Cleanup the module integrations.f90

[Juergen Reuter]

This still looks like "Kraut und Rueben" as a left-over from 2.1.1. WK promised to do this.

[Juergen Reuter]

After the commit by WK yesterday, is there still something left to do here??

[kilian]

cleaned up, but there is a potentially missing feature: handling processes without matrix element, which should be treated as zero.

Also wondered about a few lines regarding helicity selection. Probably obsolete?

[Juergen Reuter]

Ok, I checked the reset_helicity_selection setup: In 2.1.1: integration_setup_process -> process_reset_helicity_selection -> hard_interaction_reset_helicity_selection -> hard_interaction_data_reset_helicity_selection. This last subroutine calls the abstract interface proc. pointer prc_reset_helicity_selection which calls the corresponding routine in the O'Mega matrix element. In 2.2.0: integration_setup_process -> integration_t%process_t%init_component -> integration_t%process_t%process_component_t%init -> (in case of being prc_omega_t) prc_omega_activate_parameters -> prc_driver%reset_helicity_selection which is again the proc. pointer to the O'Mega matrix element mentioned above.

WK, satisfied?

For the processes without (valid) matrix element, I will have a look ...

[Juergen Reuter]

After discussion with WK, the issue with the reset_helicity_selection is settled. Dummy integration for processes with vanishing matrix element are reintroduced in r5441. This could be improved to the case of processes, where only some of the components vanish, but some don't. (E.g. g, g -> Z, Z vanishes at tree level, but not at NLO). I leave this open for discussion for the next meeting. All has been cleaned up in ingegrations.f90. Closing.