Improve access to track container in tau track tools

This is to improve how the track related tau tools are called by the algorithms. Previously we used a link from the tau to the track container to access the track container in the tool. This required a few extra steps to do. Now, the algorithm passes the container via the execute function, as is done with the rest of the r22 tau code. The base classes and trigger tau algorithms were updated with this too. The old method was also causing problems when trying to compile tauRec in the analysis base release for r22.

While tidying the algorithms I changed the if statements for the execute functions to use tool->type instead of using the tool->name (which depends on what is set in the python config file).

When testing I also found that the default file for the RNN track classifier was not available. I have uploaded a new calibration file and updated the default flag (although this tool is not scheduled by default).