Remove optional return types from session history methods

[0xZaddyy]

After PR #1014 removes the Uninitialized variant and enforces that every SEL must contain at least one event, session history methods no longer need to return Option<T> since it guarantee's data availability

Changes

• Sender methods: fallback_tx() and pj_param() now return non-optional types
• Receiver methods: pj_uri() and session_context() now return non-optional types
• FFI layer: Updated method signatures to match

Closes #1046

Pull Request Checklist

Please confirm the following before requesting review:

• I have disclosed my use of
  AI
  in the body of this PR.
• I have read CONTRIBUTING.md and rebased my branch to produce hygienic commits.

[coveralls]

Pull Request Test Coverage Report for Build 17649247435

Details

• 14 of 14 (100.0%) changed or added relevant lines in 3 files are covered.
• 4 unchanged lines in 2 files lost coverage.
• Overall coverage increased (+0.02%) to 86.041%

---

Files with Coverage Reduction	New Missed Lines	%
payjoin/src/core/receive/v2/session.rs	2	97.59%
payjoin/src/core/send/v2/session.rs	2	95.65%

Totals
Change from base Build 17593913920:	0.02%
Covered Lines:	8235
Relevant Lines:	9571

---

💛 - Coveralls

[arminsabouri]

Thanks for creating a clean diff! This was easy to review. I have one issue with a comment that was changed.

Besides that we should remove default impl from SessionHistory and replace it with a pub(crate) new method which accepts a non-empty iter of session events. Otherwise we could have a builder pattern but that may be overkill. Lmk what you think.

Point being the application should not have the visibility to create an instance of SessionHistory. It can only be obtained by replaying events -- which guarantees that we have at least one event.

[0xZaddyy]

Besides that we should remove default impl from SessionHistory and replace it with a pub(crate) new method which accepts a non-empty iter of session events. Otherwise we could have a builder pattern but that may be overkill. Lmk what you think.

looks good to me as this prevents a scenerio where SessionHistory could exist without events.

[arminsabouri]

Code ACK
Just one more case of the same rustdoc being modified

[arminsabouri]

Another case of the same comment being modified. Must have missed this in my first review.

Suggested change

	/// Fallback transaction from the session
	/// Fallback transaction from the session if present

[0xZaddyy]

Code ACK Just one more case of the same rustdoc being modified

thank you @arminsabouri
i've fixed the doc change

[arminsabouri]

utACk

[DanGould]

I'm somewhat skeptical of these methods on SessionHistory at all. It seems like fallback_tx is not used (maybe just because our payjoin-cli reference is incomplete?) and pj_url is used for an error handling hack (below) that I think we're gonna be able to internalize. Auditing the use of this is gonna have to be done.

                let pj_uri = session_history.pj_uri().extras.endpoint().clone();
                let ohttp_relay = self.unwrap_relay_or_else_fetch(Some(pj_uri)).await?;
                self.handle_recoverable_error(&ohttp_relay, &session_history).await?;

[DanGould]

My comment is not blocking merge tho

[arminsabouri]

I'm somewhat skeptical of these methods on SessionHistory at all. It seems like fallback_tx is not used (maybe just because our payjoin-cli reference is incomplete?) and pj_url is used for an error handling hack (below) that I think we're gonna be able to internalize. Auditing the use of this is gonna have to be done.

                let pj_uri = session_history.pj_uri().extras.endpoint().clone();
                let ohttp_relay = self.unwrap_relay_or_else_fetch(Some(pj_uri)).await?;
                self.handle_recoverable_error(&ohttp_relay, &session_history).await?;

OH1H The fact that its not used is indicative that the refrence is incomplete and once we complete it we may find that we dont need exact method of SessionHistory. OTOH I think there is always going to be a need to fallback based on user preference -- this could also be part of the reference.