Skip to content

docs(i18n): polish trad sweep v0.9.0 — scanner #647 = 0 finding (4 datasets, 1813 rows) #667

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
jsboige merged 1 commit into from
Jul 3, 2026
Merged

docs(i18n): polish trad sweep v0.9.0 — scanner #647 = 0 finding (4 datasets, 1813 rows) #667

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
49 changes: 49 additions & 0 deletions docs/investigations/2026-07-03-polish-trad-sweep-v090-confirm.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,49 @@
# Polish trad sweep v0.9.0 — pre-release contamination/drift = 0 (confirmed)

**Date:** 2026-07-03 (po-2024, base master `27442add`)
**Dispatch:** `29u572` (ai-01), secondaire — "sweep de vérification finale pré-release"
**Tool:** scanner #647 (`docs/investigations/scripts/2026-07-02-scan-translations.py`, merged) — 3 anti-FP heuristics
**Result:** **TOTAL: 0 finding** across all 4 datasets. Expected (audits antérieurs 100%).

## TL;DR

Final pre-release translation-quality sweep confirms the 4 datasets are clean at tag time: **0 drift**, **0 MT-contamination residual**, **0 EN-identical-to-FR**. No gpt-5.5 fill warranted. This is the release audit-trail proof that the i18n surface is ship-ready.

## Sweep result (code=truth, master `27442add`)

| Dataset | Rows | Drift | Contamination | EN=FR | Verdict |
|---|---:|---:|---:|---:|---|
| Fallacies (`Argumentum Fallacies - Taxonomy.csv`) | 1408 | 0 | 0 | 0 | ✅ clean |
| Virtues (`Argumentum Virtues - Taxonomy.csv`) | 223 | 0 | 0 | 0 | ✅ clean |
| Scenarii (`Argumentum Scenarii - Cards.csv`) | 167 | 0 | 0 | 0 | ✅ clean |
| Rules (`Argumentum Rules - Cards.csv`) | 15 | 0 | 0 | 0 | ✅ clean |
| **TOTAL** | **1813** | **0** | **0** | **0** | ✅ |

**Scanner #647 TOTAL: 0** (re-run on `27442add`). Unmapped columns reported by the scanner are structural metadata (pk/path/depth/crossLinks/AIF_*/print_and_play), not translatable content — expected, not findings.

## Context — what "clean" rests on

The 0-finding state is the cumulative result of the i18n work landed this cycle:

- **#640** Rules i18n refonte (gpt-5.5) — purged the 23 "English Channel" HIGH occurrences; 0 HIGH résiduel.
- **#653** Johnny 6.1.3 duplication fix — 6 languages; scanner TOTAL: 0 after.
- **#218/#236/#246/#290/#295** Virtues i18n (title/desc/remark × 8 langs) — 100% coverage.
- **Scenarii EN/RU/PT** — 167/167 records 100% covered (verified cell-by-cell `7206f2f9`, 2026-05-24).
- **Fallacies** — 1408 rows; #351/#308 drift (40 PK) resolved in an earlier cycle.

The #642 GDrive Rules sync (2 tabs byte-perfect) and the GDrive migration audit (#656) confirmed Rules parity repo↔GSheet (24→21, 0 char lost in the 24→15 restructure).

## Method / reproducibility

```
python docs/investigations/scripts/2026-07-02-scan-translations.py
# → TOTAL: 0
```

Scanner #647 = 3 anti-false-positive heuristics (accent-normalised EN=FR compare, FR-stopword-density "EN reading as French" probe, contamination dictionary). Read-only on the 4 CSVs. 0 write.

## Conclusion

No translation work warranted pre-tag — the i18n surface is confirmed clean at `27442add`. Per the "no fabrication" discipline: reporting the 0-finding state rather than generating work. **QA verdict on rendered output remains ai-01's lane** (verdict #140 content + #632 CMYK); this doc is the data-side pendant.

Relates #647 (scanner), #640 (Rules refonte), #653 (Johnny fix), #134 (release), #140 (QA), #29u572 (dispatch). Base `27442add`.
Loading