I need to enabled duplicate detection and blocking on text fields that only checks the alphanumeric characters in the string and is exact for those characters. The available fuzzy algorithms do not seem to offer this variation.
I can have two strings like HISTORY-101 and HISTORY 101 that must be identified as duplicates while not matching HISTORY-201.
The various fuzzy algorithms indicate that they would see HISTORY-201 and HISTORY 101 as equivalently distant from HISTORY-101 (i.e. one character change away).
It seems like my only solution may be to copy the original string to a new field that strips out non-alphanumeric characters, and then apply Exact matching as the matching method. Since Formula fields are not eligible for matching rules, that means more automation and an extra field to accomplish this. Any suggestions for a simpler way? Am I misinterpreting the capabilities of the matching algorithms? And why do I not even get just Fuzzy as a matching option on custom object fields?