coded wrote:
Quote:
|
Add wildcard ignores for usernames and chat room spam. I noticed bot names are progressively becoming harder to catch with wildcard ignores, because their usernames are starting to mimic legit usernames, and containing less patterns. However they still spam rooms with the same phrases, thus it could be easy to filter them with a spam wildcard, or something.
|
Wildcarding via standard RegExing has been added for IDs currently with character sequences starting with, ending with or even containing the strings found in the rule set (no matter the position in which they're found). Chat messages will have RegEx support for pattern matching as well as an option to block messages based on literal strings (which you can specifically supply) that are found in messages. I'll probably add nickname and location RegExing also as they could also be useful in blocking users that have that information available in their profiles.
A good example of a literal string match would be: "Too many bots, PM me to talk."
An example of a RegEx rule being fired off based on the *c2c* rule: "anyone want to c2c?"
Both pattern matching and allowing for literal string input for filtering have the potential to block lots of junk when used in combination. The 'flag as bot' system YMLite uses could also be incorporated at some point (cross-referencing flagged bot messages and using them for comparison purposes).