Not --all-features test suite

[benalleng]

Got everything compiling, however based on the current cfg flag layout in the integration tests, all the tests are run when v1 is enabled and none of the tests run when v2 is enabled

Closes #1019

Pull Request Checklist

Please confirm the following before requesting review:

• I have disclosed my use of
  AI
  in the body of this PR.

  I had chatgpt help me clean up some clones in the SenderBuilder while i was transitioning over to the NoopSessionPersister

• I have read CONTRIBUTING.md and rebased my branch to produce hygienic commits.

[coveralls]

Coverage Report for CI Build 24687887120

Coverage increased (+0.002%) to 84.504%

Details

• Coverage increased (+0.002%) from the base build.
• Patch coverage: 4 uncovered changes across 1 file (113 of 117 lines covered, 96.58%).
• No coverage regressions found.

Uncovered Changes

File	Changed	Covered	%
payjoin/src/core/uri/v1.rs	108	104	96.3%

Coverage Regressions

No coverage regressions found.

---

Coverage Stats

Relevant Lines:	12926
Covered Lines:	10923
Line Coverage:	84.5%
Coverage Strength:	409.42 hits per line

---

💛 - Coveralls

[benalleng]

based on the current cfg logic in the integration tests no tests run when v2 is the only thing enabled, perhaps someone could point me in the right direction on better splitting up the tests based on features

[DanGould]

My guess is that the v2 tests are relying on the URI builder from v1. We may need another builder that's v2-specific OR to expose this utility without feature gating it. The former seems more appropriate.

let mut pj_uri =
                build_v1_pj_uri(&pj_receiver_address, EXAMPLE_URL, OutputSubstitution::Enabled)?;

[benalleng]

My guess is that the v2 tests are relying on the URI builder from v1. We may need another builder that's v2-specific OR to expose this utility without feature gating it. The former seems more appropriate.

let mut pj_uri =
                build_v1_pj_uri(&pj_receiver_address, EXAMPLE_URL, OutputSubstitution::Enabled)?;

Its only the v2_to_v1 test that requires this setup whereas all the others rely on the usual session to create a pj_uri. I guess we could create a pj_uri with a session like that but then it really wouldn't be a v1 receiver

[DanGould]

Ah, and in that case it makes sense to test only with v1 compiled. Needs further investigation I suppose.

[DanGould]

@benalleng I may have unrequested reviewership by mistake. Is this ready for review?

[benalleng]

It is though with our source code now organized correctly #1019 (comment) we no longer fail on cargo test -p payjoin so before any actual code review I guess we need to ask ourselves if the addition of non --all-features tests still bring joy

This is incorrect we still fail on cargo test -p payjoin on master

[benalleng]

At the moment the line
https://github.com/benalleng/rust-payjoin/blob/d3bb1ab5e74f32b74dd2fb50000b920e514a03c9/payjoin/contrib/test.sh#L11
does not test anything with the v2 feature as all the v2 tests require enabling v1 with the way that the integration tests are organized at the moment.

[DanGould]

Please remind me how you'd like this one to end up? We want to test v2 only and this is the way to do it? Seems like a no-brainer as a code reduction and ehh we actually want to run our tests.

[benalleng]

I think it is good at the moment, but the noop for the v2 feature seems a bit confusing for the integration test with how I organized it. But I guess its better than failing and all the tests still run

[DanGould]

I see you just rebased but we may need another

[benalleng]

Yeah I gave it a try to get the tests properly partitioned but I think the latest commit isn't right