In your substitution, I think you need to be using Match Groups. Match groups are a good tool for reformatting data, like inserting a dash between two groups of numbers. The following seems to work for me:

Where the pattern I'm using for "Replace" is (edited to make it more flexible):

(\d{2})[\s\-]+(\d{2})[\s\-]+(\d{4})

And the pattern for "With" is:

${1}-${2}-${3}

The ${1} syntax is the Match Group: I have grouped each set of numbers (month, day, year) using parenthesis in the pattern match area, then referred to them with ${1}, ${2}, etc.

The term [\s\-] is a regular expression for a space or a hyphen.

Edited:

If you want to make the substitution more flexible, you can use [\s\-]+ to indicate one or more spaces and/or hyphens, just like you have \s+ in your original pattern. I have edited my pattern to show this.

Toyia,

No worries. If I'm understanding the situation correctly, the formatting that was suggested, you would not need to account for anything but numbers. The "Pattern Matching" will only output numbers. Even if your Zone OCR came back with "12-21-1974" or "12-----21----1974"(just something super random), it will ignore all those "-" and just output "12211974" . However, I believe i may have left out one additional information. In the "Edit Token" options in the "Pattern Matching" token, Under Pattern it should only be left as "\d". However in the drop down box below that, ( "If multiple matches exist, return:"), select "All Matches (Combined with no spaces)". that will generate all the numbers rather than just one number if it is left at the default "First Match only". I uploaded my example on the attachment. Of course the "Name:" and "Token value" will vary depending on what you called yours. Let me know if that helps!

Regards,

Andrew

Toyia,

In regards to DOB token issue...

Actually you can keep it as a "date" format. You would not need to make it a "text". You technically could, however there may be an issue if you leave it as "text". If somewhere in the future, you wish to utilize the "date" value itself, it will read it as  "text" rather than a "date". So we will shy away from the "text" idea. But upon looking at your word document, i believe that you attached the MM-dd-yyyy in the wrong section. I will do a quick recap with some screen shots to hopefully point you in the right direction.

We will assume we got the zone OCR covered and the Pattern matching done. The Pattern matching should only be generating numbers regardless of the "\" or "-" or other alphabets.  The key to always confirming your tokens is to always ultilize the "Test Value" box. The "result value" will input exactly as it would behave. So that is how to determine if your token will work the way you want it. 

The "pattern Matching" should look something like the attachment called "pattern Matching edit token option".

Then afterwards when you go back to your "Document Class" in Quick Fields and input the value into the date field, it should look something like the attachment called "Document field formatting.png"

This will make your date come out as 12/31/2014. So an input of "12312014" will transform into "12/31/2014". As mentioned, it is always a good habit to test your token in the "test value" box to see what the result is. 

However if you want your date to look like 12-31-2014, you will need go into your Laserfiche Administration Console. You can create a new date field. Its properties should look like the attachment "date field properties". That will make your date look like 12-31-2014 rather than 12/31/2014. If you wish to use this are your date format, then don't forget to add that field into your Quick Fields "fields" tab!

Hope this addresses your issues regarding the DOB token.

Regards,

Andrew