refactor: printer factory, MQTT/FTPS hardening, coverage to 86 - #119
Merged
Conversation
Continues the boundary/error-model arc (#117/#118). Architecture: - RuntimeContext.printer() now uses an injectable factory registered downward from bambu_cli.printer (set_printer_factory); context.py no longer imports printer. Removes the last check_layers ALLOWED edge (context -> printer); ALLOWED is now empty. Correctness / resources: - mqtt_session: lock _pending_payload/_command_issued across the paho network thread; cached-complete-snapshot fast path on silent pushall. - mqtt_tls: drop global socket.setdefaulttimeout mutation; bound connect via client._connect_timeout; configurable, validated MQTTS port. - mqtt_cmd/mqtt_print/mqtt_monitor: teardown in finally (fixes leak on the connect-exception path). - ftps: close the socket if connect/login/prot_p fails. - snapshot/slice: drop unreachable safe_log_error+abort tails after emit_json_error (now typed NoReturn); it already logs and raises. Tests / CI: - +40 targeted tests for the above error paths; raise CI cov floor 83 -> 86 (measured 91.0% Linux). Stronger docs-coverage drift guard. - Docs: api.md support matrix; SECURITY.md residuals marked accepted for 1.0; scoreboard refreshed.
_file_permission_check returns None on win32 by design (POSIX mode bits don't apply — see the Windows secret ACLs residual in SECURITY.md), so the 0o644/0o600 assertions only hold off-Windows. Assert the None path on win32 instead.
DLANSAMA
added a commit
that referenced
this pull request
Aug 14, 2026
Truth-sync only (grades and narrative untouched). Reconciles stale snapshots against the measured post-#119 state: - coverage: 89.1/89.2% -> 90.99% Linux / 90.68% Windows - tests: 1419/1459 -> 1499 non-live (1 live deselected) - statements: 8120 -> 8368 - Python: drop the stale "Linux 3.9" / py39 refs -> 3.10/3.12/3.14 - floor-history / ratchet-headroom margins recomputed for the 86 floor quality-roadmap.md was internally contradictory (current scoreboard vs stale header + floor-history + ratchet paragraphs); those now agree. Historical dated records (the 5b08720 re-verification blockquote) are preserved as-is and a current 2026-08-13 entry appended. docs-consistency test passes. Co-authored-by: DLANSAMA <258674612+DLANSAMA@users.noreply.github.com>
DLANSAMA
added a commit
that referenced
this pull request
Aug 14, 2026
Re-verified the 2026-07-31 post-audit residual table against current main rather than trusting the audit snapshot. Four residuals it lists as open are closed, and one the follow-up notes recorded as closed is not: - S5 closed: utils.py:63-67 is now a shim delegating to jsonio; both spellings redact the audit's own repro identically. No weaker second implementation. - A4 closed: mqtt_port is wired end to end (context -> printer -> mqtt_tls _mqtt_port -> client.connect). Doctor no longer reports a port it ignores. - T1 closed: textual is pinned >=8.0,<9.0 (#115), not <2.0. - A3/Q1/S1/B1/R1/R3 closed by #119, the camera opt-in work, and the TUI merge. - Q4 NOT closed: #116 removed 81 setdefault stubs across 33 files but tests/bambu_test_base.py:26-29 still assigns MagicMock to sys.modules for paho at import time and never restores it, which is also order-dependent. - P1, Q5, Q6, A1, A2, S3, R2, T2/WS4, W3.4 re-confirmed open with current file:line evidence; A1's target moved to bambu_cli.cliparse. Plan headers carried numbers from before #119/#120 and read as live state. Both gameplans now lead with measured current values (1499 non-live, 90.99% Linux / 90.68% Windows over 8368 statements, floor 86, Python 3.10/3.12/3.14) and keep the old figures marked as historical, including the a-plus W1 sizing that was derived from the 89.4% baseline. mutation-baseline.md claimed its per-module table reconciles to the 50.7% headline. It does not: the seven rows are the whole only_mutate scope yet sum to 2010 mutants against a headline 2091, leaving 81 unattributed (37 survived, 41 killed, 3 timeout). Also corrected the "residual survivors should be cosmetic" prediction, which the same file already disproves at 21.8% with 269 survivors, and the stale "re-run mutmut on slicer/output.py" note, which was already done on 2026-08-04. tui-plan.md kept imperative branch instructions after its own "implemented" banner; folded into the historical framing. Also drops AI-vendor attribution from these two plan docs: the audience and source lines, the "pasteable brief for a new <vendor> session" heading, and the trailer rule, which now reads "AI/session attribution trailers" instead of naming a product. References to the local-only agent notes filename are left intact where removing the name would void the rule. Docs only, no code changes. Gates: 1499 passed / 1 deselected, 91.0% total; test_docs_consistency + test_docs_links 25 passed.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Reviewed batch of changes (originally drafted with Grok). Full gate run passed locally:
ruff check,ruff format --check,mypy(clean),bandit(exit 0),check_layers, all smokes, and1499 passedat 91.0% coverage (floor now 86).Architecture
RuntimeContext.printer()now goes through an injectable factory registered downward frombambu_cli.printer(set_printer_factory);context.pyno longer importsprinter.py. This closes the lastcheck_layersALLOWEDedge — the allowlist is now empty. Every entrypoint routes throughcli.py, which importsprinterat module top, so the factory is always installed before any.printer()call. Guarded with aRuntimeError+ tests.Correctness / resources
_pending_payload/_command_issuedacross the paho network thread; cached-complete-snapshot fast path when a pushall stays silent.socket.setdefaulttimeout()mutation; bound the connect viaclient._connect_timeout(paho 2.x uses it increate_connection); configurable + validated MQTTS port.finally— fixes a socket leak on the connect-exception path.connect/login/prot_pfails.safe_log_error(...)+abort(...)tails afteremit_json_error(now typedNoReturn, already logs+raises). Continues refactor: commands raise; cli.main is the sole error emitter #117/fix: size_verified, leftover contracts, and extras on raised errors #118.Tests / CI / docs
test_docs_consistencygains a stronger drift guard (docs coverage % must be ≥ floor).api.mdgains a support matrix;SECURITY.mdresiduals marked accepted for 1.0; scoreboard refreshed.Follow-up (not in this PR)
AGENTS.mdcoverage line still cites a historical "Linux 3.9" CI run — left for the upcoming full documentation pass.