Currently, if you have segments that are alphanumeric values, they will show as repetitions with the repetition filter even though the text is different. Ideally, these would not show in the repetition filter as the text is not the same even though they are being processed as alphanumeric values with the language processing rules.

There is the option to remove alphanumeric values from the language processing rules to make sure these values do not show in the repetitions filter, but when you do that you remove the benefits of how Trados can handle the values within segments.

If the repetition filter was able to look at the actual text to know what to filter, that would be helpful.