I was handed the grunt task of creating a knowledge base that will take in PHP code, read the classes and create a KB of methods, properties, etc.
My first thought was to read in each file, and then filter out the classes w/regex and place them in the DB for future use.
Then I thought, man that's a lot of regex... so I'm debating on loading each file w/include during the indexing and introspect each class from get_declared_classes. The latter seems like it would be less error prone.
Any thoughts? I have no one to bounce ideas off today, so I figured maybe I could pick your brains.