Implementare completa U5 din plan-treapta2.md (sectiunea 13):
- _upload.html: drop zone + buton accesibil (a11y: drag nu e la tastatura),
drag-and-drop JS, mesaj 'NU se trimite nimic pana confirmi',
selector foi pt multi-sheet xlsx, stari eroare/mesaj
- _mapcoloane.html: formular mapare coloane cu .maprow/.mapcol.grow,
sugestii fuzzy pre-selectate, etiichete <label> vizibile, sample values,
format data configurabil
- _preview_import.html: tabel 6 stari, pills rezumat, filtre pe stare,
.chk per-rand pe needs_review (D11), banner declarant .banner.warn
direct deasupra input-ului N (D12), bara confirmare sticky,
text 'dubla cu randul N' pe duplicate_in_file (D10 daltonism),
link export CSV randuri esuate
- base.html: .s-needs_review (warn), .s-already_sent/.s-duplicate_in_file
(muted), .drop-zone, .banner.warn, .sticky-bar, .htmx-indicator
- routes.py: rute /_import/upload/mapare-coloane/preview/reset/confirma;
helper _web_compute_preview refoloseste _resolve_row_for_preview,
_already_sent_lookup, _signature din import_router (fara a-l edita);
commit ON CONFLICT DO NOTHING (TOCTOU); log atestare
- tests/test_import_ui.py: 15 teste (dashboard, upload, mapare, preview,
confirmare N corect/gresit, reset, erori, multi-sheet, a11y D10/D11/D12)
279 teste total, 0 esecuri.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Link-ul 'tot' era criptic. 'export CSV: trimise' / 'toate' spune ce descarca
fiecare (don't-make-me-think).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
T5 (tools/import_dbf.py): citire prestatii_rar.DBF / mapare_prestatii.DBF cu
dbfread, raport dry-run (randuri valide/duplicate/goale, mapari orfane = cod
necunoscut in nomenclator), --commit cu upsert idempotent in tranzactie.
Dashboard: browser nomenclator, indicator stare RAR (indisponibil? derivat din
ultimul login < 30h, coada arata ultima stare locala), export audit CSV
(/v1/audit/export?status=sent|all&date_from&date_to, b64Image exclus,
coloana purge_after pentru retentia 90z).
Verify: 11 teste noi (test_import_dbf 6, test_dashboard 5), suita 111 pass,
dry-run real pe DBF-urile din repo + smoke live dashboard/CSV.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
T5 reinterpretat: nu import DBF, ci editor web al maparii operatie ROAAUTO ->
cod RAR, cu fuzzy lookup si validare de catre utilizator.
- Contract hibrid: item prestatie accepta cod_prestatie (RAR direct, back-compat)
SAU cod_op_service+denumire (mapat de gateway prin operations_mapping).
- Ingestie: op intern necunoscut -> submission needs_mapping (nu pleaca la RAR);
codul rezolvat se scrie inapoi in payload_json -> payload builder + worker neatinse.
- Editor HTMX (_mapari.html + GET /_fragments/mapari, POST /mapari): listeaza
op-urile nemapate, fuzzy preselecteaza codul RAR, save -> re-rezolvare automata
(queued / needs_data).
- Fuzzy: rapidfuzz.token_sort_ratio pe denumire normalizata (fara diacritice).
- Nomenclator: seed fallback 18 coduri la boot (offline) + refresh live din worker.
- Cont default id=1 cat timp auth API-key (CORE) nu exista (account_id NULL).
- Endpointuri API: GET /v1/mapari/pending, POST /v1/mapari (respinge cod inexistent).
- 15 teste noi (tests/test_mapping.py); 69 pass total.
- Contract actualizat (docs/api-rar-contract.md), rapidfuzz==3.14.5 in requirements.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
