Improve API support for nonforking HAF applications

Even current HAF version supports nonforking application flow (by using immediate detach after context creation), it would be helpful to provide better support for this case at API level.

Initial proposal of this solution is to add to the hive.app_create_context new parameter:

hive.app_create_context( _name hive.context_name, _is_forking BOOLEAN = TRUE);

By passing _is_forking = FALSE, HAF would create an nonforking application, which would not create triggers nor shadow tables like also be tied to irreversible block. Application context specific views, also shall have simplified definition and should be based on irreversible data.

All further application flow should be enclosed in standard block processing loop, by subsuequent calls to the app_next_block API which should provide another range of irreversible blocks to be consumed.