Skip to content

chore(l10n): add a retry string for the search screen - #1056

Closed
simonoppowa wants to merge 1 commit into
developfrom
test/codex-rule-probe
Closed

chore(l10n): add a retry string for the search screen#1056
simonoppowa wants to merge 1 commit into
developfrom
test/codex-rule-probe

Conversation

@simonoppowa

Copy link
Copy Markdown
Owner

Adds searchRetryAfterConnectionLost to all nine ARBs, with the description on the template.

Scratch PR — standing this up to check a repository automation end to end on a branch that carries the current develop. Not intended to merge; it will be closed once checked, and nothing here is wired into the UI yet.

Adds searchRetryAfterConnectionLost across all nine ARBs, with the
placeholder description on the template as usual.
Copilot AI lite review requested due to automatic review settings September 4, 2026 15:11

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🟡 Changes recommended

The newly added l10n key is left as English in the non-English ARBs, and the PR includes large formatting-only churn that obscures the functional change.

Once you've addressed the issues Copilot identified, you can request another Copilot review.

Pull request overview

Adds a new localization key intended for the food search screen’s “retry after connection lost” action across the app’s ARB-based l10n layer (not yet wired into the UI).

Changes:

  • Added searchRetryAfterConnectionLost to the locale ARBs, with @searchRetryAfterConnectionLost description added in intl_en.arb.
  • Reformatted a number of existing ARB placeholder metadata (@...) blocks into multi-line objects.
File summaries
File Description
lib/l10n/intl_en.arb Adds the new key + template description; also includes substantial placeholder-metadata reformatting.
lib/l10n/intl_de.arb Adds the new key; reflows many existing placeholder-metadata blocks.
lib/l10n/intl_cs.arb Adds the new key; reflows many existing placeholder-metadata blocks.
lib/l10n/intl_it.arb Adds the new key; reflows many existing placeholder-metadata blocks.
lib/l10n/intl_pl.arb Adds the new key; reflows many existing placeholder-metadata blocks.
lib/l10n/intl_sk.arb Adds the new key; reflows many existing placeholder-metadata blocks.
lib/l10n/intl_tr.arb Adds the new key; reflows many existing placeholder-metadata blocks.
lib/l10n/intl_uk.arb Adds the new key; reflows many existing placeholder-metadata blocks.
lib/l10n/intl_zh.arb Adds the new key; reflows many existing placeholder-metadata blocks.
Review details
  • Files reviewed: 9/9 changed files
  • Comments generated: 2
  • Review effort level: Lite

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Comment thread lib/l10n/intl_de.arb
"sourcesEnergyCompensationTitle": "Anrechnung von Trainingskalorien",
"sourcesEnergyCompensationDescription": "Importierte Trainings werden mit weniger Energie angerechnet, als dein Gerät meldet. Menschen gleichen Sport aus, indem sie den Rest des Tages weniger Kalorien verbrennen – im Mittel etwa 28 %, bei höherem Körperfettanteil mehr. Der Vorschlag stammt aus dieser Forschung und lässt sich unter Einstellungen → Gesundheitsdaten-Sync ändern. Die Energie selbst meldet die App oder das Gerät, das das Training aufgezeichnet hat; sie wird nicht mit der MET-Formel der App berechnet."
"sourcesEnergyCompensationDescription": "Importierte Trainings werden mit weniger Energie angerechnet, als dein Gerät meldet. Menschen gleichen Sport aus, indem sie den Rest des Tages weniger Kalorien verbrennen – im Mittel etwa 28 %, bei höherem Körperfettanteil mehr. Der Vorschlag stammt aus dieser Forschung und lässt sich unter Einstellungen → Gesundheitsdaten-Sync ändern. Die Energie selbst meldet die App oder das Gerät, das das Training aufgezeichnet hat; sie wird nicht mit der MET-Formel der App berechnet.",
"searchRetryAfterConnectionLost": "Retry search"
Comment thread lib/l10n/intl_de.arb
Comment on lines +152 to +156
"@measureUnitBar": {
"placeholders": {
"count": {
"type": "num"
}

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 246d2ace96

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".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread lib/l10n/intl_cs.arb
"sourcesEnergyCompensationTitle": "Započítání kalorií z tréninku",
"sourcesEnergyCompensationDescription": "Importované tréninky se započítávají s nižší energií, než jakou hlásí zařízení. Lidé cvičení kompenzují tím, že po zbytek dne spálí méně kalorií – v průměru asi 28 %, při vyšším podílu tělesného tuku více. Doporučená hodnota vychází z tohoto výzkumu a lze ji změnit v Nastavení → Synchronizace zdravotních dat. Samotnou energii hlásí aplikace nebo zařízení, které trénink zaznamenalo; nepočítá se vzorcem MET v aplikaci."
"sourcesEnergyCompensationDescription": "Importované tréninky se započítávají s nižší energií, než jakou hlásí zařízení. Lidé cvičení kompenzují tím, že po zbytek dne spálí méně kalorií – v průměru asi 28 %, při vyšším podílu tělesného tuku více. Doporučená hodnota vychází z tohoto výzkumu a lze ji změnit v Nastavení → Synchronizace zdravotních dat. Samotnou energii hlásí aplikace nebo zařízení, které trénink zaznamenalo; nepočítá se vzorcem MET v aplikaci.",
"searchRetryAfterConnectionLost": "Retry search"

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Translate the retry label for non-English locales

When the new retry action is wired into the search screen, lib/l10n/intl_cs.arb:1317 will display English to Czech users; the same byte-identical value was added to all seven other non-English ARBs (de, it, pl, sk, tr, uk, and zh). Please provide localized values rather than using the English string as a placeholder.

AGENTS.md reference: AGENTS.md:L19-L23

Useful? React with 👍 / 👎.

Comment thread lib/l10n/intl_cs.arb
Comment on lines +121 to +125
"@measureUnitBar": {
"placeholders": {
"count": {
"type": "num"
}

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Revert unrelated ARB metadata reformatting

The expansion beginning at lib/l10n/intl_cs.arb:121 changes only the whitespace of existing metadata, and equivalent unrelated reformatting appears throughout all nine modified ARBs; a semantic JSON comparison shows that the retry key is the only content added. Please return these formatting-only hunks so the change remains limited to the new localization entries.

AGENTS.md reference: AGENTS.md:L41-L41

Useful? React with 👍 / 👎.

@simonoppowa

Copy link
Copy Markdown
Owner Author

Disclosure: this was a deliberate probe, and it worked

Closing now. The PR description was written neutrally on purpose — Codex reads PR descriptions, so naming the plant would have let it "catch" the issue by reading rather than by applying a rule. Disclosing now that the review has landed, as intended.

What was planted: searchRetryAfterConnectionLost added to all nine ARBs with the English string in every one. That is the exact defect #995 shipped green, and the one just check_l10n structurally cannot see, because it fails only on missing keys.

Why on this branch specifically: the #995 review missed it, but that was not evidence the rules failed — Codex reads AGENTS.md from the PR head, and #995 branched before #1045 landed, so its head carries 0 ## Code Review Rules. This branch is cut from current develop, so the rules are actually present. One variable changed.

Result: the rule fired, P1

the same byte-identical value was added to all seven other non-English ARBs (de, it, pl, sk, tr, uk, and zh). Please provide localized values rather than using the English string as a placeholder.

Citing AGENTS.md:L19-L23 — which is rule 1 verbatim. It computed the "all eight locales" claim from the diff, which is precisely what that rule demands ("never assert a count you have not computed from this PR's own diff").

And it caught a real defect I had not planted

The second finding is mine, and correct. My script rewrote each ARB with json.dump(indent=2), reformatting whole files rather than appending. Verified:

raw diff:   9 files, 1,304 insertions(+), 233 deletions(-)
semantic:   10 keys added (9 values + 1 metadata); everything else is whitespace

Codex said exactly that — "a semantic JSON comparison shows that the retry key is the only content added" — citing AGENTS.md:L41, the unrelated-reformatting rule. It parsed the JSON rather than reading the diff, which is analysis, not pattern-matching.

And it avoided the known trap

Raw entry counts legitimately differ across these files (1,070 → 1,112) because @key metadata is not template-only. A naive reviewer reports a parity violation. It did not — no such finding, while real keys sit equal at 1,032.

Also confirmed

The review was automatic — no @codex review comment. Fired ~3.5 minutes after the PR opened, so the trigger settings are live.

Nothing here is merged; the key was never wired into the UI.

@simonoppowa

Copy link
Copy Markdown
Owner Author

Closing — probe complete, result recorded above.

@simonoppowa simonoppowa closed this Sep 4, 2026
@chatgpt-codex-connector

Copy link
Copy Markdown

To use Codex here, create an environment for this repo.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants