-
Notifications
You must be signed in to change notification settings - Fork 759
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Use async functions within request handling code #1542
Comments
I'm actually a bit convinced the stack trace breakage was at one point being done intentionally from this utility function, at least for some methods. While if you do things in the most straightforward way, retries will lead to deeper callstacks, there is another option that is also available and easy to express with I might be totally offbase on the purpose of that utility function or even on how much using the |
@yejia-stripe sorry for taking a while to get back. Unfortunate the workaround does not work for usages around async iteration:
^ this code snippet doesn't work with the proxying (I imagine because the list implements both async iteration and a standard promise mechanism). |
After working on this more, and trying to get the proxying to work, I found that using the proxy means that To be honest this is a really frustrating issue, doubly so because the cleanup job is straightforward, and because full stack traces are such a basic expectation for debugging issues. I understand everyone has their priorities but please consider this a very enthusiastic whine about this library's usability. |
Thanks for getting back to us @rtpg |
Hey @rtpg Thanks for your patience. |
thank you for the comment! I'm glad to hear it, and I think that it will make many people happy |
Is your feature request related to a problem? Please describe.
Currently, there is a lot of
setTimeout
s and other non-async-y code that means that we cannot take advantage of V8's async stack trace feature. This means that we lose out on stack traces, the bare minimum that one would want for debugging an issue in production.While support for the async stack trace feature came in Node 12, support for the async/await syntax (which would allow for users of newer node to get a stack trace) has been present from Node 7.
Describe the solution you'd like
At the very least, please use
async
/await
syntax within the core request calling methods (StripeResource
,StripeMethod
), and remove usage of rawsetTimeout
, instead doing things likeawait Promise((resolve, reject) => setTimeout(reject, 0))
to make sure that async stack traces are maintained.Doing this pervasively should make the code easier to manage, reduce a lot of indentation, and be functionally equivalent given how this all works.
Describe alternatives you've considered
Monkeypatching is theoretically possible, and there are options to capture error traces myself, but that involves creating spurious error objects all over and increasing difficulties of updating
stripe-node
.Additional context
#1066 is an example of this issue
The text was updated successfully, but these errors were encountered: