redact: UX overhaul — engine simplification, credits, spinner, save-to-docs, badges

- Remove GPU/regex engine options; keep only azure_mini (1 credit) and azure_full (2 credits)
- Variable credit cost: engine-aware pre-check and charge in api/redact.php; PricingCatalog base = 1
- Fix ATTORNEY not preserved when keepOfficials=true: add to LLM prompt, generic-tag, pseudonym regexes
- Replace Azure credits hint with per-engine credit cost text (all 4 languages)
- Single-file upload only (was: up to 5); simplify status messages
- Clear previous redaction output and show pulsing spinner when a new run starts
- Add "Save to My Docs" button in redact output panel (corpus-save.js path)
- corpus-save.js: capture source_doc_ids from button dataset, pass in POST payload
- api/save-to-corpus.php: accept source_doc_ids, store first as source_url=corpus-doc:{id}
- doc-picker.js: show "✂ Redacted" badge for documents saved from the redact tool
- CSS: .redact-working spinner, doc-item__badge--redact pill styles

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
2026-05-25 08:18:51 +02:00
parent a821d39dcd
commit 56cd87dd7b
10 changed files with 117 additions and 50 deletions
+10 -3
View File
@@ -5,8 +5,15 @@ require_once __DIR__ . '/../includes/LegalTools.php';
dbnToolsRequireMethod('POST'); dbnToolsRequireMethod('POST');
dbnToolsRequireAuth(); dbnToolsRequireAuth();
$ftUid = dbnToolsFreeTierCheck('redact');
// Determine engine and its credit cost before the pre-flight credit check
$input = dbnToolsJsonInput(400000); $input = dbnToolsJsonInput(400000);
$_validEngines = ['azure_mini', 'azure_full'];
$_engine = in_array((string)($input['engine'] ?? ''), $_validEngines, true)
? (string)$input['engine'] : 'azure_mini';
$_engineCredits = $_engine === 'azure_full' ? 2 : 1;
$ftUid = dbnToolsFreeTierCheckAmount('redact', $_engineCredits);
dbnToolsWithChargedTelemetry('redact', '', $ftUid, function () use ($input): array { dbnToolsWithChargedTelemetry('redact', '', $ftUid, function () use ($input): array {
$text = dbnToolsInjectDocContent($input, dbnToolsString($input, 'text', 128000, false)); $text = dbnToolsInjectDocContent($input, dbnToolsString($input, 'text', 128000, false));
@@ -17,7 +24,7 @@ dbnToolsWithChargedTelemetry('redact', '', $ftUid, function () use ($input): arr
$region = dbnToolsNormalizeRegion($input['region'] ?? 'nordic'); $region = dbnToolsNormalizeRegion($input['region'] ?? 'nordic');
$language = dbnToolsNormalizeLanguage($input['language'] ?? 'en'); $language = dbnToolsNormalizeLanguage($input['language'] ?? 'en');
$validEngines = ['azure_mini', 'azure_full', 'gpu', 'regex']; $validEngines = ['azure_mini', 'azure_full'];
$engine = in_array((string)($input['engine'] ?? ''), $validEngines, true) $engine = in_array((string)($input['engine'] ?? ''), $validEngines, true)
? (string)$input['engine'] ? (string)$input['engine']
: 'azure_mini'; : 'azure_mini';
@@ -67,4 +74,4 @@ dbnToolsWithChargedTelemetry('redact', '', $ftUid, function () use ($input): arr
$text, $mode, $region, $language, $aliases, $text, $mode, $region, $language, $aliases,
$engine, $outputFormat, $keepOfficials, $exemptNames, $redactTypes $engine, $outputFormat, $keepOfficials, $exemptNames, $redactTypes
); );
}); }, $_engineCredits);
+10 -3
View File
@@ -35,6 +35,13 @@ $tags = json_encode(
JSON_UNESCAPED_UNICODE JSON_UNESCAPED_UNICODE
); );
$rawSourceDocIds = $input['source_doc_ids'] ?? null;
$sourceDocIdArr = is_array($rawSourceDocIds)
? $rawSourceDocIds
: (is_string($rawSourceDocIds) ? array_filter(array_map('trim', explode(',', $rawSourceDocIds))) : []);
$firstSourceDocId = (int)(reset($sourceDocIdArr) ?: 0);
$sourceUrl = $firstSourceDocId > 0 ? "corpus-doc:{$firstSourceDocId}" : null;
if ($title === '') { if ($title === '') {
dbnToolsError('title is required.', 400, 'bad_request'); dbnToolsError('title is required.', 400, 'bad_request');
} }
@@ -71,10 +78,10 @@ $wordCount = str_word_count($content);
$ins = $db->prepare(" $ins = $db->prepare("
INSERT INTO client_documents INSERT INTO client_documents
(client_id, corpus_id, title, source_type, content, category, (client_id, corpus_id, title, source_type, content, category,
tags, import_method, source_tool, word_count, status) tags, import_method, source_tool, source_url, word_count, status)
VALUES (?, ?, ?, 'text', ?, 'tool-output', ?, 'tool_output', ?, ?, 'pending') VALUES (?, ?, ?, 'text', ?, 'tool-output', ?, 'tool_output', ?, ?, ?, 'pending')
"); ");
$ins->execute([$clientId, $corpusId, $title, $content, $tags, $sourceTool, $wordCount]); $ins->execute([$clientId, $corpusId, $title, $content, $tags, $sourceTool, $sourceUrl, $wordCount]);
$docId = (int)$db->lastInsertId(); $docId = (int)$db->lastInsertId();
try { try {
+20
View File
@@ -270,3 +270,23 @@
transition: border-color 0.15s, color 0.15s; transition: border-color 0.15s, color 0.15s;
} }
.audio-corpus-upload:hover { border-color: var(--dbn-accent, #00205B); color: var(--dbn-accent, #00205B); } .audio-corpus-upload:hover { border-color: var(--dbn-accent, #00205B); color: var(--dbn-accent, #00205B); }
/* ── Redacted document badge ─────────────────────────────────────────────── */
.doc-item__badge {
display: inline-block;
margin-left: 0.45em;
padding: 0.1em 0.45em;
border-radius: 4px;
font-size: 0.72rem;
font-weight: 600;
vertical-align: middle;
line-height: 1.4;
}
.doc-item__badge--redact {
background: #ede8f7;
color: #5b35a8;
border: 1px solid rgba(91, 53, 168, 0.2);
}
+32
View File
@@ -2266,6 +2266,38 @@ p {
cursor: progress; cursor: progress;
} }
@keyframes redact-spin {
to { transform: rotate(360deg); }
}
.redact-working {
display: flex;
flex-direction: column;
align-items: center;
justify-content: center;
gap: 12px;
padding: 32px 16px;
border: 1px solid var(--line);
border-radius: 8px;
background: #fbfcfe;
}
.redact-working p {
margin: 0;
color: var(--muted);
font-size: 0.92rem;
}
.redact-working__spinner {
display: block;
width: 22px;
height: 22px;
border: 3px solid var(--line);
border-top-color: var(--teal);
border-radius: 50%;
animation: redact-spin 0.9s linear infinite;
}
/* ── Feedback widget ──────────────────────────────────────────── */ /* ── Feedback widget ──────────────────────────────────────────── */
.feedback-widget { .feedback-widget {
margin-top: 1.5rem; margin-top: 1.5rem;
+16 -10
View File
@@ -41,6 +41,7 @@
} }
function bodyFor(kind, payload) { function bodyFor(kind, payload) {
const sourceDocIds = (payload.sourceDocIds || '').split(',').map(s => s.trim()).filter(Boolean);
if (window.DBN_DASHBOARD) { if (window.DBN_DASHBOARD) {
return JSON.stringify({ return JSON.stringify({
title: payload.title, title: payload.title,
@@ -48,6 +49,7 @@
source_tool: payload.tool || 'dashboard-save', source_tool: payload.tool || 'dashboard-save',
tags: payload.tags, tags: payload.tags,
kind, kind,
...(sourceDocIds.length ? { source_doc_ids: sourceDocIds } : {}),
}); });
} }
return JSON.stringify({ return JSON.stringify({
@@ -55,6 +57,7 @@
content: payload.content, content: payload.content,
source_tool: payload.tool || '', source_tool: payload.tool || '',
tags: payload.tags, tags: payload.tags,
...(sourceDocIds.length ? { source_doc_ids: sourceDocIds } : {}),
}); });
} }
@@ -74,9 +77,10 @@
} }
_pendingBtn = btn; _pendingBtn = btn;
dlg.dataset.pendingContent = content; dlg.dataset.pendingContent = content;
dlg.dataset.pendingTool = btn.dataset.tool || ''; dlg.dataset.pendingTool = btn.dataset.tool || '';
dlg.dataset.pendingKind = 'tool_output'; dlg.dataset.pendingKind = 'tool_output';
dlg.dataset.pendingSourceDocIds = btn.dataset.sourceDocIds || '';
titleIn.value = btn.dataset.suggestedTitle || ''; titleIn.value = btn.dataset.suggestedTitle || '';
tagsIn.value = ''; tagsIn.value = '';
@@ -90,12 +94,13 @@
e.preventDefault(); e.preventDefault();
dlg.close(); dlg.close();
const btn = _pendingBtn; const btn = _pendingBtn;
const content = dlg.dataset.pendingContent || ''; const content = dlg.dataset.pendingContent || '';
const tool = dlg.dataset.pendingTool || ''; const tool = dlg.dataset.pendingTool || '';
const kind = dlg.dataset.pendingKind || 'tool_output'; const kind = dlg.dataset.pendingKind || 'tool_output';
const title = titleIn.value.trim(); const sourceDocIds = dlg.dataset.pendingSourceDocIds || '';
const tags = tagsIn.value.trim(); const title = titleIn.value.trim();
const tags = tagsIn.value.trim();
if (!title || !content) return; if (!title || !content) return;
@@ -109,7 +114,7 @@
method: 'POST', method: 'POST',
credentials: 'same-origin', credentials: 'same-origin',
headers: { 'Content-Type': 'application/json' }, headers: { 'Content-Type': 'application/json' },
body: bodyFor(kind, { title, content, tool, tags }), body: bodyFor(kind, { title, content, tool, tags, sourceDocIds }),
}); });
const data = await resp.json().catch(() => ({})); const data = await resp.json().catch(() => ({}));
@@ -146,6 +151,7 @@
delete dlg.dataset.pendingContent; delete dlg.dataset.pendingContent;
delete dlg.dataset.pendingTool; delete dlg.dataset.pendingTool;
delete dlg.dataset.pendingKind; delete dlg.dataset.pendingKind;
delete dlg.dataset.pendingSourceDocIds;
}); });
function showToast(msg, isError) { function showToast(msg, isError) {
+4 -1
View File
@@ -97,10 +97,13 @@
.toLocaleDateString(undefined, { dateStyle: 'medium' })); .toLocaleDateString(undefined, { dateStyle: 'medium' }));
} catch (_) {} } catch (_) {}
} }
var redactBadge = doc.source_tool === 'redact'
? '<span class="doc-item__badge doc-item__badge--redact">✂ Redacted</span>'
: '';
return '<div class="doc-item' + (sel ? ' is-selected' : '') + '" data-id="' + id + '" role="option" aria-selected="' + sel + '">' return '<div class="doc-item' + (sel ? ' is-selected' : '') + '" data-id="' + id + '" role="option" aria-selected="' + sel + '">'
+ '<input type="checkbox" ' + (sel ? 'checked' : '') + ' tabindex="-1" aria-hidden="true">' + '<input type="checkbox" ' + (sel ? 'checked' : '') + ' tabindex="-1" aria-hidden="true">'
+ '<div>' + '<div>'
+ '<div class="doc-item__title">' + esc(doc.title || 'Untitled') + '</div>' + '<div class="doc-item__title">' + esc(doc.title || 'Untitled') + redactBadge + '</div>'
+ (meta.length ? '<div class="doc-item__meta">' + esc(meta.join(' · ')) + '</div>' : '') + (meta.length ? '<div class="doc-item__meta">' + esc(meta.join(' · ')) + '</div>' : '')
+ '</div>' + '</div>'
+ '</div>'; + '</div>';
+18 -24
View File
@@ -8,9 +8,7 @@ const REDACT_I18N = {
redactEngine: 'Engine', redactEngine: 'Engine',
redactEngineAzureMini: 'Azure gpt-4o-mini', redactEngineAzureMini: 'Azure gpt-4o-mini',
redactEngineAzureFull: 'Azure gpt-4o', redactEngineAzureFull: 'Azure gpt-4o',
redactEngineGpu: 'GPU (cuttlefish)', redactEngineHint: 'gpt-4o-mini: 1 credit — fast, handles most documents well. gpt-4o: 2 credits — higher accuracy for complex or multi-person cases.',
redactEngineRegex: 'Regex only',
redactEngineHint: 'Azure engines use your BNL Azure credits. GPU runs the local LiteLLM proxy. Regex-only is instant and free but finds no names or organisations.',
redactMode: 'Mode', redactMode: 'Mode',
redactModeStandard: 'Standard', redactModeStandard: 'Standard',
redactModeStrict: 'Strict', redactModeStrict: 'Strict',
@@ -42,7 +40,7 @@ const REDACT_I18N = {
redactAliasAdd: 'Add', redactAliasAdd: 'Add',
redactAliasHint: 'Replace a specific name with a custom bracketed label, e.g. "David Jr" → [Junior].', redactAliasHint: 'Replace a specific name with a custom bracketed label, e.g. "David Jr" → [Junior].',
redactUploadAria: 'File upload', redactUploadAria: 'File upload',
redactUploadDrop: 'Drop up to 5 files here, or', redactUploadDrop: 'Drop one file here, or',
redactUploadBrowse: 'browse', redactUploadBrowse: 'browse',
redactUploadHint: 'text extracted in memory, never stored', redactUploadHint: 'text extracted in memory, never stored',
redactUploadClear: '× Clear', redactUploadClear: '× Clear',
@@ -63,9 +61,7 @@ const REDACT_I18N = {
redactEngine: 'Motor', redactEngine: 'Motor',
redactEngineAzureMini: 'Azure gpt-4o-mini', redactEngineAzureMini: 'Azure gpt-4o-mini',
redactEngineAzureFull: 'Azure gpt-4o', redactEngineAzureFull: 'Azure gpt-4o',
redactEngineGpu: 'GPU (cuttlefish)', redactEngineHint: 'gpt-4o-mini: 1 kreditt — rask, håndterer de fleste dokumenter godt. gpt-4o: 2 kreditter — høyere nøyaktighet for komplekse eller flerpersonssaker.',
redactEngineRegex: 'Kun regex',
redactEngineHint: 'Azure-motorer bruker BNL Azure-kreditter. GPU kjører lokal LiteLLM-proxy. Kun regex er øyeblikkelig og gratis, men finner ingen navn eller organisasjoner.',
redactMode: 'Modus', redactMode: 'Modus',
redactModeStandard: 'Standard', redactModeStandard: 'Standard',
redactModeStrict: 'Strikt', redactModeStrict: 'Strikt',
@@ -97,7 +93,7 @@ const REDACT_I18N = {
redactAliasAdd: 'Legg til', redactAliasAdd: 'Legg til',
redactAliasHint: 'Erstatt et spesifikt navn med en egendefinert merkelapp, f.eks. «David Jr» → [Junior].', redactAliasHint: 'Erstatt et spesifikt navn med en egendefinert merkelapp, f.eks. «David Jr» → [Junior].',
redactUploadAria: 'Filopplasting', redactUploadAria: 'Filopplasting',
redactUploadDrop: 'Slipp opptil 5 filer her, eller', redactUploadDrop: 'Slipp én fil her, eller',
redactUploadBrowse: 'bla', redactUploadBrowse: 'bla',
redactUploadHint: 'tekst hentes i minnet, lagres aldri', redactUploadHint: 'tekst hentes i minnet, lagres aldri',
redactUploadClear: '× Tøm', redactUploadClear: '× Tøm',
@@ -118,9 +114,7 @@ const REDACT_I18N = {
redactEngine: 'Рушій', redactEngine: 'Рушій',
redactEngineAzureMini: 'Azure gpt-4o-mini', redactEngineAzureMini: 'Azure gpt-4o-mini',
redactEngineAzureFull: 'Azure gpt-4o', redactEngineAzureFull: 'Azure gpt-4o',
redactEngineGpu: 'GPU (cuttlefish)', redactEngineHint: 'gpt-4o-mini: 1 кредит — швидко, добре обробляє більшість документів. gpt-4o: 2 кредити — вища точність для складних або багатоособових справ.',
redactEngineRegex: 'Лише регулярні вирази',
redactEngineHint: 'Рушії Azure використовують кредити BNL Azure. GPU запускає локальний проксі LiteLLM. Лише regex — миттєво і безкоштовно, але не знаходить імен або організацій.',
redactMode: 'Режим', redactMode: 'Режим',
redactModeStandard: 'Стандартний', redactModeStandard: 'Стандартний',
redactModeStrict: 'Суворий', redactModeStrict: 'Суворий',
@@ -152,7 +146,7 @@ const REDACT_I18N = {
redactAliasAdd: 'Додати', redactAliasAdd: 'Додати',
redactAliasHint: 'Замініть конкретне ім\'я на власну мітку, напр. «David Jr» → [Junior].', redactAliasHint: 'Замініть конкретне ім\'я на власну мітку, напр. «David Jr» → [Junior].',
redactUploadAria: 'Завантаження файлів', redactUploadAria: 'Завантаження файлів',
redactUploadDrop: 'Перетягніть до 5 файлів сюди, або', redactUploadDrop: 'Перетягніть один файл сюди, або',
redactUploadBrowse: 'огляд', redactUploadBrowse: 'огляд',
redactUploadHint: 'текст обробляється в пам\'яті, ніколи не зберігається', redactUploadHint: 'текст обробляється в пам\'яті, ніколи не зберігається',
redactUploadClear: '× Очистити', redactUploadClear: '× Очистити',
@@ -173,9 +167,7 @@ const REDACT_I18N = {
redactEngine: 'Silnik', redactEngine: 'Silnik',
redactEngineAzureMini: 'Azure gpt-4o-mini', redactEngineAzureMini: 'Azure gpt-4o-mini',
redactEngineAzureFull: 'Azure gpt-4o', redactEngineAzureFull: 'Azure gpt-4o',
redactEngineGpu: 'GPU (cuttlefish)', redactEngineHint: 'gpt-4o-mini: 1 kredyt — szybko, dobrze radzi sobie z większością dokumentów. gpt-4o: 2 kredyty — wyższa dokładność dla złożonych lub wieloosobowych spraw.',
redactEngineRegex: 'Tylko regex',
redactEngineHint: 'Silniki Azure używają kredytów Azure BNL. GPU korzysta z lokalnego proxy LiteLLM. Tylko regex jest natychmiastowy i bezpłatny, ale nie znajdzie imion ani organizacji.',
redactMode: 'Tryb', redactMode: 'Tryb',
redactModeStandard: 'Standardowy', redactModeStandard: 'Standardowy',
redactModeStrict: 'Ścisły', redactModeStrict: 'Ścisły',
@@ -207,7 +199,7 @@ const REDACT_I18N = {
redactAliasAdd: 'Dodaj', redactAliasAdd: 'Dodaj',
redactAliasHint: 'Zastąp konkretną nazwę własną etykietą, np. «David Jr» → [Junior].', redactAliasHint: 'Zastąp konkretną nazwę własną etykietą, np. «David Jr» → [Junior].',
redactUploadAria: 'Przesyłanie pliku', redactUploadAria: 'Przesyłanie pliku',
redactUploadDrop: 'Upuść do 5 plików tutaj lub', redactUploadDrop: 'Upuść jeden plik tutaj lub',
redactUploadBrowse: 'przeglądaj', redactUploadBrowse: 'przeglądaj',
redactUploadHint: 'tekst wyodrębniany w pamięci, nigdy nie przechowywany', redactUploadHint: 'tekst wyodrębniany w pamięci, nigdy nie przechowywany',
redactUploadClear: '× Wyczyść', redactUploadClear: '× Wyczyść',
@@ -1126,6 +1118,9 @@ async function runTool(event) {
lastToolPayload = { ...payload }; lastToolPayload = { ...payload };
setBusy(true); setBusy(true);
if (state.activeTool === 'redact') {
els.results.innerHTML = '<div class="redact-working" role="status" aria-live="polite"><span class="redact-working__spinner" aria-hidden="true"></span><p>Redacting document…</p></div>';
}
renderTrace([ renderTrace([
{ label: 'Query interpretation', detail: 'Preparing request.', status: 'running' }, { label: 'Query interpretation', detail: 'Preparing request.', status: 'running' },
]); ]);
@@ -1218,7 +1213,7 @@ function setupUpload() {
async function handleFiles(fileList) { async function handleFiles(fileList) {
const allowed = ['pdf', 'docx', 'txt']; const allowed = ['pdf', 'docx', 'txt'];
const files = Array.from(fileList).slice(0, 5); const files = Array.from(fileList).slice(0, 1);
for (const file of files) { for (const file of files) {
const ext = file.name.split('.').pop().toLowerCase(); const ext = file.name.split('.').pop().toLowerCase();
@@ -1228,7 +1223,7 @@ async function handleFiles(fileList) {
} }
} }
els.status.textContent = files.length === 1 ? `Extracting ${files[0].name}` : `Extracting ${files.length} files…`; els.status.textContent = `Extracting ${files[0].name}`;
setBusy(true); setBusy(true);
const parts = []; const parts = [];
@@ -1256,9 +1251,7 @@ async function handleFiles(fileList) {
if (data.truncated) anyTruncated = true; if (data.truncated) anyTruncated = true;
} }
const combined = parts.length === 1 const combined = parts[0].text;
? parts[0].text
: parts.map((p) => `--- Document: ${p.filename} ---\n\n${p.text}`).join('\n\n');
const MAX_COMBINED = 128000; const MAX_COMBINED = 128000;
const combinedTruncated = combined.length > MAX_COMBINED; const combinedTruncated = combined.length > MAX_COMBINED;
@@ -1271,9 +1264,7 @@ async function handleFiles(fileList) {
els.uploadFileInfo.classList.remove('is-hidden'); els.uploadFileInfo.classList.remove('is-hidden');
const truncNote = (anyTruncated || combinedTruncated) ? ' — truncated to 128000 char limit' : ''; const truncNote = (anyTruncated || combinedTruncated) ? ' — truncated to 128000 char limit' : '';
els.status.textContent = parts.length === 1 els.status.textContent = `Extracted ${totalChars.toLocaleString()} chars from ${parts[0].filename}${truncNote}.`;
? `Extracted ${totalChars.toLocaleString()} chars from ${parts[0].filename}${truncNote}.`
: `Extracted ${totalChars.toLocaleString()} chars total from ${parts.length} files${truncNote}.`;
} catch (err) { } catch (err) {
els.status.textContent = err.message; els.status.textContent = err.message;
resetUpload(); resetUpload();
@@ -1560,10 +1551,13 @@ function renderMainFinding(data) {
? `<button type="button" class="upgrade-engine-btn" id="rerunBetterBtn">Re-run with gpt-4o for higher accuracy →</button>` ? `<button type="button" class="upgrade-engine-btn" id="rerunBetterBtn">Re-run with gpt-4o for higher accuracy →</button>`
: ''; : '';
const sourceDocIds = lastToolPayload?.doc_ids?.join(',') || '';
const suggestedTitle = `Redacted document — ${new Date().toLocaleDateString()}`;
const dlRow = `<div class="redact-downloads"> const dlRow = `<div class="redact-downloads">
<button type="button" class="redact-dl-btn" id="rdlCopy">${t('redactCopy')}</button> <button type="button" class="redact-dl-btn" id="rdlCopy">${t('redactCopy')}</button>
<button type="button" class="redact-dl-btn" id="rdlTxt">${t('redactDownloadTxt')}</button> <button type="button" class="redact-dl-btn" id="rdlTxt">${t('redactDownloadTxt')}</button>
<button type="button" class="redact-dl-btn" id="rdlDocx">${t('redactDownloadDocx')}</button> <button type="button" class="redact-dl-btn" id="rdlDocx">${t('redactDownloadDocx')}</button>
<button type="button" class="redact-dl-btn js-save-corpus" data-content-id="redactOutputPre" data-tool="redact" data-suggested-title="${escapeHtml(suggestedTitle)}" data-source-doc-ids="${escapeHtml(sourceDocIds)}">Save to My Docs</button>
</div>`; </div>`;
return `${viewToggle}<pre class="redacted-output" id="redactOutputPre">${highlightRedactedText(lastRedactedText)}</pre>${inventoryHtml}${upgradeBtn}${dlRow}`; return `${viewToggle}<pre class="redacted-output" id="redactOutputPre">${highlightRedactedText(lastRedactedText)}</pre>${inventoryHtml}${upgradeBtn}${dlRow}`;
+3 -3
View File
@@ -1120,7 +1120,7 @@ PROMPT;
// Build officials note // Build officials note
$officialsNote = ''; $officialsNote = '';
if ($keepOfficials) { if ($keepOfficials) {
$officialsNote = "\n\nOFFICIALS — for persons identified as JUDGE, EXPERT_WITNESS, or CASEWORKER in an official capacity: do NOT replace their name with a plain bracket tag. Instead use the format [ROLE: Name], e.g. [JUDGE: Andersen] or [EXPERT_WITNESS: Dr. Larsen]. Their name must remain visible inside the tag."; $officialsNote = "\n\nOFFICIALS — for persons identified as JUDGE, ATTORNEY, EXPERT_WITNESS, or CASEWORKER in an official capacity: do NOT replace their name with a plain bracket tag. Instead use the format [ROLE: Name], e.g. [JUDGE: Andersen], [ATTORNEY: Skretting] or [EXPERT_WITNESS: Dr. Larsen]. Their name must remain visible inside the tag.";
} }
$allowedTypesNote = ''; $allowedTypesNote = '';
@@ -1363,7 +1363,7 @@ PROMPT;
private function applyGenericTags(string $text): string private function applyGenericTags(string $text): string
{ {
// Collapse contextual role tags (e.g. [FATHER], [JUDGE: Andersen], [CHILD_1]) → [PERSON] // Collapse contextual role tags (e.g. [FATHER], [JUDGE: Andersen], [CHILD_1]) → [PERSON]
$text = preg_replace('/\[(?:FATHER|MOTHER|CHILD(?:_\d+)?|GRANDPARENT|SIBLING|ATTORNEY|JUDGE(?::\s*[^\]]+)?|CASEWORKER(?::\s*[^\]]+)?|EXPERT_WITNESS(?::\s*[^\]]+)?|PERSON(?:_\d+)?)\]/u', '[PERSON]', $text) ?? $text; $text = preg_replace('/\[(?:FATHER|MOTHER|CHILD(?:_\d+)?|GRANDPARENT|SIBLING|ATTORNEY(?::\s*[^\]]+)?|JUDGE(?::\s*[^\]]+)?|CASEWORKER(?::\s*[^\]]+)?|EXPERT_WITNESS(?::\s*[^\]]+)?|PERSON(?:_\d+)?)\]/u', '[PERSON]', $text) ?? $text;
return $text; return $text;
} }
@@ -1382,7 +1382,7 @@ PROMPT;
// Replace named role tags (keeping consistent mapping per unique tag) // Replace named role tags (keeping consistent mapping per unique tag)
$text = preg_replace_callback( $text = preg_replace_callback(
'/\[(FATHER|MOTHER|CHILD(?:_\d+)?|GRANDPARENT|SIBLING|ATTORNEY|JUDGE(?::\s*[^\]]+)?|CASEWORKER(?::\s*[^\]]+)?|EXPERT_WITNESS(?::\s*[^\]]+)?|PERSON(?:_\d+)?)\]/u', '/\[(FATHER|MOTHER|CHILD(?:_\d+)?|GRANDPARENT|SIBLING|ATTORNEY(?::\s*[^\]]+)?|JUDGE(?::\s*[^\]]+)?|CASEWORKER(?::\s*[^\]]+)?|EXPERT_WITNESS(?::\s*[^\]]+)?|PERSON(?:_\d+)?)\]/u',
function (array $m) use (&$nameCursor, &$personMap, $norwegianNames): string { function (array $m) use (&$nameCursor, &$personMap, $norwegianNames): string {
$key = $m[1]; $key = $m[1];
if (!isset($personMap[$key])) { if (!isset($personMap[$key])) {
+1 -1
View File
@@ -133,7 +133,7 @@ final class PricingCatalog
'translate' => 1, 'translate' => 1,
'korrespond_refine' => 1, 'korrespond_refine' => 1,
'timeline' => 2, 'timeline' => 2,
'redact' => 2, 'redact' => 1, // minimum (gpt-4o-mini); azure_full overrides to 2 in api/redact.php
'barnevernet' => 3, 'barnevernet' => 3,
'advocate' => 3, 'advocate' => 3,
'korrespond' => 3, 'korrespond' => 3,
+3 -5
View File
@@ -19,10 +19,8 @@ require_once __DIR__ . '/includes/layout.php';
<span class="control-label" data-i18n="redactEngine">Engine</span> <span class="control-label" data-i18n="redactEngine">Engine</span>
<label><input type="radio" name="redactEngine" value="azure_mini" checked id="redactEngineAzureMini"> <span data-i18n="redactEngineAzureMini">Azure gpt-4o-mini</span> &#9733; <small class="control-hint">(fast)</small></label> <label><input type="radio" name="redactEngine" value="azure_mini" checked id="redactEngineAzureMini"> <span data-i18n="redactEngineAzureMini">Azure gpt-4o-mini</span> &#9733; <small class="control-hint">(fast)</small></label>
<label><input type="radio" name="redactEngine" value="azure_full" id="redactEngineAzureFull"> <span data-i18n="redactEngineAzureFull">Azure gpt-4o</span> <small class="control-hint">(best)</small></label> <label><input type="radio" name="redactEngine" value="azure_full" id="redactEngineAzureFull"> <span data-i18n="redactEngineAzureFull">Azure gpt-4o</span> <small class="control-hint">(best)</small></label>
<label><input type="radio" name="redactEngine" value="gpu" id="redactEngineGpu"> <span data-i18n="redactEngineGpu">GPU (cuttlefish)</span> <small class="control-hint">(local)</small></label>
<label><input type="radio" name="redactEngine" value="regex" id="redactEngineRegex"> <span data-i18n="redactEngineRegex">Regex only</span> <small class="control-hint">(free)</small></label>
</div> </div>
<p class="upload-hint" data-i18n="redactEngineHint">Azure engines use your BNL Azure credits. GPU runs the local LiteLLM proxy. Regex-only is instant and free but finds no names or organisations.</p> <p class="upload-hint" data-i18n="redactEngineHint">gpt-4o-mini: 1 credit fast, handles most documents well. gpt-4o: 2 credits higher accuracy for complex or multi-person cases.</p>
<details class="advanced-panel" id="redactAdvanced"> <details class="advanced-panel" id="redactAdvanced">
<summary class="advanced-toggle" data-i18n="redactAdvancedToggle">Advanced settings</summary> <summary class="advanced-toggle" data-i18n="redactAdvancedToggle">Advanced settings</summary>
@@ -95,10 +93,10 @@ require_once __DIR__ . '/includes/layout.php';
</div> </div>
<div class="upload-zone" id="uploadZone" role="region" aria-label="File upload" data-i18n-aria="redactUploadAria"> <div class="upload-zone" id="uploadZone" role="region" aria-label="File upload" data-i18n-aria="redactUploadAria">
<input type="file" id="uploadInput" multiple accept=".pdf,.docx,.txt" aria-label="Choose files"> <input type="file" id="uploadInput" accept=".pdf,.docx,.txt" aria-label="Choose file">
<div id="uploadPrompt" class="upload-prompt"> <div id="uploadPrompt" class="upload-prompt">
<span class="upload-icon" aria-hidden="true">&#8679;</span> <span class="upload-icon" aria-hidden="true">&#8679;</span>
<p><span data-i18n="redactUploadDrop">Drop up to 5 files here, or</span> <label for="uploadInput" class="upload-browse" data-i18n="redactUploadBrowse">browse</label></p> <p><span data-i18n="redactUploadDrop">Drop one file here, or</span> <label for="uploadInput" class="upload-browse" data-i18n="redactUploadBrowse">browse</label></p>
<p class="upload-hint"><strong>PDF</strong>, <strong>DOCX</strong>, <strong>TXT</strong> &mdash; <span data-i18n="redactUploadHint">text extracted in memory, never stored</span></p> <p class="upload-hint"><strong>PDF</strong>, <strong>DOCX</strong>, <strong>TXT</strong> &mdash; <span data-i18n="redactUploadHint">text extracted in memory, never stored</span></p>
</div> </div>
<div id="uploadFileInfo" class="upload-file is-hidden"> <div id="uploadFileInfo" class="upload-file is-hidden">