BounceDetector 1.1 Released

I updated my BounceDetector project over at RIAForge last night to version 1.1. You can use this simple two-file package to analyze the content of bounced email and determine severity and cause using an extensive production-tested signature database. Paired with CFPOP, this lets you easily fulfill Zawinski’s Law. New in this release:

• Changed reFind() to find() for a (substantial) speed increase as well as more precise signature matching. The signatures in the database weren’t intended as regular expressions previously. I began the process of converting the many signatures into fewer RegExps but found that it made it more difficult to determine if a signature was in the database which would make future updates more difficult. The find() method is fast enough to make improving the reFind() method unnecessary.
• Removed some duplicate signatures
• Added new signatures collected from our production system
• Reorganized signatures slightly to put common numeric-based reasons at the top which will increase the likelihood of an early match and exit from the loop.

I renamed the files as well to reduce the likelihood that it would conflict with your existing systems. The goal was to make it easier to use as an svn:external which is how I’m including it in my applications now.