A Case for Reflection 1
W. E. Brown, Philippe Canal, Mark S. Fischler, Jim Kowalkowski, Pere Mato, M. Paterno, S. Roiser, L. Tuura · 2005
The HEP (high-energy physics) community has long embraced C++ as its primary language for obtaining, saving to mass storage, restoring from mass storage, and analyzing large amounts of physics event data. Current experiments as well as experiments under development make extensive use of large bodies of code written in C++. Several ad hoc toolkits and libraries have been written to support the physicist user base in its efforts to process petabytes of scientific data. We believe that many of these activities have been significantly more difficult than they might have been. We attribute much of this difficulty to the lack of a standard means to obtain information from a C++ language processor regarding a compiled piece of code, a facility historically termed reflection.