Fix payment_intent_unexpected_state error when voiding payment intents

[jfturcot]

Solves #64

If a payment intent has been created but the order is never completed, Stripe will eventually time out the intent and change its state. When a new payment intent is created, it first tries to void all the previous uncompleted payment intents of the user and this would break with the Stripe error payment_intent_unexpected_state since Stripe's state machine does not allow that change.

[codecov-commenter]

Codecov Report

Merging #115 (fe96667) into master (a57ff9d) will decrease coverage by 0.76%.
The diff coverage is 83.33%.

@@            Coverage Diff             @@
##           master     #115      +/-   ##
==========================================
- Coverage   95.45%   94.69%   -0.77%     
==========================================
  Files          15       15              
  Lines         352      358       +6     
==========================================
+ Hits          336      339       +3     
- Misses         16       19       +3     

Impacted Files	Coverage Δ
app/decorators/models/spree/payment_decorator.rb	90.90% <83.33%> (-9.10%)	⬇️
...dels/solidus_stripe/address_from_params_service.rb	96.42% <0.00%> (-3.58%)	⬇️
.../models/spree/payment_method/stripe_credit_card.rb	97.97% <0.00%> (-1.02%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a57ff9d...fe96667. Read the comment docs.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It might be closed if no further activity occurs. Thank you for your contributions.

[kennyadsl]

Thanks for your contribution @jfturcot!

I was wondering if there's a way to add this logic to SolidusStripe::PaymentMethod instead, in order to avoid overriding the payment class (which will be used by other payment methods as well).

The class I referenced has Spree::PaymentMethod as parent class, which implements a generic try_void method, which is called when the subclass doesn't implement one.

What do you think? Would it make sense to move this logic there?

[kennyadsl]

Hello @jfturcot! Are you still interested in moving this PR forward? Otherwise, we can probably continue the work to fix the issue. Thanks!

[jfturcot]

@kennyadsl Sure! Sorry I have been very busy lately, but I would love to get this merged. I will take a look at it over the weekend.