[Lash-L]

Add pure cursor and page planners for V1 dynamic map elements, following #1001 and #1005. Track the accepted generation and logical offset, then return explicit decisions for appends, replacements, empty generations, stale or gapped pages, and bounded generation-reset retries. An overlapping path page contributes only its unseen suffix; other element overlaps are rejected.

Port the element-level portion of the existing local planner, with signed/unsigned 32-bit nonce comparisons, the merged result == 0 or count == 0 acceptance rule, and checks that malformed pages cannot advance a cursor. Whole-map orchestration, byte merging, and device polling follow separately.

Also address Allen's #1005 review: explain how incremental updates keep the displayed path and robot position current, and rename the model/test modules to dynamic_map.py and test_dynamic_map.py under their existing V1 directories.

Validation:

• Focused planner/model/block tests: 149 passed.
• Full suite on Python 3.14: 2,067 passed, 53 xfailed, 11 deselected; 99 snapshots passed.
• All pre-commit hooks passed, including Ruff, Mypy, and Codespell.
• Replayed four saved response cases: path overlap, pose-only path, empty element, and element-generation mismatch. Input captures remained unchanged; capture contents are not included in this PR.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota. Try requesting a review at a lower effort level.

[codecov]

Codecov Report

❌ Patch coverage is 98.73418% with 6 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
roborock/map/dynamic_state.py	97.05%	3 Missing and 3 partials ⚠️

Files with missing lines	Coverage Δ
roborock/data/v1/__init__.py	100.00% <100.00%> (ø)
roborock/data/v1/dynamic_map.py	100.00% <ø> (ø)
tests/data/v1/test_dynamic_map.py	100.00% <ø> (ø)
tests/map/test_dynamic_state.py	100.00% <100.00%> (ø)
roborock/map/dynamic_state.py	97.05% <97.05%> (ø)

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[allenporter]

whats the key?

[Lash-L]

key is the consts above

[Lash-L]

Automated code review by Gemini 4 Argon in Antigravity using our /review (/code-review) skill (.agents/skills/code-review/SKILL.md / AGENTS.md).

Summary

Clean, pure state-transition planner for V1 dynamic map element updates.

• Layer 1 (Architecture & Boundaries): roborock/map/dynamic_state.py is purely functional/immutable (frozen=True dataclasses) with zero RPC or transport coupling.
• Layer 2 (Data Models & Wire Mapping): Block header validation (_PATH_HEADER_LEN = 20, _STANDARD_NUM_HEADER_LEN = 12), path/mop flag count agreement, 32-bit signed/unsigned nonce comparison, and overlap trimming on element 3 match the dynamic map wire format. See inline notes on PAGE_LIMITS and an empty-page edge case during replacement_pending.
• Layer 3 & 4 (Style & Error Handling): Structural block problems raise RoborockException in low-level helpers and are converted into PageOutcome.SKIP reasons by validate_page_content so malformed pages never advance a cursor.
• Layer 5 (Testing Standards): Thorough parameterized table tests in tests/map/test_dynamic_state.py; see inline note for the 3 uncovered lines reported by Codecov.

[Lash-L]

Layer 2 & 5 (PAGE_LIMITS & Coverage):

1. PAGE_LIMITS includes 32: 32, 37: 50, 38: 50, and 58: 50, which are not in MERGEABLE_ELEMENTS (so page_limit() always returns 0 for them), while other dynamic element IDs that have default request limits (1: 1, 5: 500, 17: 4096, 26: 1, 63: 50, 64: 50) are omitted. Either scoping PAGE_LIMITS strictly to MERGEABLE_ELEMENTS or including the full default table (1, 3, 5, 6, 11, 15, 16, 17, 21, 24, 25, 26, 27, 29, 30, 31, 32, 36, 37, 38, 42, 44, 45, 46, 50, 51, 52, 53, 58, 60, 63, 64) would make the table's role clearer.
2. page_limit() is currently only called in tests with unsupported element IDs, leaving line 77 (return PAGE_LIMITS.get(element_id, 0)), ElementState.is_complete (line 110), and element_record_count when a _STANDARD_NUM_ELEMENTS block is absent (line 237) uncovered.

[Lash-L]

If state.replacement_pending is True (at offset == 0) and an accepted page at start == 0 reports length == 0 and max_len == 0 (accepted_total == 0), this returns PageOutcome.NO_RECORDS with known_total=0, replacement_pending=True, and replace_content=False. Because offset == known_total == 0, no further page will be requested and the previous generation's cached content is never cleared. When accepted_total == 0 and state.replacement_pending (or state.known_total > 0), should this clear replacement_pending=False and set replace_content=True to match reconcile_announcement?