Beyond the Hoc Language

As mentioned in the introduction, the abilities of the hoc language are sufficient for many typical tasks in physical analysis. There are, however, special cases for which the speed and/or the interpreter language capabilities are not sufficient.

For these cases, there are two ways of enhancing YODA's abilities while mainly retaining the comfortable features.

• One way is to add new functions to the set of built-in hoc library routines. After adding own, customized routines, they are indistinguishable from the ''native'' hoc functions. They can be used to perform lengthy calculations at full, compiled speed or memory-greedy algorithms with complex datastructures.. Since they are written in C (C++), they may use the full spectrum of the language and all libraries available. See section 7.1 on how to to add new functions to the hoc language. The interface to and from these functions is still restricted by the hoc language, however.

• Introduce own hard callbacks into YODA. This means to write one or more complete C++ classes (derived from a basecall providing the framework), which perform arbitrary operations on the incoming data. There is no restriction on how to operate on the data, and the selection criteria (when to process and when to ignore data) is completely in the user's code. On the other hand, YODA' principle framework is preserved. These callbacks have full access to YODA's histograms and cuts, and their appearance to the user is like a user task (which cannot be loaded and unloaded, however). See section 7.2 on how to implement such a class.