Catch Payment Processor Exception if thrown when registering via bac…

[seamuslee001]

…k office participant registration

Overview

This catches any Exception thrown by the Payment Processor as it doesn't appear to be caught anywhere else and then sets a standard style status message

Before

Payment Processor Exception not caught

After

Payment Processor Exception caught

ping @eileenmcnaughton

[civibot]

(Standard links)

• If this is your first pull-request for CiviCRM, please browse CONTRIBUTING.md for information about the development and testing processes.
• If you are reviewing this pull-request, you may wish to consult the test sites and the Review Standards (long template, short template).

[mattwire]

@seamuslee001 There seems to be quite a bit of inconsistency in how we handle the call to doPayment() - I do agree with catching the exception, but looking at other points in the code I think the right thing to do is call CRM_Contribute_BAO_Contribution::failPayment() and then call CRM_Core_Session::singleton()->setStatus($e->getMessage()); (ie without adding any extra text).

[seamuslee001]

@mattwire i'm not sure if we have created a contribution at this point in this code so not sure if we can call failPayment() but will update on the other part then.

[eileenmcnaughton]

We should have created a contribution & I'm pretty sure on this form we have - the front end event form is still a gap

[seamuslee001]

@eileenmcnaughton @mattwire having done a bit of debugging i think the contribution is actually getting created at L1292 after we have done the payment when we call https://github.com/civicrm/civicrm-core/blob/master/CRM/Event/Form/Registration/Confirm.php#L960

[eileenmcnaughton]

ug - I've started digging into this event stuff due to the paypal checkout flow issue - but so far have only put up minor cleanups like #14917

[mattwire]

:-( @seamuslee001 Ok, can you add the failPayment in commented if it definitely does not with a link in the comments to https://lab.civicrm.org/dev/financial/issues/53 or if it makes sense wrap failPayment in an if (!empty($contributionID)) { failPayment } with the same comment? At least we can work towards consistency - even if we can't quite get there yet!

[seamuslee001]

@mattwire @eileenmcnaughton ok added comments and made changes now

[eileenmcnaughton]

OK - I wrote a unit test (will push up separately) & stepped through this in it & agree it makes sense as a change. I'd rather see us doing the redirect in the postProcess rather than submit & fix the order of events - but this does fix a problem where the code intended to be called on a failed payment isn't