test(sessions): the picker search check no longer depends on the temp directory's name

[Lstarsky0]

What and why

session_control_acceptance::tests::tui_and_api_listings_agree fails now and then on my Mac:

assertion `left == right` failed
  left: ["charlie", "bravo", "alpha"]
 right: ["bravo"]

The test saves its sessions with a workspace under TempDir, searches for brav and expects only bravo. The matcher looks at title id workspace and falls back to a subsequence match, so the random .tmpXXXXXX part of the path can supply the missing letters. When it does, every session in that workspace matches.

How often depends on the machine's temp path. On my Mac the per-user part of the path already contains b…r, so the random part only needs a…v: about 1 run in 70 by simulation. Under /tmp it is about 1 in 57,000, which is why CI doesn't show it. With TMPDIR pointing at a directory named brav it fails every time.

search_sort_and_preview_are_one_projection in the same file already avoids this, and its comment says why: it pins all three haystack fields. This test now records /repo and /other as the workspaces, the same way. The session store stays in the private temp directory and the assertions are unchanged.

#5929 put this test's failures down to the redirected CODEWHALE_HOME and gave it a private store. It has still failed since then with the output above, which a session written into the store by another test would not produce.

Issue

Refs #5929

How I tested it

• With TMPDIR set to a directory named brav, the test fails on main with the output above and passes with this change. With my normal TMPDIR it passes on both.
• The whole session_control_acceptance module: 24 passed.
• The 1-in-70 and 1-in-57,000 figures come from a small script that applies the matcher's rule (substring, else subsequence, over the lowercased title id workspace) to .tmp plus six random alphanumeric characters, 400,000 draws per path.
• cargo fmt --all -- --check, cargo clippy -p codewhale-tui --all-targets --all-features with CI's flags, and cargo check -p codewhale-tui --all-targets: clean. Clippy also on 1.99, plus cargo +1.89.0 check -p codewhale-tui.
• cargo nextest run -p codewhale-tui --lib: 15247 passed, 46 failed. The same 46 fail on main (2e8bc7b) on my machine. Two of them are the command-count checks that 7be1f48 has since updated; most of the rest need an extension-host Node version I don't have.

No new tests; this changes the fixture of an existing one.

Checklist

• One focused change, rebased on current main
• cargo fmt --all passes, and tests cover new or changed behavior
• If this adds or changes a user-facing feature, I updated its docs and its row in docs/features.toml (cargo test -p codewhale-tui --test feature_registry checks it) — n/a

[claude]

Claude Code Review

This pull request is from a fork — automated review is disabled. A repository maintainer can comment @claude review to run a one-time review.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-11T14:16:32.665005Z	82f2d0f	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[Hmbown]

Thank you @Lstarsky0 for tracing the picker’s intermittent search failure to the random temporary-directory name and pinning the complete workspace metadata. Your #6958 merged as its own PR at 73e993a295, with your commit and branch preserved. Exact own-head CI passed: macOS19,035/0failed/33skipped, Ubuntu19,032/0/34, Windows18,301passed(oneleaky)/0/27 and Lint. The listing check and owner recovery passed on all three. This is now on main for the 0.10.2 candidate; the assembled release and final packages still require their gates.

[Hmbown]

What and why

session_control_acceptance::tests::tui_and_api_listings_agree fails now and then on my Mac:

assertion `left == right` failed
  left: ["charlie", "bravo", "alpha"]
 right: ["bravo"]

The test saves its sessions with a workspace under TempDir, searches for brav and expects only bravo. The matcher looks at title id workspace and falls back to a subsequence match, so the random .tmpXXXXXX part of the path can supply the missing letters. When it does, every session in that workspace matches.

How often depends on the machine's temp path. On my Mac the per-user part of the path already contains b…r, so the random part only needs a…v: about 1 run in 70 by simulation. Under /tmp it is about 1 in 57,000, which is why CI doesn't show it. With TMPDIR pointing at a directory named brav it fails every time.

search_sort_and_preview_are_one_projection in the same file already avoids this, and its comment says why: it pins all three haystack fields. This test now records /repo and /other as the workspaces, the same way. The session store stays in the private temp directory and the assertions are unchanged.

#5929 put this test's failures down to the redirected CODEWHALE_HOME and gave it a private store. It has still failed since then with the output above, which a session written into the store by another test would not produce.

Issue

Refs #5929

How I tested it

• With TMPDIR set to a directory named brav, the test fails on main with the output above and passes with this change. With my normal TMPDIR it passes on both.
• The whole session_control_acceptance module: 24 passed.
• The 1-in-70 and 1-in-57,000 figures come from a small script that applies the matcher's rule (substring, else subsequence, over the lowercased title id workspace) to .tmp plus six random alphanumeric characters, 400,000 draws per path.
• cargo fmt --all -- --check, cargo clippy -p codewhale-tui --all-targets --all-features with CI's flags, and cargo check -p codewhale-tui --all-targets: clean. Clippy also on 1.99, plus cargo +1.89.0 check -p codewhale-tui.
• cargo nextest run -p codewhale-tui --lib: 15247 passed, 46 failed. The same 46 fail on main (2e8bc7b) on my machine. Two of them are the command-count checks that 7be1f48 has since updated; most of the rest need an extension-host Node version I don't have.

No new tests; this changes the fixture of an existing one.

Checklist

• One focused change, rebased on current main
• cargo fmt --all passes, and tests cover new or changed behavior
• If this adds or changes a user-facing feature, I updated its docs and its row in docs/features.toml (cargo test -p codewhale-tui --test feature_registry checks it) — n/a

Thank you!!! Wonderful contribution as always