was a way to compute a mathematical hash such
as MD5 (CRC32 isn't long enough) in Filemaker. Such a computation is
impossible in FMP script. Is this something a plug-in could do? I
couldn't find one, so it looks like I'm stuck with the
export-compute-import strategy. But even there, I'm going to have a
problem with key length: a decent hash like MD5 is 128 bits, which
would be a 32 byte hex string.

What are the requirements for writing a plug-in of your own? I'm
guessing a "developer" version of Filemaker is the first. If I could
justify that cost, are there any other direct costs/royalties to
produce a plug-in? Has anyone written up the generalities of
developing and using a Filemaker plug-in?

Sam



--

.

Relevant Pages

  • Re: "Collision for Hash Functions MD4, MD5, HAVAL-128 and RIPEMD"
    ... this was the Year of Doom for cryptographic hash functions. ... These go into great detail on the SHA-0 and MD5 collisions ... Difficulty in the former is called "collision resistance", ... you probably meant to say was "I can find a *different* string whose ...
    (comp.os.linux.security)
  • Repairing damaged MD5 values
    ... The result is that whenever the MD5 hashed bytes contained a byte ... it was stored as one hex digit instead of two. ... So instead of uniform string lengths of 32, ... in a 26 digit hash than in a 32 digit. ...
    (microsoft.public.sqlserver.programming)
  • RE: Need C# coding for MD5 Algorithm
    ... implementations of the MD5 hash algorithm inherit using C# ... // a 32 character hexadecimal string. ... // The MD5 hash of Hello World! ... Need C# coding for MD5 Algorithm ...
    (microsoft.public.dotnet.general)
  • md5.hexdigest() converting unicode string to ascii
    ... I'm trying to get a MD5 hash of a registry key value on Windows 2000 ... The registry value is a string which is stored as a unicode ...
    (comp.lang.python)
  • Re: long index strings
    ... Filemaker doesn't support enough functions to properly write your own hash ... I never even finished a trial version of a plug-in, ... I'm sure breaking a long string into 20 byte segments would work, ...
    (comp.databases.filemaker)