source-braintree-native: concatenate search limit error message

[Alex-Bair]

Description:

The Braintree SDK's API calls are all synchronous, so they block the event loop, meaning only a subset of streams can make progress at a time. Adding await asyncio.sleep(0) to the top of the various fetch functions allows other streams to make progress since these functions now yield to the event loop.

Using await asyncio.sleep(0) within the connector-specific code for this purpose is a temporary measure until I refactor the connector to make asynchronous API requests.

I also fixed an issue with the search limit error message where the f-strings were not concatenated together & the full error message wasn't getting displayed.

Workflow steps:

(How does one use this feature, and how has it changed)

Documentation links affected:

(list any documentation links that you created, or existing ones that you've identified as needing updates, along with a brief description)

Notes for reviewers:

(anything that might help someone review this PR)

---

This change is 

[williamhbaker]

I don't think this is going to do what you're after, which I assume is to have multiple calls to the Braintree API in progress at a time, concurrently. The calls to braintree_gateway.transaction etc. are not async, so they will still block the event loop. I may be mis-reading / mis-interpreting what you're after here though, and adding this isn't going to break anything, so just FYI.

[Alex-Bair]

I agree, this isn't intended to try and let multiple calls progress concurrently. I won't be able to achieve that without making the API calls asynchronous. My main concern was that if a stream isn't caught up to the present, it would continue trying to get caught up by immediately reinvoking fetch_foobar & blocking any of the other streams from making API calls since there aren't any places within fetch_foobar that yields to the event loop. I also thought this might prevent the SIGQUIT mechanism from working if one of the streams was trying to catch up to the present. But I tested locally, and sending a SIGQUIT still works without these await asyncio.sleep(0) lines, so there's still an await in the CDK that's yielding to the event loop and preventing these issues (I think it's here).

I'll update the PR to remove these await asyncio.sleep(0) lines since they aren't needed. I'll keep the error message fix and merge that.

[williamhbaker]

LGTM