[brettheap]

What was wrong

lanes-edit.sh swapped's first field is what its callers hand to tmux rename-window, to --name, to lane <name> and to a read of the register, so it has to be the register row's spelling. It was the raw name the swap record is read under. swapped_candidates took it from nm[k] = $2, and the comment beside it called that "the spelling the register row carries".

It is not. LOG_AWK keys every line to its log FILE (Amendment 16(c): lane = FL, the file's base name), so field 2 is the base name of lanes/log/<name>.md. A lane whose log was written before Amendment 15 still has a lowercase file beside a mixed-case row (the code's own example is lanes/log/openxfactory-4.md beside openXfactory-4), and swapped printed the lowercase name. window-lane already canonicalised its answer (canon_lane, then ${wl_can:-$wl_l}); swapped did not.

Evidence, run in a temp HOME with a temp bare origin and the fixture below, against origin/main's lanes-edit.sh (f87fd5a):

stdout=reposwlow-1	2026-09-12T09:30:00Z	lowwin:5
FIELD1 repoSwLow-1 (window lowwin:5, file reposwlow-1.md) = [reposwlow-1]

A note on the issue's framing: it names the log LINE's spelling. The parser never lets a line's own spelling reach this read; the raw name is the log FILE's base name. So a row whose log file already carries the row's spelling but whose PAUSED line is lowercase (the existing repoWin-1 fixture) reads correctly today. The bug is the lowercase file beside a mixed-case row.

What changed

• swapped_lanes takes canon_lane of the raw name and prints ${sw_can:-$sw_l} as the first field, the same pattern as window-lane. Where the register cannot answer (the 15(d) pair, an unreadable alias table) it keeps the raw name, as window-lane does.
• The raw name stays what log_path_for and the pickaxe key use, because they read the log. (With a lowercase file, canonicalising first would have pointed the pickaxe at lanes/log/<Mixed>.md, which does not exist on origin, and degraded the landing rank.)
• The comment in swapped_candidates that said the raw name is the row's spelling is corrected, and swapped_lanes gets a header paragraph saying which spelling the first field is and why.
• docs/README-lanes.md, ### swapped: one sentence saying the lane is the row's spelling, with the lowercase-file case.
• Callers checked: lane-start compares with tolower; lane-handoff takes the first field and runs canon_lane on it; nothing else parses the output (git grep -n swapped -- lane lane-start lane-end lane-handoff lanes skills).

Functions and regions edited:

• lanes-edit.sh: swapped_lanes (the loop body and the closing printf, plus its header comment)
• lanes-edit.sh: swapped_candidates (the comment above the final printf in the awk program, no code change)
• docs/README-lanes.md: the ### swapped [<workstation>] section
• tests/test_lane_helpers.sh: the seed block beside repoSwap-1 (new repoSwLow-1 row and reposwlow-1.md log), and the assertions after the window-lane Raven casewin:3 cases

Tests

• New fixture repoSwLow-1 (Raven, PAUSED ... swap; window lowwin:5) whose only log is the lowercase file reposwlow-1.md, and run "$E" swapped Raven asserting its row's first field is repoSwLow-1. Rows are picked by window, because Raven holds other swapped lanes.
• The existing repoWin-1 fixture gets an assertion too (casewin:3 row reads repoWin-1). It passes with or without the fix, and the comment says so.

Fail-before and pass-after, by a scratch script that builds the two Raven fixtures above in a temp HOME and temp bare origin, and runs lanes-edit.sh swapped Raven with LANES_EDIT, LANES_WORKSPACE_ROOT, LANES_REPO, LANES_LANE unset:

BEFORE (origin/main):  FIELD1 repoSwLow-1 = [reposwlow-1]   FIELD1 repoWin-1 = [repoWin-1]
AFTER  (this branch):  FIELD1 repoSwLow-1 = [repoSwLow-1]   FIELD1 repoWin-1 = [repoWin-1]

bash -n passes on lanes-edit.sh and tests/test_lane_helpers.sh. The full suite was not run locally (AGENTS.md ruling of 2026-09-16); CI's tests job is the record.

Closes #55

🤖 Generated with Claude Code

Summary by Sourcery

Report canonical register-row lane names from swapped while preserving log-file names for log lookups and landing-order calculations.

Bug Fixes:

• Ensure swapped reports the lane spelling from the register row instead of the log file name, including lanes with legacy lowercase log files.

Enhancements:

• Document the spelling contract for swapped and clarify the distinction between register names and log file names.

Documentation:

• Update the swapped command documentation with its register-row spelling behavior and legacy lowercase-log example.

Tests:

• Add coverage for swapped lanes whose lowercase log file differs in case from the register row, while retaining coverage for matching log and row spellings.

[sourcery-ai]

Reviewer's Guide

Fix swapped so its first field is the canonical register-row lane name rather than the raw log-file basename, without changing log reads or landing ranking, and document and test both spelling scenarios.

Sequence diagram for canonical swapped lane output

sequenceDiagram
    participant Caller
    participant swapped_lanes
    participant canon_lane
    participant Log
    Caller->>swapped_lanes: swapped_lanes
    swapped_lanes->>canon_lane: canon_lane(sw_l)
    canon_lane-->>swapped_lanes: register row spelling
    swapped_lanes->>Log: log_path_for(sw_l)
    Log-->>swapped_lanes: log data and landing key
    swapped_lanes-->>Caller: first field = sw_can or sw_l

Loading

File-Level Changes

Change	Details	Files
Canonicalize the first swapped output field to the lane spelling stored in the register while preserving the log-file spelling for log lookups and landing-order ranking.	Resolve each raw swapped-lane name through canon_lane, falling back to the raw name when canonicalization is unavailable. Continue using the raw log-derived name for log_path_for and the pickaxe key. Document the distinction between register-row and log-file spellings in implementation comments and the lane documentation.	lanes-edit.sh docs/README-lanes.md
Add regression coverage for mixed-case register rows backed by lowercase log files and retain coverage for logs that already use the row spelling.	Seed a swapped lane with a mixed-case register row and lowercase-only log file. Assert swapped Raven emits the register spelling for that lane. Assert the existing matching-file case remains correctly spelled.	tests/test_lane_helpers.sh

Assessment against linked issues

Issue	Objective	Addressed	Explanation
#55	Make swapped print the register row’s canonical lane spelling as its first field, rather than the raw spelling used to read the log.	✅
#55	Document the first-field spelling contract and add regression coverage for a lane whose log name differs in case from its register row.	✅

Possibly linked issues

• swapped's first field is the LOG's spelling, not the register row's — Amendment 15's canonical-name claim is made where the register is not read #55: The PR fixes the same incorrect swapped first-field spelling by canonicalizing it through the register; it clarifies the raw name comes from the log filename.
• park and status read a root's legs out of .gitmodules with git config -z, so a submodule name with a space no longer loses the leg and a valueless key is never a path #15: The PR implements the issue’s canonical-spelling requirement for swapped, resolving the row spelling when a legacy log filename differs by case.

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[sourcery-ai]

Sorry @brettheap, you've used your own review budget of 250,000 diff characters for the last 7 days.

You can request another review in 2 hours and 49 minutes by commenting @sourcery-ai review. Upgrade to get a review now.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud