Component accumulator in DAODs - part 3b: common trigger matching

This MR brings the trigger matching designed for DAOD-level analysis into the component accumulator setup. It consists of two parts:

(1) Run 3 setup: largely this work has already been done on the trigger analysis side and this MR involves unwrapping the hidden CA configs. It makes one change: instead of using the method AddRun3TrigNavSlimmingCollectionsToEDM to add the relevant branches directly to the stream, a new method AddRun3TrigNavSlimmingCollectionsToSlimmingHelper now sets them in a slimming helper which is itself responsible for building the overall output list and then passing that to the stream config. The old method is left alone to ensure compatibility with the non-CA version.

(2) Run 2 setup: this has involved replacing the TriggerMatchingHelper class with normal config scripts for the TriggerMatchingTool. These in turn use the TrigDecisionTool which has an existing config script. Note that a simple helper class is still used to obtain and pass on the list of triggers, which is needed in the main config for the slimming, and also to configure the matching itself. This helper is established in the main config and then passed down.

Note that this MR does not impact any ongoing production because it only touches the new configuration, run from the Derivation_tf.py transform, which isn't yet in use.