Expand RTC payout parser variants

[Yasuno-5555]

Fixes #6645

Summary:

• expand the PR body parser to accept payout wallet/address variants
• accept miner ID lazy-pay variants alongside wallet directives
• fall back to contextual RTC address extraction when payout labels are present
• add unit coverage for the new parser cases

Verification:

• python3 .github/actions/rtc-auto-bounty/test_award_rtc.py
• python3 tests/test_award_rtc_workflow.py

[github-actions]

Welcome to RustChain! Thanks for your first pull request.

Before we review, please make sure:

• Non-doc PRs have a BCOS-L1 or BCOS-L2 label
• Doc-only PRs are exempt from BCOS tier labels when they only touch docs/**, *.md, or common image/PDF files
• New code files include an SPDX license header
• You've tested your changes against the live node

Bounty tiers: Micro (1-10 RTC) | Standard (20-50) | Major (75-100) | Critical (100-150)

A maintainer will review your PR soon. Thanks for contributing!

[Le-Who]

Thanks for adding coverage for the explicit variants from #6645. I think this still needs one tightening pass before merge because the new directive path can return an invalid recipient string for a common PR-body shape.

The new _DIRECTIVE_RE captures (\S.*?) through the end of the line and resolve_wallet_from_pr_body() returns that full capture after only trimming a trailing comma. That means a body like:

Payout address if accepted: RTC4730a64f821a590972e8b37781fae5d568b5865c (preferred wallet)

would return:

RTC4730a64f821a590972e8b37781fae5d568b5865c (preferred wallet)

instead of the actual RTC address. The contextual fallback already extracts the exact RTC[0-9a-f]{40} token, but it never runs for directive lines because the directive match returns first. This can reintroduce the same manual-pay routing problem whenever contributors add a parenthetical note, Markdown punctuation, or other text after the address.

Please either make RTC-address directives extract the _RTC_ADDRESS_RE token from the captured value before returning it, or split the handling so RTC address labels return only the address while lazy-pay miner-id labels keep returning the full miner id. A focused regression test for an annotated Payout address if accepted: line would cover the failure mode.

[MolhamHamwi]

Reviewed the payout parser expansion in .github/actions/rtc-auto-bounty/award_rtc.py and the new parser coverage in .github/actions/rtc-auto-bounty/test_award_rtc.py.

Two specific observations:

1. The broadening from the old wallet/.rtc-wallet regex to explicit payout labels is a useful compatibility improvement for real claim bodies, and the new tests cover the common Payout address if accepted, RTC Wallet, contextual payout-address, and miner-id variants. I also ran the two affected test commands locally (python3 .github/actions/rtc-auto-bounty/test_award_rtc.py and python3 tests/test_award_rtc_workflow.py), and they passed.

2. I agree with the existing requested-change path on annotated directive values: _DIRECTIVE_RE returns the whole post-colon capture before the contextual RTC-token extraction can run. For address-style labels, a line such as Payout address if accepted: RTC4730a64f821a590972e8b37781fae5d568b5865c (preferred) would currently return the address plus parenthetical text, while the contextual branch would have returned only the RTC token. A focused regression for an annotated address directive would protect the intended manual-pay routing behavior.

Why I liked it: this PR targets a real payout-automation edge case and keeps the parser small instead of adding a heavier body parser.

I received RTC compensation for this review.

[perezjefry53-cpu]

Good improvement in expanding the wallet directive parser. A few observations:

1. Edge case: If someone writes Payout Wallet: (capital W) vs Payout wallet: (lowercase w), the current regex handles this via re.IGNORECASE which is good.

2. Potential ambiguity: If a PR body contains both wallet: RTCxxx AND miner ID: some-id, the function will return the first match. Consider if there should be priority ordering (e.g., explicit wallet directives should take precedence over miner IDs).

3. Line-based matching: The switch from multi-line regex to line-by-line matching is a good change that makes the code more readable and easier to debug.

4. Test coverage: The new test cases cover the added directive formats well. Consider adding a test for when both a wallet directive AND a miner ID directive exist in the same body to document the expected priority behavior.

Overall the implementation is clean and well-structured.

[jaxint]

Thanks for contributing! 🦀

Code review completed. Nice work! 🚀

---

Wallet: AhqbFaPBPLMMiaLDzA9WhQcyvv4hMxiteLhPk3NhG1iG

[Yasuno-5555]

Address-style directives now extract the RTC token before returning, so annotated lines like Payout address if accepted: RTC... (preferred wallet) no longer leak the trailing note into the payout target.

Targeted verification:

• python3 .github/actions/rtc-auto-bounty/test_award_rtc.py
• python3 tests/test_award_rtc_workflow.py

[sanphandinh]

Re-reviewing current head 62b0778e: the RTC-address annotation case is fixed, but I think the miner-id branch now preserves the same class of routing bug for lazy-pay recipients.

For non-miner labels, the parser extracts only the RTC[0-9a-f]{40} token from an annotated value. For miner labels, it returns the whole captured value, and the new test explicitly locks in:

body = "miner ID for payout if accepted: my-miner-id (fallback worker)\n"
assert resolve_wallet_from_pr_body(body) == "my-miner-id (fallback worker)"

That means a very common annotated claim line will send the payout automation a recipient string that is not the actual miner id. This is especially likely because the accepted prose variant is long (miner ID for payout if accepted:), and contributors often append notes like (fallback) or (preferred). The earlier address bug was fixed because extra prose breaks payout routing; the same reasoning applies here unless the downstream lazy-pay system intentionally accepts spaces and parentheses as part of a miner id.

Suggested fix: define the allowed miner-id token shape and extract only that token from miner-id directives too, or reject annotated miner-id values with a clear error so maintainers can ask for an exact recipient. A regression should assert that miner ID for payout if accepted: my-miner-id (fallback worker) resolves to my-miner-id or fails closed, not to the annotated full line.

Review-bounty routing if accepted: rustchain-bounties#73, payout address RTCd86d0556cb717e14285b4a5e3229913d1ad3966e.

[Yasuno-5555]

Follow-up pushed on current head:

• address-style directives still extract the exact RTC... token
• miner-id directives now strip trailing annotations and return the leading recipient token only, so miner ID for payout if accepted: my-miner-id (fallback) resolves to my-miner-id
• added a focused regression for the annotated miner-id case

Re-ran:

• python3 .github/actions/rtc-auto-bounty/test_award_rtc.py
• python3 tests/test_award_rtc_workflow.py

[FakerHideInBush]

Reviewed the latest head dfb4ad6 after the two earlier parser-routing concerns. The current implementation now handles both annotation classes that matter for payout safety: RTC-address directives extract the exact RTC[0-9a-f]{40} token from annotated values, and miner-id directives extract the leading miner-id token instead of returning parenthetical notes as part of the recipient.

Validation run locally from a clean checkout:

• python -m pytest .github/actions/rtc-auto-bounty/test_award_rtc.py -q -> 67 passed, 9 subtests passed
• python -m pytest tests/test_award_rtc_workflow.py -q -> 3 passed
• python -m py_compile .github/actions/rtc-auto-bounty/award_rtc.py .github/actions/rtc-auto-bounty/test_award_rtc.py
• git diff --check origin/main...HEAD -- .github/actions/rtc-auto-bounty/award_rtc.py .github/actions/rtc-auto-bounty/test_award_rtc.py

I also smoke-tested the key recipient cases directly:

• annotated payout address returns only the RTC address
• annotated miner-id returns only my-miner-id
• contextual payout-address prose returns only the RTC address
• legacy wallet: github-user, still trims the trailing comma

The broad CI / test check is still red, but the changed action parser tests and the workflow-level award tests pass locally, and the earlier review blockers appear addressed on this head.

wallet: RTCe0961d6b54f2fa96db57a373c84d8ad8986153f8

[jaxint]

Great contribution! 🔍 Reviewed and looks solid.

[jaxint]

LGTM! Thanks for contributing to RustChain. Approved.

[jaxint]

Great contribution! This PR looks well-structured and follows best practices.

[jaxint]

LGTM! Thanks for contributing to RustChain. Approved.

[jaxint]

LGTM