Files
breakpilot-compliance/backend-compliance/compliance/services/doc_checks/cookie_checks.py
T
Benjamin Admin b363c28539 feat: Add 76 Level-2 regex checks for document correctness verification
Split dsi_document_checker.py (466 LOC) into doc_checks/ package (9 files).
Two-pass L1→L2 logic: L1 checks "Is it mentioned?", L2 checks "Is it correct?"
(e.g. controller has full address, specific Art. 6 lit., concrete time periods).

138 total checks (62 L1 + 76 L2) across 7 doc types:
- DSE Art. 13: 31, Impressum §5 TMG: 16, Cookie §25 TDDDG: 15
- Widerruf §355: 15, AGB §305ff: 21, Social Media Art. 26: 20, DSFA Art. 35: 18

Frontend: hierarchical L1→L2 display with dual progress bars
(green=completeness, blue=correctness).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-05-07 12:37:03 +02:00

180 lines
6.9 KiB
Python

"""
Cookie-Richtlinie checks — §25 TDDDG / ePrivacy.
Level 1: Pflichtangabe erwaehnt?
Level 2: Pflichtangabe korrekt/vollstaendig?
"""
COOKIE_CHECKLIST = [
# ── L1: Arten der Cookies ─────────────────────────────────────────
{
"id": "cookie_types",
"label": "Arten der Cookies",
"level": 1, "parent": None,
"patterns": [
r"(?:notwendig|essentiell|funktional|statistik|marketing|tracking)",
r"cookie.*(?:art|typ|kategori)",
],
"severity": "HIGH",
},
{
"id": "cookie_names_listed",
"label": "Konkrete Cookie-Namen aufgelistet",
"level": 2, "parent": "cookie_types",
"patterns": [
r"(?:_ga|_gid|_gat|_fbp|_gcl|phpsessid|jsessionid|csrf|xsrf)",
r"cookie[\-_]?name\s*[:\|]",
r"name\s+des\s+cookie",
],
"severity": "MEDIUM",
},
{
"id": "cookie_essential_justified",
"label": "Essentiell/Notwendig-Cookies begruendet",
"level": 2, "parent": "cookie_types",
"patterns": [
r"(?:essentiell|notwendig|technisch\s+(?:erforderlich|notwendig)).*(?:funktion|betrieb|sicherheit|warenkorb|session|anmeldung)",
r"(?:unbedingt|zwingend)\s+erforderlich",
],
"severity": "LOW",
},
# ── L1: Zwecke der Cookies ────────────────────────────────────────
{
"id": "purposes",
"label": "Zwecke der Cookies",
"level": 1, "parent": None,
"patterns": [
r"zweck.*cookie", r"cookie.*zweck",
r"(?:wofuer|wozu|warum).*cookie",
r"cookies?\s+(?:ein|ver)?\s*,?\s*um\s+",
r"(?:setzen|verwenden|nutzen)\s+.*cookies?\s+.*(?:um|fuer|für)",
r"(?:analyse|marketing|tracking|funktional)\w*\s*cookies?\s*\.?\s*(?:um|damit|diese|sie)",
r"cookies?\s+(?:dienen|helfen|erm(?:oe|ö)glichen)",
],
"severity": "HIGH",
},
{
"id": "cookie_providers_named",
"label": "Konkrete Anbieter/Dienste benannt",
"level": 2, "parent": "purposes",
"patterns": [
r"(?:google\s+(?:analytics|tag\s+manager|ads)|matomo|piwik|hotjar|hubspot|facebook\s+pixel|meta\s+pixel|linkedin\s+insight|microsoft\s+clarity)",
r"(?:anbieter|provider|dienst)\s*[:\|]\s*[A-Z]",
],
"severity": "MEDIUM",
},
{
"id": "cookie_analytics_named",
"label": "Analytics-/Statistik-Tools konkret benannt",
"level": 2, "parent": "purposes",
"patterns": [
r"google\s+analytics|matomo|piwik|plausible|fathom|adobe\s+analytics|microsoft\s+clarity|hotjar|etracker",
],
"severity": "LOW",
},
{
"id": "cookie_marketing_named",
"label": "Marketing-/Tracking-Tools konkret benannt",
"level": 2, "parent": "purposes",
"patterns": [
r"(?:facebook|meta)\s+pixel|google\s+ads|linkedin\s+insight|tiktok\s+pixel|pinterest\s+tag|criteo|adroll|taboola",
],
"severity": "LOW",
},
# ── L1: Speicherdauer ─────────────────────────────────────────────
{
"id": "retention",
"label": "Speicherdauer der Cookies",
"level": 1, "parent": None,
"patterns": [
r"(?:speicherdauer|laufzeit|g(?:ue|ü)ltigk|ablauf).*cookie",
r"cookie.*(?:\d+\s+(?:tag|monat|jahr)|session)",
],
"severity": "MEDIUM",
},
{
"id": "cookie_duration_values",
"label": "Konkrete Speicherdauern pro Cookie",
"level": 2, "parent": "retention",
"patterns": [
r"\d+\s+(?:tag|monat|jahr|minute|stunde|day|month|year)",
r"session[\-\s]?cookie",
r"(?:ablauf|expiry|laufzeit)\s*[:\|]\s*\d+",
],
"severity": "LOW",
},
# ── L1: Drittanbieter ─────────────────────────────────────────────
{
"id": "third_party",
"label": "Drittanbieter-Cookies",
"level": 1, "parent": None,
"patterns": [
r"drittanbieter", r"third.?party",
r"(?:google|facebook|meta|microsoft).*cookie",
],
"severity": "MEDIUM",
},
{
"id": "cookie_legal_basis",
"label": "Rechtsgrundlage fuer Cookies (§25 TDDDG / Art. 6(1)(a))",
"level": 2, "parent": "third_party",
"patterns": [
r"§\s*25\s*(?:abs\.)?\s*(?:1|2)?\s*tdddg",
r"art\.\s*6\s*(?:abs\.\s*)?1\s*(?:lit\.\s*)?a.*(?:cookie|einwilligung)",
r"einwilligung.*(?:cookie|tracking|marketing)",
r"ttdsg|tdddg|§\s*25",
],
"severity": "MEDIUM",
},
# ── L1: Widerspruch ───────────────────────────────────────────────
{
"id": "opt_out",
"label": "Widerspruchsmoeglichkeit",
"level": 1, "parent": None,
"patterns": [
r"(?:widerspruch|opt.?out|ablehnen|deaktivieren).*cookie",
r"cookie.*(?:ablehnen|deaktivieren|l(?:oe|ö)schen)",
],
"severity": "MEDIUM",
},
{
"id": "cookie_consent_mechanism",
"label": "Consent-Tool/Banner beschrieben",
"level": 2, "parent": "opt_out",
"patterns": [
r"(?:cookie|consent)\s*[\-\s]?(?:banner|hinweis|tool|management|einstellung)",
r"(?:cookiebot|usercentrics|onetrust|borlabs|complianz|klaro|tarteaucitron)",
r"einwilligung\s+(?:jederzeit|widerrufen|zurueckziehen|zur(?:ue|ü)ckziehen)",
],
"severity": "LOW",
},
{
"id": "cookie_browser_settings",
"label": "Browser-Einstellungen zum Cookie-Management",
"level": 2, "parent": "opt_out",
"patterns": [
r"browser[\-\s]?einstellung",
r"(?:in\s+ihrem|im)\s+browser.*(?:cookie|deaktivieren|l(?:oe|ö)schen|blockieren)",
r"(?:chrome|firefox|safari|edge).*(?:cookie|einstellung)",
],
"severity": "LOW",
},
# ── Neue L1: Cookie-Tabelle ───────────────────────────────────────
{
"id": "cookie_table",
"label": "Strukturierte Cookie-Tabelle/Liste",
"level": 1, "parent": None,
"patterns": [
r"(?:cookie[\-\s])?(?:tabelle|uebersicht|übersicht|liste|aufstellung)",
r"(?:name|bezeichnung)\s*[\|\t]\s*(?:anbieter|zweck|dauer|laufzeit)",
r"(?:first[\-\s]?party|third[\-\s]?party)\s*[\|\t]",
],
"severity": "LOW",
},
]