A lot of Addex features rely on your ability to put extra information onto address book entries. A really simple example is its generation of procmail rules. If it's going to filter mail from Mom into the "family/mom" folder, I need to be able to tell it that I want that kind of filtering. Address book programs, though, are not usually designed to store all the information Addex wants.
An address book entry in Addex only has a few attributes:
The other attributes could all be set up with some kind of attribute declaration blah blah whatever. Instead, they use Mixin::ExtraFields::Param. It's about one line of code and it just goes.
The first three data are pretty easy to get out of most address books. After all, they're some of the data that address books are meant to hold. The "bunch of other attributes" data are a little iffier. These attributes are called the entry's "fields." Here are some things that you can use fields for:
Using the Apple address book plugin, fields are gathered by looking at the "notes" section of the entry, which would be really useful for keeping actual notes about contacts, if the editing UI wasn't so awful. Addex looks for lines that look like this:
For example, the entry for my brother in law has this in its notes:
Using the abook plugin, there are even fewer ways to add arbirary data, so you have to configure plugin to relate the limit number of extra abook fields with Addex "fields."
What I like about fields is that they're just so stupid. I can add arbirary snippets of data to an entry for any plugin to consume. By avoiding all the problems that won't really come up (like namespaces, datatypes, escaping) I've made it really easy for myself to add new features.
I may write one more little piece about the (current) Addex workflow, if I decide it's worth the time. Then I'll be done talking about Addex for a while, at least until I write some fun new plugins.