I would like to know in which language a mail is. The customer didn’t want to buy content analyzer so I must use screening rules. Is this possible? As far as I found out the screening rules only return true or false. Can apply more than one rule for a mail and get a percentage of these rules being true/false?
In other words if I get the most 50 words used in english, can I apply a rule to see if some of them are used?
You’re correct - screening rules return true or false. Even MultiScreen object won’t help as it always returns just single matching rule (if any).
In the past, I played a bit with language identification/guessing engines and the results were quite interesting. It should be possible to use such engine and develop a media server (ESP) that will analyse the text and returns identified language.