You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We won't break old functionality, but adding a generic source and exact term that parses a string to select which term to evaluate would be nifty. This would then look like,
No rubber to road yet, so we can plan this out, happy for any feedback/suggestions. Under the hood I'm inclined to build a map between string and a pointer to source/exact functions, e.g. "u":"masa_eval_exact_u()", after which this method calls the appropriate function evaluated at the points it was given. The trick is that we have man different methods due to the numerous overloaded functions for the various dimensionality we might have. We could circumvent this by pointing the map to a container of all the available functions for that variable and then selecting only the one that has the appropriate dimensionality, but I suspect there is a more elegant solution available.
This was an old MASA svn ticket that is being resurrected for @pbauman and @vikramgarg.
The text was updated successfully, but these errors were encountered:
I'm inclined to output a message to the log (e.g. "MASA ERROR: No such source term") and call masa_exit(1) -- this is how we handle the user attempting to initialize a non-existent solution.
If the user has enabled exception handling in masa, they can catch this, otherwise it calls exit and ends.
Presently, MASA requires calling a specific method for each source term, e.g.
etc.
We won't break old functionality, but adding a generic source and exact term that parses a string to select which term to evaluate would be nifty. This would then look like,
No rubber to road yet, so we can plan this out, happy for any feedback/suggestions. Under the hood I'm inclined to build a map between string and a pointer to source/exact functions, e.g. "u":"masa_eval_exact_u()", after which this method calls the appropriate function evaluated at the points it was given. The trick is that we have man different methods due to the numerous overloaded functions for the various dimensionality we might have. We could circumvent this by pointing the map to a container of all the available functions for that variable and then selecting only the one that has the appropriate dimensionality, but I suspect there is a more elegant solution available.
This was an old MASA svn ticket that is being resurrected for @pbauman and @vikramgarg.
The text was updated successfully, but these errors were encountered: