Version:            (using KDE KDE 3.5.2)

For matching against fields containing addresses, you have a number of operators to choose from: "is equal to", "contains", "matches regexp", "is in address book in category", and simply "is in address book".

For "is equal to" to be true, the whole content of the field must match exactly, but the content can vary without any real significance.

"contains" matches any substring, meaning that "@example.com" also matches "@example.com.tw", while "@example.com>" only matches if the address is actually in brackets, which is not necessarily the case. In practice this may rarely be a problem, but in some cases you may have to resort to a regexp.

Matching against address book requires you to create a new category specifically for one filter, which can lead to a lot of categories.

And finally, regular expressions are unnecessarily complicated for many users.


Therefore I propose a simpler address matching mechanism, which parses the field(s) in question for the effective address(es) instead of simply doing a substring search, and which allows for at least the simple * wildcard. Examples:

"*@example.com" matches all addresses within the example.com domain.
"root@*" matches all addresses with local part "root".
"*.example.com" matches all addresses within subdomains of example.com, or List-Id's ending with example.com.
"kde-*@kde.org" matches several KDE mailing lists.