General Information

The basic idea behind a plugin flow is as follows;

• Plugin is loaded.
• Plugin registers event handlers for any events it cares to know about.
• BZFS runs.
• BZFS calls installed event handlers as the event conditions happen.
• Code in plugin handlers modify the data passed in with the event, or perform new actions based on the event using other API functions.
• BZFS returns from the event call and uses any modified data.
• BZFS runs, performing any event calls as they happen, until it finally terminates normally.
• Plugin is unloaded when bzfs is quit, or plugin is manually unloaded using the /unloadplugin command.
• Plugin unregisters any event handlers it had installed.

This allows for a plugin to do many different things depending on what events it hooks itself into. Plugins can hook into any number of events they wish and are allowed to do anything the developer wishes to do. A plugin does not have access to 100% of the internals of BZFS. It must access it by using the BZFS API set of functions as described below. Most of these API functions will directly map to internal bzfs functions. In fact there are many cases where internal features of bzfs use the same API functions that plugins can use. Plugins may also link in any external library and call any external functions or code that they wish.

As always open source projects are evolving. If you do not see an API function that you need, please let us know and we can try to add it for you, or code it in yourself, and submit it as a patch.