use std::collections::{HashMap, HashSet};
use chrono::{DateTime, Utc};
use serde::{Deserialize, Serialize};
use crate::audit::interpretation::Interpretation;
use crate::audit::module_scores::build_module_scores;
use crate::audit::report::{AuditReport, PerformanceResults, ViewportScores};
use crate::audit::risk_assessment::compute_risk_assessment;
use crate::audit::scoring::{AccessibilityScorer, PrincipleCoverage};
use crate::cli::WcagLevel;
use crate::dark_mode::DarkModeAnalysis;
use crate::mobile::MobileFriendliness;
use crate::security::SecurityAnalysis;
use crate::seo::SeoAnalysis;
use crate::taxonomy::{ReportVisibility, RuleLookup, Scaling, Severity};
use crate::wcag::WcagResults;
use crate::audit::finding_derive::{
calculate_priority_score, derive_bfsg_relevance, derive_complexity, derive_confidence,
derive_expected_impact, derive_false_positive_risk, derive_remediation_priority,
derive_verification,
};
pub use crate::audit::finding_derive::{
complexity_text, expected_impact_text, ComplexityKind, ExpectedImpactKind, ScoreEffect,
};
pub use crate::audit::interactive_finding::{
finding_uncertainty_text, interactive_finding_text, AccessibilityJourney, FindingUncertainty,
InteractiveFinding, InteractiveFindingKind, InteractiveFindingValues, JourneyExecution,
JourneyRun, JourneyStep, JourneyTrace,
};
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct NormalizedReport {
pub url: String,
pub wcag_level: WcagLevel,
pub timestamp: DateTime<Utc>,
pub duration_ms: u64,
pub nodes_analyzed: usize,
pub score: u32,
pub grade: String,
pub certificate: String,
pub findings: Vec<NormalizedFinding>,
pub severity_counts: SeverityCounts,
#[serde(default)]
pub occurrence_counts: SeverityCounts,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub accessibility_assessments: Vec<AccessibilityAssessment>,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub rule_outcomes: Vec<crate::wcag::RuleRun>,
#[serde(default)]
pub execution: crate::audit::AuditExecution,
pub module_scores: Vec<ModuleScoreEntry>,
pub overall_score: u32,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub accessibility_subcategory_scores: Vec<SubcategoryScoreEntry>,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub security_category_scores: Vec<SecurityCategoryScoreEntry>,
pub risk: RiskAssessment,
#[serde(default)]
pub principle_coverage: PrincipleCoverage,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub audit_flags: Vec<AuditFlag>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub consent_privacy: Option<crate::audit::ConsentPrivacySnapshot>,
#[serde(default)]
pub has_screenshots: bool,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub viewport_scores: Option<ViewportScores>,
pub score_calculation_method: String,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub score_breakdown: Option<ScoreBreakdown>,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub interactive_findings: Vec<InteractiveFinding>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub accessibility_journey: Option<AccessibilityJourney>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub screen_reader: Option<crate::screen_reader::ScreenReaderSummary>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub interpretation: Option<Interpretation>,
}
#[derive(Clone)]
pub struct AuditContext<'a> {
pub normalized: NormalizedReport,
pub raw_dual_viewport: Option<&'a crate::audit::report::DualViewportResults>,
pub raw_performance: Option<&'a PerformanceResults>,
pub raw_performance_desktop: Option<&'a PerformanceResults>,
pub raw_seo: Option<&'a SeoAnalysis>,
pub raw_security: Option<&'a SecurityAnalysis>,
pub raw_html_conform: Option<&'a crate::html_conform::HtmlConformAnalysis>,
pub raw_mobile: Option<&'a MobileFriendliness>,
pub raw_ux: Option<&'a crate::ux::UxAnalysis>,
pub raw_journey: Option<&'a crate::journey::JourneyAnalysis>,
pub raw_dark_mode: Option<&'a DarkModeAnalysis>,
pub raw_design_quality: Option<&'a crate::design_quality::DesignQualityAnalysis>,
pub raw_ai_transparency: Option<&'a crate::ai_transparency::AiTransparencyAnalysis>,
pub raw_network_dns: Option<&'a crate::network::dns::NetworkDnsAnalysis>,
pub raw_source_quality: Option<&'a crate::source_quality::SourceQualityAnalysis>,
pub raw_ai_visibility: Option<&'a crate::ai_visibility::AiVisibilityAnalysis>,
pub raw_tech_stack: Option<&'a crate::tech_stack::TechStackAnalysis>,
pub raw_content_visibility: Option<&'a crate::content_visibility::ContentVisibilityAnalysis>,
pub raw_wcag: &'a WcagResults,
pub raw_patterns: Option<&'a crate::patterns::PatternAnalysis>,
pub raw_throttled_performance: &'a [crate::audit::report::ThrottledPerfResult],
pub raw_best_practices: Option<&'a crate::best_practices::BestPracticesAnalysis>,
pub raw_commerce: Option<&'a crate::commerce::CommerceAnalysis>,
}
#[derive(Debug, Clone, Default, Serialize, Deserialize)]
pub struct SeverityCounts {
pub critical: usize,
pub high: usize,
pub medium: usize,
pub low: usize,
pub total: usize,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct AccessibilityAssessment {
pub kind: String,
pub rule_id: String,
pub wcag_criterion: String,
pub severity: Severity,
pub message: String,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub fix_suggestion: Option<String>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub selector: Option<String>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub viewport: Option<String>,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub evidence: Vec<crate::wcag::Evidence>,
}
pub(crate) fn normalize_assessments(results: &WcagResults) -> Vec<AccessibilityAssessment> {
let sources = [
("warning", results.warnings.as_slice()),
("manual_review", results.not_testables.as_slice()),
("positive", results.positives.as_slice()),
];
sources
.into_iter()
.flat_map(|(kind, findings)| {
findings.iter().map(move |finding| AccessibilityAssessment {
kind: kind.to_string(),
rule_id: finding
.rule_id
.clone()
.unwrap_or_else(|| finding.rule.clone()),
wcag_criterion: wcag_criterion_of(&finding.rule),
severity: finding.severity,
message: finding.message.clone(),
fix_suggestion: finding.fix_suggestion.clone(),
selector: finding.selector.clone(),
viewport: finding
.tags
.iter()
.find(|tag| {
matches!(
tag.as_str(),
"desktop-only" | "mobile-only" | "both-viewports"
)
})
.cloned(),
evidence: finding.evidence.clone(),
})
})
.collect()
}
fn default_finding_category() -> String {
"wcag".to_string()
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct NormalizedFinding {
#[serde(default = "default_finding_category")]
pub category: String,
pub rule_id: String,
pub wcag_criterion: String,
pub axe_id: Option<String>,
pub wcag_level: String,
pub dimension: String,
pub subcategory: String,
pub issue_class: String,
#[serde(skip)]
pub dimension_kind: crate::taxonomy::Dimension,
#[serde(skip)]
pub subcategory_kind: crate::taxonomy::Subcategory,
#[serde(skip)]
pub issue_class_kind: crate::taxonomy::IssueClass,
pub severity: Severity,
#[serde(default, skip_serializing_if = "String::is_empty")]
pub user_impact: String,
#[serde(default, skip_serializing_if = "String::is_empty")]
pub technical_impact: String,
pub score_impact: ScoreImpactData,
#[serde(skip)]
pub report_visibility: ReportVisibilityData,
pub aggregation_key: String,
pub title: String,
pub description: String,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub help_url: Option<String>,
pub occurrence_count: usize,
pub priority_score: f32,
#[serde(default, skip_serializing_if = "String::is_empty")]
pub confidence: String,
#[serde(default, skip_serializing_if = "String::is_empty")]
pub false_positive_risk: String,
#[serde(default, skip_serializing_if = "String::is_empty")]
pub verification: String,
#[serde(default, skip_serializing_if = "String::is_empty")]
pub complexity: String,
#[serde(default, skip_serializing_if = "String::is_empty")]
pub complexity_reason: String,
#[serde(default)]
pub complexity_kind: ComplexityKind,
#[serde(default, skip_serializing_if = "String::is_empty")]
pub expected_impact: String,
#[serde(default)]
pub expected_impact_kind: ExpectedImpactKind,
#[serde(default, skip_serializing_if = "String::is_empty")]
pub bfsg_relevance: String,
#[serde(default, skip_serializing_if = "String::is_empty")]
pub remediation_priority: String,
pub occurrences: Vec<OccurrenceDetail>,
}
impl NormalizedFinding {
pub fn is_legal_flag(&self) -> bool {
self.wcag_level == "A"
&& matches!(self.severity, Severity::Critical | Severity::High)
&& !crate::taxonomy::is_convention_rule(&self.rule_id)
}
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ScoreImpactData {
pub base_penalty: f32,
pub max_penalty: f32,
pub scaling: String,
}
#[derive(Debug, Clone, Copy, Serialize, Deserialize)]
pub struct ReportVisibilityData {
pub executive: bool,
pub standard: bool,
pub technical: bool,
}
impl From<&ReportVisibility> for ReportVisibilityData {
fn from(rv: &ReportVisibility) -> Self {
Self {
executive: rv.executive,
standard: rv.standard,
technical: rv.technical,
}
}
}
impl Default for ReportVisibilityData {
fn default() -> Self {
Self {
executive: true,
standard: true,
technical: true,
}
}
}
#[derive(Debug, Clone, Default, Serialize, Deserialize)]
pub struct OccurrenceDetail {
pub node_id: String,
pub message: String,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub selector: Option<String>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub fix_suggestion: Option<String>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub html_snippet: Option<String>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub suggested_code: Option<String>,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub tags: Vec<String>,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub evidence: Vec<crate::wcag::Evidence>,
#[serde(skip)]
pub evidence_screenshot: Option<Vec<u8>>,
#[serde(skip)]
pub evidence_viewport: Option<&'static str>,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ModuleScoreEntry {
pub name: String,
pub score: u32,
#[serde(skip_serializing_if = "Option::is_none")]
pub grade: Option<String>,
pub band: String,
pub weight_pct: u32,
pub contributes_to_overall: bool,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub derived_from: Vec<String>,
pub measurement_type: String,
}
impl ModuleScoreEntry {
pub(crate) fn new(
name: &str,
score: u32,
measurement_type: &str,
contributes_to_overall: bool,
) -> Self {
let weight_pct = crate::taxonomy::module_weight(name);
Self {
name: name.to_string(),
score,
grade: (weight_pct > 0).then(|| {
crate::registry::LETTER_GRADE
.label(score as f32, false)
.to_string()
}),
band: crate::registry::FIVE_BAND
.label(score as f32, true)
.to_string(),
derived_from: crate::taxonomy::module_derived_from(name)
.iter()
.map(|m| m.to_string())
.collect(),
weight_pct,
contributes_to_overall,
measurement_type: measurement_type.to_string(),
}
}
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct SubcategoryScoreEntry {
pub name: String,
#[serde(skip)]
pub subcategory_kind: crate::taxonomy::Subcategory,
pub score: u32,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct SecurityCategoryScoreEntry {
pub name: String,
#[serde(skip)]
pub category_kind: crate::security::SecurityCategory,
pub score: u32,
}
#[derive(
Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash, Serialize, Deserialize, Default,
)]
#[serde(rename_all = "lowercase")]
pub enum RiskLevel {
#[default]
Low,
Medium,
High,
Critical,
}
impl std::fmt::Display for RiskLevel {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
match self {
RiskLevel::Low => write!(f, "Gering"),
RiskLevel::Medium => write!(f, "Mittel"),
RiskLevel::High => write!(f, "Hoch"),
RiskLevel::Critical => write!(f, "Kritisch"),
}
}
}
impl RiskLevel {
pub fn label_localized(&self, i18n: &crate::i18n::I18n) -> String {
let key = match self {
RiskLevel::Low => "risk-level-low",
RiskLevel::Medium => "risk-level-medium",
RiskLevel::High => "risk-level-high",
RiskLevel::Critical => "risk-level-critical",
};
i18n.t(key)
}
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct RiskAssessment {
pub level: RiskLevel,
pub score: u32,
pub threshold: u32,
pub driven_by: String,
pub critical_issues: usize,
pub high_issues: usize,
pub legal_flags: usize,
pub blocking_issues: usize,
pub interactive_critical_issues: usize,
#[serde(default)]
pub interactive_high_issues: usize,
pub summary: String,
}
impl RiskAssessment {
pub fn summary_for(&self, locale: &str) -> String {
if locale != "en" {
return self.summary.clone();
}
match self.level {
RiskLevel::Critical if self.legal_flags >= 3 => format!(
"Critical risk: {} WCAG Level A violations with legal relevance (BFSG). {} blocking issues on interactive controls.",
self.legal_flags, self.blocking_issues
),
RiskLevel::Critical => format!(
"Critical risk: {} critical violations on interactive controls and content.",
self.critical_issues
),
RiskLevel::High => format!(
"High risk: {} critical and {} severe issues. Users are actively excluded.",
self.critical_issues, self.high_issues
),
RiskLevel::Medium => {
if self.interactive_critical_issues > 0 {
format!(
"Medium risk: {} critical interactive findings detected.",
self.interactive_critical_issues
)
} else if self.interactive_high_issues > 0 {
format!(
"Medium risk: {} severe interactive findings detected.",
self.interactive_high_issues
)
} else {
format!(
"Medium risk: {} severe issues detected. Limitations for certain user groups.",
self.high_issues + self.critical_issues
)
}
}
RiskLevel::Low => {
let notable = self.interactive_high_issues + self.interactive_critical_issues;
if notable > 0 {
format!(
"Low risk: no critical violations — keyboard journey has {} requiring manual review.",
if notable == 1 { "1 notable finding".to_string() } else { format!("{notable} notable findings") }
)
} else {
"Low risk: no critical violations — improvement potential remains.".to_string()
}
}
}
}
}
fn gate_certificate_by_risk(
certificate: String,
risk_level: &RiskLevel,
legal_flags: usize,
blocking_issues: usize,
) -> String {
if matches!(risk_level, RiskLevel::Critical) {
return "NICHT BESTANDEN".to_string();
}
let is_positive = matches!(
certificate.as_str(),
"AUSBAUFÄHIG" | "STABIL" | "GUT" | "SEHR GUT"
);
let does_not_pass =
matches!(risk_level, RiskLevel::High) || legal_flags > 0 || blocking_issues > 0;
if is_positive && does_not_pass {
return "EINGESCHRÄNKT".to_string();
}
certificate
}
fn viewport_score_calculation_note() -> String {
"accessibility_score is the 70% mobile / 30% desktop blend of the displayed \
viewport accessibility scores. viewport_scores.weighted_overall blends the \
viewport module scores before security; overall_score is the canonical final \
score after the optional security blend."
.to_string()
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ScoreBreakdown {
pub calculation_note: String,
pub desktop_weight_pct: u32,
pub mobile_weight_pct: u32,
pub desktop_accessibility: u32,
pub mobile_accessibility: u32,
pub viewport_blended_accessibility: u32,
pub desktop_overall: u32,
pub mobile_overall: u32,
pub viewport_blended_overall: u32,
pub viewport_blend_weight_pct: u32,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub security_score: Option<u32>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub security_weight_pct: Option<u32>,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct AuditFlag {
pub kind: String,
pub related_rule: Option<String>,
pub source: String,
pub message: String,
}
const MAX_OCCURRENCES: usize = 5;
fn build_wcag_findings(violations: &[crate::wcag::Violation]) -> Vec<NormalizedFinding> {
let mut groups: HashMap<&str, Vec<&crate::wcag::Violation>> = HashMap::new();
for v in violations {
groups.entry(wcag_group_key(v)).or_default().push(v);
}
let findings: Vec<NormalizedFinding> = groups
.into_iter()
.map(|(rule_id, violations)| {
let first = violations[0];
let taxonomy_rule = RuleLookup::by_legacy_wcag_id(rule_id);
use crate::taxonomy::{Dimension, IssueClass, Subcategory};
let (
tax_id,
dimension_kind,
subcategory_kind,
issue_class_kind,
dimension,
subcategory,
issue_class,
user_impact,
technical_impact,
score_impact,
visibility,
) = if let Some(rule) = taxonomy_rule {
(
rule.id.to_string(),
rule.dimension,
rule.subcategory,
rule.issue_class,
rule.dimension.label(true).to_string(),
rule.subcategory.label(true).to_string(),
rule.issue_class.label(true).to_string(),
rule.user_impact_en.to_string(),
rule.technical_impact_en.to_string(),
ScoreImpactData {
base_penalty: rule.score_impact.base_penalty,
max_penalty: rule.score_impact.max_penalty,
scaling: match rule.score_impact.occurrence_scaling {
Scaling::Logarithmic => "logarithmic".to_string(),
Scaling::Linear => "linear".to_string(),
Scaling::Fixed => "fixed".to_string(),
},
},
ReportVisibilityData::from(&rule.report_visibility),
)
} else {
(
format!("unknown.{}", rule_id),
Dimension::Accessibility,
Subcategory::ContentAlternatives,
IssueClass::Missing,
"Accessibility".to_string(),
"Unknown".to_string(),
"Unknown".to_string(),
String::new(),
String::new(),
ScoreImpactData {
base_penalty: 0.0,
max_penalty: 0.0,
scaling: "unknown".to_string(),
},
ReportVisibilityData::default(),
)
};
let occurrence_count = violations.len();
let mut seen_selectors = std::collections::HashSet::new();
let mut occurrences: Vec<OccurrenceDetail> = violations
.iter()
.filter(|v| {
let key = v.selector.as_deref().unwrap_or(&v.node_id);
seen_selectors.insert(key.to_string())
})
.take(MAX_OCCURRENCES)
.map(|v| OccurrenceDetail {
node_id: v.node_id.clone(),
message: v.message.clone(),
selector: v.selector.clone(),
fix_suggestion: v.fix_suggestion.clone(),
html_snippet: v.html_snippet.clone(),
suggested_code: v.suggested_code.clone(),
tags: v.tags.clone(),
evidence: v.evidence.clone(),
evidence_screenshot: v.evidence_screenshot.clone(),
evidence_viewport: v.evidence_viewport,
})
.collect();
if let Some(shared_fix) = occurrences.first().and_then(|o| o.fix_suggestion.clone()) {
if occurrences[1..]
.iter()
.all(|o| o.fix_suggestion.as_deref() == Some(shared_fix.as_str()))
{
for occ in &mut occurrences[1..] {
occ.fix_suggestion = None;
}
}
}
let axe_id = first
.rule_id
.clone()
.or_else(|| taxonomy_rule.and_then(|r| r.axe_id).map(String::from));
let severity = violations
.iter()
.map(|v| v.severity)
.max()
.unwrap_or(first.severity);
let priority_score = calculate_priority_score(severity, occurrence_count, &tax_id);
let subcategory_de = subcategory_kind.label(false);
let issue_class_de = issue_class_kind.label(false);
let confidence = derive_confidence(&tax_id, subcategory_de, issue_class_de);
let false_positive_risk =
derive_false_positive_risk(&tax_id, subcategory_de, issue_class_de);
let verification = derive_verification(&false_positive_risk);
let (complexity, complexity_kind) =
derive_complexity(occurrence_count, &tax_id, issue_class_de);
let complexity_reason = complexity_text(complexity_kind, true);
let expected_impact_kind = derive_expected_impact(
severity,
occurrence_count,
"wcag",
first.level.to_string().as_str(),
);
let expected_impact = expected_impact_text(&expected_impact_kind, true);
let bfsg_relevance = derive_bfsg_relevance(
&tax_id,
"wcag",
&wcag_criterion_of(&first.rule),
first.level.to_string().as_str(),
severity,
);
let remediation_priority =
derive_remediation_priority(severity, occurrence_count, &complexity);
let display_title = taxonomy_rule
.map(|r| r.title_en.to_string())
.unwrap_or_else(|| first.rule_name.clone());
NormalizedFinding {
category: "wcag".to_string(),
rule_id: tax_id.clone(),
wcag_criterion: wcag_criterion_of(&first.rule),
axe_id,
wcag_level: first.level.to_string(),
dimension,
subcategory,
issue_class,
dimension_kind,
subcategory_kind,
issue_class_kind,
severity,
user_impact,
technical_impact,
score_impact,
report_visibility: visibility,
aggregation_key: tax_id,
title: display_title,
description: first.message.clone(),
help_url: first.help_url.clone(),
occurrence_count,
priority_score,
confidence,
false_positive_risk,
verification,
complexity,
complexity_reason,
complexity_kind,
expected_impact,
expected_impact_kind,
bfsg_relevance,
remediation_priority,
occurrences,
}
})
.collect();
findings
}
#[derive(Debug, Clone)]
pub struct BatchFinding {
pub finding: NormalizedFinding,
pub urls: Vec<String>,
}
pub fn aggregate_batch_findings(reports: &[NormalizedReport]) -> Vec<BatchFinding> {
let mut aggregated: Vec<BatchFinding> = Vec::new();
let mut index: HashMap<String, usize> = HashMap::new();
for report in reports {
for finding in &report.findings {
match index.get(&finding.aggregation_key) {
Some(&i) => {
let entry = &mut aggregated[i];
entry.finding.occurrence_count += finding.occurrence_count;
entry.finding.severity = entry.finding.severity.max(finding.severity);
if !entry.urls.contains(&report.url) {
entry.urls.push(report.url.clone());
}
}
None => {
index.insert(finding.aggregation_key.clone(), aggregated.len());
aggregated.push(BatchFinding {
finding: finding.clone(),
urls: vec![report.url.clone()],
});
}
}
}
}
let wcag_titles: HashSet<String> = aggregated
.iter()
.filter(|b| b.finding.category == "wcag")
.map(|b| b.finding.title.trim().to_lowercase())
.collect();
aggregated.retain(|b| {
b.finding.category == "wcag"
|| !wcag_titles.contains(&b.finding.title.trim().to_lowercase())
});
for entry in &mut aggregated {
rederive_count_dependent_fields(&mut entry.finding);
}
aggregated.sort_by(|a, b| crate::audit::prioritization::action_order(&a.finding, &b.finding));
aggregated
}
fn rederive_count_dependent_fields(f: &mut NormalizedFinding) {
let count = f.occurrence_count;
let issue_class_de = if f.category == "wcag" {
f.issue_class_kind.label(false)
} else {
"issue"
};
f.priority_score = calculate_priority_score(f.severity, count, &f.rule_id);
let (complexity, complexity_kind) = derive_complexity(count, &f.rule_id, issue_class_de);
f.complexity_reason = complexity_text(complexity_kind, true);
f.complexity = complexity;
f.complexity_kind = complexity_kind;
f.expected_impact_kind = derive_expected_impact(f.severity, count, &f.category, &f.wcag_level);
f.expected_impact = expected_impact_text(&f.expected_impact_kind, true);
f.remediation_priority = derive_remediation_priority(f.severity, count, &f.complexity);
f.bfsg_relevance = derive_bfsg_relevance(
&f.rule_id,
&f.category,
&f.wcag_criterion,
&f.wcag_level,
f.severity,
);
}
pub(crate) fn wcag_group_key(violation: &crate::wcag::Violation) -> &str {
if let Some(axe_id) = violation.rule_id.as_deref() {
if RuleLookup::by_legacy_wcag_id(axe_id).is_some() {
return axe_id;
}
}
violation.rule.as_str()
}
fn compute_accessibility_subcategory_scores(
violations: &[crate::wcag::Violation],
) -> Vec<SubcategoryScoreEntry> {
use crate::taxonomy::Subcategory;
const ACCESSIBILITY_SUBCATEGORIES: [Subcategory; 7] = [
Subcategory::ContentAlternatives,
Subcategory::StructureSemantics,
Subcategory::NavigationInteraction,
Subcategory::FormsInteraction,
Subcategory::LanguageClarity,
Subcategory::TechnicalRobustness,
Subcategory::VisualPresentation,
];
let subcategory_for = |v: &crate::wcag::Violation| -> Subcategory {
RuleLookup::by_legacy_wcag_id(wcag_group_key(v))
.map(|r| r.subcategory)
.unwrap_or_default()
};
ACCESSIBILITY_SUBCATEGORIES
.into_iter()
.map(|subcategory| {
let subset: Vec<crate::wcag::Violation> = violations
.iter()
.filter(|v| subcategory_for(v) == subcategory)
.cloned()
.collect();
let score = AccessibilityScorer::calculate_score(&subset).round() as u32;
SubcategoryScoreEntry {
name: subcategory.label(true).to_string(),
subcategory_kind: subcategory,
score,
}
})
.collect()
}
pub fn seo_heading_finding_text(
issue_type: &str,
en: bool,
fallback_message: &str,
) -> (String, String) {
let title = match (issue_type, en) {
("long_heading", true) => "Heading too long".to_string(),
("long_heading", false) => "Überschrift zu lang".to_string(),
("missing_h1", true) => "Missing H1 heading".to_string(),
("missing_h1", false) => "Fehlende H1-Überschrift".to_string(),
("multiple_h1", true) => "Multiple H1 headings".to_string(),
("multiple_h1", false) => "Mehrere H1-Überschriften".to_string(),
("skipped_level", true) => "Skipped heading level".to_string(),
("skipped_level", false) => "Übersprungene Überschriftenebene".to_string(),
("empty_heading", true) => "Empty heading".to_string(),
("empty_heading", false) => "Leere Überschrift".to_string(),
(other, _) => other.replace('_', " "),
};
let technical_impact = match (issue_type, en) {
("skipped_level", true) => "Skipped heading levels break the tree structure for screen readers and SEO crawlers — keep a logical H1→H2→H3 hierarchy.".to_string(),
("skipped_level", false) => "Übersprungene Heading-Ebenen zerstören die Baumstruktur für Screenreader und SEO-Crawler — logische Hierarchie H1→H2→H3 einhalten.".to_string(),
("missing_h1", true) => "Missing H1 heading — page purpose not recognizable for search engines and screen readers.".to_string(),
("missing_h1", false) => "Fehlende H1-Überschrift — Seitenzweck für Suchmaschinen und Screenreader nicht erkennbar.".to_string(),
("multiple_h1", true) => "Multiple H1 headings undermine the content hierarchy; search engines cannot derive a single main focus.".to_string(),
("multiple_h1", false) => "Mehrere H1-Überschriften untergraben die inhaltliche Hierarchie; Suchmaschinen können keinen eindeutigen Hauptfokus ableiten.".to_string(),
("long_heading", true) => "Overly long headings are truncated in SERPs and make quick scanning harder for users.".to_string(),
("long_heading", false) => "Überlange Überschriften werden in SERPs abgeschnitten und erschweren das schnelle Scannen für Nutzer.".to_string(),
("empty_heading", true) => "Empty headings cause navigation problems for screen readers and are treated as a poor signal by SEO crawlers.".to_string(),
("empty_heading", false) => "Leere Überschriften erzeugen Navigationsprobleme für Screenreader und werden von SEO-Crawlern als schlechtes Signal gewertet.".to_string(),
_ => fallback_message.to_string(),
};
(title, technical_impact)
}
fn aggregate_seo_findings(
seo: &crate::seo::SeoAnalysis,
max_occurrences: usize,
) -> Vec<NormalizedFinding> {
let mut heading_groups: HashMap<&str, Vec<&crate::seo::HeadingIssue>> = HashMap::new();
for issue in &seo.headings.issues {
heading_groups
.entry(&issue.issue_type)
.or_default()
.push(issue);
}
let mut findings = Vec::new();
for (issue_type, issues) in heading_groups {
let first = issues[0];
let occurrence_count = issues.len();
let rule_id = format!("seo.headings.{}", issue_type);
let (title, technical_impact) = seo_heading_finding_text(issue_type, true, &first.message);
let priority_score = calculate_priority_score(first.severity, occurrence_count, &rule_id);
let confidence = derive_confidence(&rule_id, "Content", "issue");
let false_positive_risk = derive_false_positive_risk(&rule_id, "Content", "issue");
let verification = derive_verification(&false_positive_risk);
let (complexity, complexity_kind) = derive_complexity(occurrence_count, &rule_id, "issue");
let complexity_reason = complexity_text(complexity_kind, true);
let expected_impact_kind =
derive_expected_impact(first.severity, occurrence_count, "seo", "");
let expected_impact = expected_impact_text(&expected_impact_kind, true);
let bfsg_relevance = derive_bfsg_relevance("", "seo", "", "", first.severity);
let remediation_priority =
derive_remediation_priority(first.severity, occurrence_count, &complexity);
findings.push(NormalizedFinding {
category: "seo".to_string(),
rule_id: rule_id.clone(),
wcag_criterion: String::new(),
axe_id: None,
wcag_level: String::new(),
dimension: "SEO".to_string(),
subcategory: "Content".to_string(),
issue_class: "issue".to_string(),
dimension_kind: crate::taxonomy::Dimension::Seo,
subcategory_kind: crate::taxonomy::Subcategory::ContentStructure,
issue_class_kind: crate::taxonomy::IssueClass::Weak,
severity: first.severity,
user_impact: String::new(),
technical_impact,
score_impact: ScoreImpactData {
base_penalty: 0.0,
max_penalty: 0.0,
scaling: "none".to_string(),
},
report_visibility: ReportVisibilityData::default(),
aggregation_key: rule_id,
title,
description: first.message.clone(),
help_url: None,
occurrence_count,
priority_score,
confidence,
false_positive_risk,
verification,
complexity,
complexity_reason,
complexity_kind,
expected_impact,
expected_impact_kind,
bfsg_relevance,
remediation_priority,
occurrences: issues
.iter()
.take(max_occurrences)
.map(|i| OccurrenceDetail {
node_id: i.issue_type.clone(),
message: i.message.clone(),
selector: None,
fix_suggestion: None,
html_snippet: None,
suggested_code: None,
tags: vec!["seo".to_string()],
..Default::default()
})
.collect(),
});
}
findings
}
fn filter_aria_hidden_interactive(
interactive_findings: &mut Vec<InteractiveFinding>,
findings: &[NormalizedFinding],
wcag_violations: &[crate::wcag::Violation],
) {
let mut aria_hidden_selectors: std::collections::HashSet<String> = wcag_violations
.iter()
.filter(|v| v.rule_id.as_deref() == Some("keyboard/hidden-focusable"))
.filter_map(|v| v.selector.clone())
.collect();
aria_hidden_selectors.extend(
findings
.iter()
.filter(|f| f.rule_id == "a11y.aria_hidden_focus.invalid")
.flat_map(|f| f.occurrences.iter().filter_map(|o| o.selector.clone())),
);
if aria_hidden_selectors.is_empty() {
return;
}
fn normalize_selector(sel: &str) -> String {
let mut normalized = String::new();
let mut chars = sel.chars().peekable();
while let Some(c) = chars.next() {
if c == '.' {
while let Some(&next_c) = chars.peek() {
if next_c.is_alphanumeric() || next_c == '-' || next_c == '_' {
chars.next();
} else {
break;
}
}
} else if c.is_alphabetic() {
let mut tag = String::new();
tag.push(c);
while let Some(&next_c) = chars.peek() {
if next_c.is_alphanumeric() || next_c == '-' || next_c == '_' {
tag.push(chars.next().unwrap());
} else {
break;
}
}
if chars.peek() == Some(&'#') {
} else {
normalized.push_str(&tag);
}
} else if c == ' ' {
let is_around_gt = normalized.ends_with('>') || chars.peek() == Some(&'>');
if !is_around_gt {
normalized.push(' ');
}
} else {
normalized.push(c);
}
}
normalized.trim().replace(" >", ">").replace("> ", ">")
}
fn extract_selector_from_message(message: &str) -> Option<String> {
let start_idx = message.find('(')?;
let mut depth = 0;
let mut end_idx = None;
for (i, c) in message[start_idx..].char_indices() {
if c == '(' {
depth += 1;
} else if c == ')' {
depth -= 1;
if depth == 0 {
end_idx = Some(start_idx + i);
break;
}
}
}
let mut sel = &message[start_idx + 1..end_idx?];
if let Some(colon_idx) = sel.find(": display:") {
sel = &sel[..colon_idx];
} else if let Some(colon_idx) = sel.find(": visibility:") {
sel = &sel[..colon_idx];
} else if let Some(colon_idx) = sel.find(": opacity:") {
sel = &sel[..colon_idx];
}
Some(sel.trim().to_string())
}
let normalized_aria_hidden_selectors: std::collections::HashSet<String> = aria_hidden_selectors
.iter()
.map(|s| normalize_selector(s))
.collect();
interactive_findings.retain(|inf| {
if inf.category == "HiddenFocusable" {
if let Some(sel) = extract_selector_from_message(&inf.message) {
let norm_sel = normalize_selector(&sel);
!normalized_aria_hidden_selectors.iter().any(|s| {
norm_sel == *s
|| norm_sel.ends_with(&format!(">{}", s))
|| s.ends_with(&format!(">{}", norm_sel))
})
} else {
true
}
} else {
true
}
});
}
pub fn normalize<'a>(report: &'a AuditReport) -> AuditContext<'a> {
let violations = &report.accessibility.wcag_results.violations;
let seo_reports_lang = report
.discoverability
.seo
.as_ref()
.is_some_and(|s| s.technical.has_lang);
let had_311 = violations.iter().any(|v| v.rule == "3.1.1");
let mut findings = build_wcag_findings(violations);
if let Some(seo) = &report.discoverability.seo {
findings.extend(aggregate_seo_findings(seo, MAX_OCCURRENCES));
}
findings.sort_by(crate::audit::prioritization::action_order);
let mut interactive_findings = report.interactive_findings.clone();
filter_aria_hidden_interactive(&mut interactive_findings, &findings, violations);
let score = report
.viewport_scores
.as_ref()
.map(ViewportScores::weighted_accessibility)
.unwrap_or_else(|| report.accessibility.score.round().max(1.0) as u32);
let wcag_findings: Vec<_> = findings.iter().filter(|f| f.category == "wcag").collect();
let severity_counts = SeverityCounts {
critical: wcag_findings
.iter()
.filter(|f| f.severity == Severity::Critical)
.count(),
high: wcag_findings
.iter()
.filter(|f| f.severity == Severity::High)
.count(),
medium: wcag_findings
.iter()
.filter(|f| f.severity == Severity::Medium)
.count(),
low: wcag_findings
.iter()
.filter(|f| f.severity == Severity::Low)
.count(),
total: wcag_findings.len(),
};
let occurrence_counts = SeverityCounts {
critical: wcag_findings
.iter()
.filter(|f| f.severity == Severity::Critical)
.map(|f| f.occurrence_count)
.sum(),
high: wcag_findings
.iter()
.filter(|f| f.severity == Severity::High)
.map(|f| f.occurrence_count)
.sum(),
medium: wcag_findings
.iter()
.filter(|f| f.severity == Severity::Medium)
.map(|f| f.occurrence_count)
.sum(),
low: wcag_findings
.iter()
.filter(|f| f.severity == Severity::Low)
.map(|f| f.occurrence_count)
.sum(),
total: wcag_findings.iter().map(|f| f.occurrence_count).sum(),
};
let vuln_security_penalty: u32 = report
.best_practices
.as_ref()
.map(|bp| {
bp.vulnerable_libraries
.vulnerable
.iter()
.map(|v| match v.severity.as_str() {
"high" => 15,
"medium" => 8,
_ => 3,
})
.sum::<u32>()
.min(30)
})
.unwrap_or(0);
let module_scores =
build_module_scores(report, score, &occurrence_counts, vuln_security_penalty);
let (overall_score, score_calculation_method, score_breakdown) =
if let Some(ref vs) = report.viewport_scores {
let security_adjusted = report
.security
.as_ref()
.map(|s| s.score.saturating_sub(vuln_security_penalty));
let blend_weight = if security_adjusted.is_some() {
90u32
} else {
100u32
};
let mut weighted = vs.weighted_overall as f64 * blend_weight as f64;
let mut total = blend_weight as f64;
if let Some(sec) = security_adjusted {
weighted += sec as f64 * 10.0;
total += 10.0;
}
let breakdown = ScoreBreakdown {
calculation_note: viewport_score_calculation_note(),
desktop_weight_pct: 30,
mobile_weight_pct: 70,
desktop_accessibility: vs.desktop.accessibility,
mobile_accessibility: vs.mobile.accessibility,
viewport_blended_accessibility: vs.weighted_accessibility(),
desktop_overall: vs.desktop.overall,
mobile_overall: vs.mobile.overall,
viewport_blended_overall: vs.weighted_overall,
viewport_blend_weight_pct: blend_weight,
security_score: security_adjusted,
security_weight_pct: security_adjusted.map(|_| 10u32),
};
(
(weighted / total).round() as u32,
"viewport_weighted".to_string(),
Some(breakdown),
)
} else {
let contributing_modules = module_scores.iter().filter(|m| m.contributes_to_overall);
let (weighted_sum, total_weight) =
contributing_modules.fold((0.0, 0.0), |(sum, total), module| {
(
sum + module.score as f64 * module.weight_pct as f64,
total + module.weight_pct as f64,
)
});
(
(weighted_sum / total_weight).round() as u32,
"module_weighted".to_string(),
None,
)
};
let grade = AccessibilityScorer::calculate_grade(score as f32).to_string();
let certificate = AccessibilityScorer::calculate_certificate(score as f32).to_string();
let mut audit_flags = Vec::new();
if report.accessibility.execution.quality.qualified_results {
audit_flags.push(AuditFlag {
kind: "incomplete_audit".to_string(),
related_rule: None,
source: "audit.execution".to_string(),
message: format!(
"Audit coverage is {:?}: {} rule checks failed and {} modules were partial or failed. Scores only describe the successfully measured scope.",
report.accessibility.execution.quality.status,
report.accessibility.execution.quality.failed_rule_checks,
report.accessibility.execution.quality.partial_or_failed_modules,
),
});
}
if seo_reports_lang && had_311 {
audit_flags.push(AuditFlag {
kind: "conflicting_signal".to_string(),
related_rule: Some("3.1.1".to_string()),
source: "seo.technical.has_lang".to_string(),
message: "SEO detected a language declaration while WCAG still reported 3.1.1. The finding remains in the report and should be verified against the rendered DOM.".to_string(),
});
}
if let Some(ref vs) = report.viewport_scores {
let desktop_a11y = vs.desktop.accessibility as i32;
let mobile_a11y = vs.mobile.accessibility as i32;
let gap = (desktop_a11y - mobile_a11y).abs();
if gap >= 20 {
let (higher, lower, higher_score, lower_score) = if desktop_a11y >= mobile_a11y {
("Desktop", "Mobile", desktop_a11y, mobile_a11y)
} else {
("Mobile", "Desktop", mobile_a11y, desktop_a11y)
};
audit_flags.push(AuditFlag {
kind: "viewport_gap".to_string(),
related_rule: None,
source: "viewport_scores.accessibility".to_string(),
message: format!(
"{higher} scored {higher_score}, {lower} scored {lower_score} — a {gap}-point gap suggests {lower}-specific rendering differences (e.g. lazy-loaded components, injected markup, or different DOM paths) rather than a site-wide failure.",
),
});
}
}
if report.consent_banner_detected {
let msg = if report.consent_banner_dismissed {
format!(
"Consent banner detected and automatically dismissed{}. Audit results reflect the page content after consent.",
report.consent_banner_cmp.as_ref().map(|c| format!(" ({})", c)).unwrap_or_default()
)
} else {
format!(
"Consent banner detected{} — audit performed without consent. Accessibility and SEO results may be incomplete. Recommendation: use --dismiss-consent.",
report.consent_banner_cmp.as_ref().map(|c| format!(" ({})", c)).unwrap_or_default()
)
};
audit_flags.push(AuditFlag {
kind: "consent_banner".to_string(),
related_rule: None,
source: "browser.consent".to_string(),
message: msg,
});
if !report.consent_banner_dismissed {
let violation_count: usize = report.accessibility.wcag_results.violations.len();
if report.accessibility.nodes_analyzed > 0
&& violation_count > report.accessibility.nodes_analyzed
{
audit_flags.push(AuditFlag {
kind: "consent_wall_artifact".to_string(),
related_rule: None,
source: "browser.consent".to_string(),
message: format!(
"Possible consent wall artifact: {} violations with only {} analyzed nodes. \
Scores may be measuring the consent dialog rather than the actual page content. \
Recommendation: re-run the audit with --dismiss-consent.",
violation_count, report.accessibility.nodes_analyzed
),
});
}
}
}
let has_broken_skip_link = interactive_findings
.iter()
.any(|f| f.category == "SkipLink");
let has_bypass_blocks_violation = findings
.iter()
.any(|f| f.rule_id == "a11y.bypass_blocks.missing");
if has_broken_skip_link && !has_bypass_blocks_violation {
audit_flags.push(AuditFlag {
kind: "bypass_blocks_untested".to_string(),
related_rule: Some("a11y.bypass_blocks.missing".to_string()),
source: "a11y_journey.skip_link".to_string(),
message: "A skip link is present (WCAG 2.4.1 static check passed) but the \
journey found it does not move keyboard focus to the target. The page \
effectively fails WCAG 2.4.1 — verify and fix the skip-link target."
.to_string(),
});
}
let screen_reader = report
.screen_reader_audit
.as_ref()
.map(crate::screen_reader::ScreenReaderSummary::from_report);
let risk = compute_risk_assessment(
&findings,
&occurrence_counts,
&interactive_findings,
screen_reader.as_ref(),
score,
overall_score,
);
let certificate = gate_certificate_by_risk(
certificate,
&risk.level,
risk.legal_flags,
risk.blocking_issues,
);
let accessibility_subcategory_scores = compute_accessibility_subcategory_scores(violations);
let security_category_scores = report
.security
.as_ref()
.map(|s| {
crate::security::calculate_security_category_scores(&s.issues)
.into_iter()
.map(|(category, score)| SecurityCategoryScoreEntry {
name: category.label(true).to_string(),
category_kind: category,
score,
})
.collect()
})
.unwrap_or_default();
let normalized_data = NormalizedReport {
url: report.url.clone(),
wcag_level: report.wcag_level,
timestamp: report.timestamp,
duration_ms: report.duration_ms,
nodes_analyzed: report.accessibility.nodes_analyzed,
score,
grade,
certificate,
findings,
severity_counts,
occurrence_counts,
accessibility_assessments: normalize_assessments(&report.accessibility.wcag_results),
rule_outcomes: report.accessibility.wcag_results.rule_outcomes.clone(),
execution: report.accessibility.execution.clone(),
module_scores,
overall_score,
accessibility_subcategory_scores,
security_category_scores,
risk,
principle_coverage: AccessibilityScorer::calculate_coverage(violations),
audit_flags,
consent_privacy: report.consent_privacy.clone(),
has_screenshots: report.page_screenshots.is_some(),
viewport_scores: report.viewport_scores.clone(),
score_calculation_method,
score_breakdown,
interactive_findings,
accessibility_journey: report.accessibility_journey.clone(),
screen_reader,
interpretation: None,
};
let mut ctx = AuditContext {
normalized: normalized_data,
raw_dual_viewport: report.dual_viewport.as_ref(),
raw_performance: report.performance.as_ref(),
raw_performance_desktop: report
.dual_viewport
.as_ref()
.and_then(|d| d.desktop.performance.as_ref()),
raw_seo: report.discoverability.seo.as_ref(),
raw_security: report.security.as_ref(),
raw_html_conform: report.html_conform.as_ref(),
raw_mobile: report.experience.mobile.as_ref(),
raw_ux: report.ux.as_ref(),
raw_journey: report.journey.as_ref(),
raw_dark_mode: report.experience.dark_mode.as_ref(),
raw_design_quality: report.experience.design_quality.as_ref(),
raw_ai_transparency: report.experience.ai_transparency.as_ref(),
raw_network_dns: report.experience.network_dns.as_ref(),
raw_source_quality: report.discoverability.source_quality.as_ref(),
raw_ai_visibility: report.discoverability.ai_visibility.as_ref(),
raw_tech_stack: report.discoverability.tech_stack.as_ref(),
raw_content_visibility: report.discoverability.content_visibility.as_ref(),
raw_wcag: &report.accessibility.wcag_results,
raw_patterns: report.patterns.as_ref(),
raw_throttled_performance: &report.throttled_performance,
raw_best_practices: report.best_practices.as_ref(),
raw_commerce: report.commerce.as_ref(),
};
ctx.normalized.interpretation = Some(Interpretation::from_context(&ctx));
ctx
}
pub(crate) fn wcag_criterion_of(rule: &str) -> String {
crate::taxonomy::criterion_for_rule(rule).unwrap_or_else(|| rule.to_string())
}
#[cfg(test)]
mod tests {
use super::*;
use crate::wcag::{Violation, WcagResults};
#[test]
fn slug_page_rules_normalize_to_their_criterion() {
for slug in ["aria-prohibited-attr", "aria-hidden-focus", "frame-tested"] {
assert_eq!(wcag_criterion_of(slug), "4.1.2", "{slug}");
}
}
#[test]
fn display_convention_rules_are_not_legal_flags() {
let mut results = WcagResults::new();
results.add_violation(
Violation::new(
"1.1.1",
"Display modes: text layer reachable",
WcagLevel::A,
Severity::High,
"hidden",
"div#desc",
)
.with_rule_id("display/text-hidden")
.with_tags(vec!["best-practice".into()]),
);
let report = AuditReport::new("https://example.com".into(), WcagLevel::AA, results, 1);
let norm = normalize(&report).normalized;
let finding = &norm.findings[0];
assert_eq!(finding.rule_id, "a11y.display_text_layer.hidden");
assert_eq!(finding.bfsg_relevance, "low");
assert!(!finding.is_legal_flag());
assert_eq!(norm.risk.legal_flags, 0);
let mut results = WcagResults::new();
results.add_violation(Violation::new(
"1.1.1",
"Non-text Content",
WcagLevel::A,
Severity::High,
"Missing alt",
"n1",
));
let report = AuditReport::new("https://example.com".into(), WcagLevel::AA, results, 1);
let norm = normalize(&report).normalized;
assert_eq!(norm.findings[0].bfsg_relevance, "high");
assert_eq!(norm.risk.legal_flags, 1);
}
#[test]
fn aaa_reduced_motion_is_not_a_legal_flag() {
let mut results = WcagResults::new();
results.add_violation(
Violation::new(
"2.3.3",
"Animation from Interactions",
WcagLevel::AAA,
Severity::High,
"motion",
"stylesheet",
)
.with_rule_id("prefers-reduced-motion"),
);
let report = AuditReport::new("https://example.com".into(), WcagLevel::AAA, results, 1);
let norm = normalize(&report).normalized;
let finding = &norm.findings[0];
assert_eq!(finding.rule_id, "a11y.motion.reduced_motion");
assert_eq!(finding.wcag_level, "AAA");
assert_eq!(finding.bfsg_relevance, "low");
assert!(!finding.is_legal_flag());
assert_eq!(norm.risk.legal_flags, 0);
}
#[test]
fn best_practice_name_rules_have_own_taxonomy_and_no_legal_flag() {
for (axe_id, taxonomy_id) in [
("title-only-description", "a11y.title_only_description.weak"),
(
"description-duplicates-name",
"a11y.description_duplicates_name.redundant",
),
] {
let mut results = WcagResults::new();
results.add_violation(
Violation::new("4.1.2", "x", WcagLevel::A, Severity::High, "x", "n1")
.with_rule_id(axe_id)
.with_tags(vec!["best-practice".into()]),
);
let report = AuditReport::new("https://example.com".into(), WcagLevel::AA, results, 1);
let norm = normalize(&report).normalized;
let finding = &norm.findings[0];
assert_eq!(finding.rule_id, taxonomy_id, "{axe_id}");
assert_eq!(finding.wcag_criterion, "4.1.2", "{axe_id}");
assert_eq!(finding.bfsg_relevance, "low", "{axe_id}");
assert!(!finding.is_legal_flag(), "{axe_id}");
assert_eq!(norm.risk.legal_flags, 0, "{axe_id}");
}
}
#[test]
fn test_normalize_empty() {
let report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
WcagResults::new(),
100,
);
let norm = normalize(&report);
assert_eq!(norm.normalized.score, 100);
assert_eq!(norm.normalized.grade, "A");
assert_eq!(norm.normalized.certificate, "SEHR GUT");
assert!(norm.normalized.findings.is_empty());
assert_eq!(norm.normalized.severity_counts.total, 0);
}
#[test]
fn test_normalize_groups_by_rule() {
let mut results = WcagResults::new();
results.add_violation(Violation::new(
"1.1.1",
"Non-text Content",
WcagLevel::A,
Severity::High,
"Missing alt 1",
"n1",
));
results.add_violation(Violation::new(
"1.1.1",
"Non-text Content",
WcagLevel::A,
Severity::High,
"Missing alt 2",
"n2",
));
results.add_violation(Violation::new(
"2.4.4",
"Link Purpose",
WcagLevel::A,
Severity::Medium,
"Unclear link",
"n3",
));
let report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
results,
100,
);
let norm = normalize(&report);
assert_eq!(norm.normalized.findings.len(), 2);
let alt = norm
.normalized
.findings
.iter()
.find(|f| f.wcag_criterion == "1.1.1")
.unwrap();
assert_eq!(alt.occurrence_count, 2);
assert!(alt.priority_score > 0.0);
assert_eq!(alt.occurrences.len(), 2);
assert_eq!(alt.dimension, "Accessibility");
assert!(!alt.rule_id.is_empty());
}
#[test]
fn test_normalize_taxonomy_fields() {
let mut results = WcagResults::new();
results.add_violation(Violation::new(
"1.1.1",
"Non-text Content",
WcagLevel::A,
Severity::High,
"Missing alt",
"n1",
));
let report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
results,
100,
);
let norm = normalize(&report);
let finding = &norm.normalized.findings[0];
assert_eq!(finding.rule_id, "a11y.alt_text.missing");
assert_eq!(finding.dimension, "Accessibility");
assert_eq!(finding.subcategory, "Content & Alternatives");
assert_eq!(finding.issue_class, "Missing");
assert!(finding.score_impact.base_penalty > 0.0);
assert!(finding.score_impact.max_penalty >= finding.score_impact.base_penalty);
assert!(!finding.score_impact.scaling.is_empty());
assert!(!finding.user_impact.is_empty());
assert_eq!(finding.title, "Missing alternative text on images");
assert_eq!(
finding.user_impact,
"Screen reader users receive no image information."
);
assert_eq!(finding.technical_impact, "Non-conformant image markup.");
for field in [
&finding.title,
&finding.user_impact,
&finding.technical_impact,
] {
assert!(
!field.chars().any(|c| "äöüÄÖÜß".contains(c)),
"canonical-English JSON field contains German diacritics: {field}"
);
}
}
#[test]
fn test_frame_title_keeps_wcag_241_and_specific_taxonomy() {
let mut results = WcagResults::new();
results.add_violation(
Violation::new(
"2.4.1",
"Frame title",
WcagLevel::A,
Severity::High,
"Iframe is missing an accessible name",
"iframe:nth-of-type(1)",
)
.with_rule_id("frame-title"),
);
results.add_violation(
Violation::new(
"2.4.1",
"Bypass Blocks",
WcagLevel::A,
Severity::High,
"No bypass mechanism found",
"document",
)
.with_rule_id("bypass"),
);
let report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
results,
100,
);
let norm = normalize(&report);
let frame = norm
.normalized
.findings
.iter()
.find(|f| f.rule_id == "a11y.frame_title.missing")
.expect("frame-title finding should keep its own taxonomy rule");
assert_eq!(frame.wcag_criterion, "2.4.1");
assert_eq!(frame.axe_id.as_deref(), Some("frame-title"));
assert!(norm
.normalized
.findings
.iter()
.any(|f| f.rule_id == "a11y.bypass_blocks.missing"));
}
#[test]
fn test_dom_parity_rules_keep_specific_taxonomy() {
let mut results = WcagResults::new();
results.add_violation(
Violation::new(
"3.2.2",
"On Input",
WcagLevel::A,
Severity::Medium,
"Form has input controls but no explicit submit button",
"form",
)
.with_rule_id("form-no-submit"),
);
results.add_violation(
Violation::new(
"1.3.1",
"Info and Relationships",
WcagLevel::A,
Severity::Medium,
"Presentational container contains semantic child",
"div[role=\"presentation\"]",
)
.with_rule_id("presentation-semantic-children"),
);
results.add_violation(
Violation::new(
"1.3.1",
"Landmark Main Present",
WcagLevel::A,
Severity::High,
"Page has no main landmark",
"document",
)
.with_rule_id("landmark-main-present"),
);
results.add_violation(
Violation::new(
"1.3.1",
"Landmark Unique",
WcagLevel::A,
Severity::Medium,
"Multiple navigation landmarks share the same accessible name",
"nav",
)
.with_rule_id("landmark-unique"),
);
let report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
results,
100,
);
let norm = normalize(&report);
let form = norm
.normalized
.findings
.iter()
.find(|f| f.rule_id == "a11y.form_no_submit.missing")
.expect("form-no-submit finding should keep its own taxonomy rule");
assert_eq!(form.wcag_criterion, "3.2.2");
assert_eq!(form.axe_id.as_deref(), Some("form-no-submit"));
let presentation = norm
.normalized
.findings
.iter()
.find(|f| f.rule_id == "a11y.presentation_semantic_children.invalid")
.expect("presentation-semantic-children should keep its own taxonomy rule");
assert_eq!(presentation.wcag_criterion, "1.3.1");
assert_eq!(
presentation.axe_id.as_deref(),
Some("presentation-semantic-children")
);
assert!(norm
.normalized
.findings
.iter()
.any(|f| f.rule_id == "a11y.landmark_main.missing"
&& f.axe_id.as_deref() == Some("landmark-main-present")));
assert!(norm
.normalized
.findings
.iter()
.any(|f| f.rule_id == "a11y.landmark_unique.invalid"
&& f.axe_id.as_deref() == Some("landmark-unique")));
}
#[test]
fn test_normalize_severity_counts() {
let mut results = WcagResults::new();
results.add_violation(Violation::new(
"1.1.1",
"Alt",
WcagLevel::A,
Severity::High,
"Err",
"n1",
));
results.add_violation(Violation::new(
"1.1.1",
"Alt",
WcagLevel::A,
Severity::High,
"Err",
"n2",
));
results.add_violation(Violation::new(
"2.4.4",
"Link",
WcagLevel::A,
Severity::Medium,
"Warn",
"n3",
));
let report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
results,
100,
);
let norm = normalize(&report);
assert_eq!(norm.normalized.severity_counts.high, 1);
assert_eq!(norm.normalized.severity_counts.medium, 1);
assert_eq!(norm.normalized.severity_counts.total, 2);
assert_eq!(norm.normalized.occurrence_counts.high, 2);
assert_eq!(norm.normalized.occurrence_counts.medium, 1);
assert_eq!(norm.normalized.occurrence_counts.total, 3);
}
#[test]
fn critical_interactive_finding_raises_risk_without_wcag_counts() {
let mut report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
WcagResults::new(),
100,
);
report.interactive_findings.push(InteractiveFinding {
category: "FocusTrap".to_string(),
kind: InteractiveFindingKind::FocusTrapEscaped,
maps_to_finding: None,
severity: Severity::Critical,
journey: "modal".to_string(),
before_snapshot_label: None,
after_snapshot_label: None,
message: "Modal has no focus trap.".to_string(),
fix_suggestion: None,
values: InteractiveFindingValues::default(),
uncertainty: None,
});
let norm = normalize(&report);
assert_eq!(norm.normalized.risk.level, RiskLevel::Medium);
assert_eq!(norm.normalized.severity_counts.total, 0);
assert_eq!(norm.normalized.score, 100);
assert_eq!(norm.normalized.risk.interactive_critical_issues, 1);
}
#[test]
fn high_interactive_finding_raises_risk_without_wcag_counts() {
let mut report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
WcagResults::new(),
100,
);
report.interactive_findings.push(InteractiveFinding {
category: "SkipLink".to_string(),
kind: InteractiveFindingKind::SkipLinkFocusNotMoved,
maps_to_finding: None,
severity: Severity::High,
journey: "skip-link".to_string(),
before_snapshot_label: None,
after_snapshot_label: None,
message: "Skip link is present but does not move focus to the target.".to_string(),
fix_suggestion: None,
values: InteractiveFindingValues::default(),
uncertainty: None,
});
let norm = normalize(&report);
assert_eq!(norm.normalized.risk.level, RiskLevel::Medium);
assert_eq!(norm.normalized.risk.score, 5);
assert_eq!(norm.normalized.risk.interactive_high_issues, 1);
assert_eq!(norm.normalized.severity_counts.total, 0);
assert_eq!(norm.normalized.score, 100);
}
#[test]
fn test_normalize_lang_conflict_flag_keeps_finding() {
let mut results = WcagResults::new();
results.add_violation(Violation::new(
"3.1.1",
"Language",
WcagLevel::A,
Severity::High,
"Missing lang",
"n1",
));
let mut report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
results,
100,
);
let norm_no_seo = normalize(&report);
assert_eq!(norm_no_seo.normalized.findings.len(), 1);
assert!(norm_no_seo.normalized.audit_flags.is_empty());
report.discoverability.seo = Some(crate::seo::SeoAnalysis {
technical: crate::seo::TechnicalSeo {
has_lang: true,
..Default::default()
},
..Default::default()
});
let norm_with_seo = normalize(&report);
assert_eq!(norm_with_seo.normalized.findings.len(), 1);
assert_eq!(
norm_with_seo.normalized.score,
report.accessibility.score.round() as u32
);
assert_eq!(norm_with_seo.normalized.audit_flags.len(), 1);
assert_eq!(
norm_with_seo.normalized.audit_flags[0]
.related_rule
.as_deref(),
Some("3.1.1")
);
}
#[test]
fn test_score_consistency() {
let mut results = WcagResults::new();
results.add_violation(Violation::new(
"1.4.3",
"Contrast",
WcagLevel::AA,
Severity::Low,
"Low contrast",
"n1",
));
let report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
results,
100,
);
let norm = normalize(&report);
let expected_grade =
AccessibilityScorer::calculate_grade(norm.normalized.overall_score as f32);
let expected_cert =
AccessibilityScorer::calculate_certificate(norm.normalized.overall_score as f32);
assert_eq!(norm.normalized.grade, expected_grade);
assert_eq!(norm.normalized.certificate, expected_cert);
}
#[test]
fn high_risk_vetoes_positive_certificate() {
assert_eq!(
gate_certificate_by_risk("STABIL".to_string(), &RiskLevel::High, 0, 0),
"EINGESCHRÄNKT"
);
assert_eq!(
gate_certificate_by_risk("SEHR GUT".to_string(), &RiskLevel::Critical, 0, 0),
"NICHT BESTANDEN"
);
assert_eq!(
gate_certificate_by_risk("AUSBAUFÄHIG".to_string(), &RiskLevel::High, 0, 0),
"EINGESCHRÄNKT"
);
assert_eq!(
gate_certificate_by_risk("UNGENÜGEND".to_string(), &RiskLevel::High, 0, 0),
"UNGENÜGEND"
);
}
#[test]
fn legal_or_blocking_vetoes_positive_certificate_at_medium_risk() {
assert_eq!(
gate_certificate_by_risk("GUT".to_string(), &RiskLevel::Medium, 1, 0),
"EINGESCHRÄNKT"
);
assert_eq!(
gate_certificate_by_risk("STABIL".to_string(), &RiskLevel::Medium, 0, 3),
"EINGESCHRÄNKT"
);
assert_eq!(
gate_certificate_by_risk("GUT".to_string(), &RiskLevel::Low, 0, 0),
"GUT"
);
assert_eq!(
gate_certificate_by_risk("GUT".to_string(), &RiskLevel::Medium, 0, 0),
"GUT"
);
}
#[test]
fn test_normalize_with_best_practices_produces_module_entry() {
use crate::best_practices::{
BestPracticesAnalysis, ConsoleErrorsAnalysis, VulnerableLibrariesAnalysis,
};
let mut report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
WcagResults::new(),
100,
);
report.best_practices = Some(BestPracticesAnalysis {
console_errors: ConsoleErrorsAnalysis {
errors: vec![],
warnings: vec![],
error_count: 0,
warning_count: 0,
},
vulnerable_libraries: VulnerableLibrariesAnalysis {
detected: vec![],
vulnerable: vec![],
has_vulnerabilities: false,
duplicate_libraries: vec![],
},
score: 90,
});
let norm = normalize(&report);
let bp_entry = norm
.normalized
.module_scores
.iter()
.find(|m| m.name == "Best Practices");
assert!(
bp_entry.is_some(),
"Best Practices module score must be present"
);
let entry = bp_entry.unwrap();
assert_eq!(entry.score, 90);
assert!(!entry.contributes_to_overall);
assert!(
norm.raw_best_practices.is_some(),
"raw_best_practices must be passed through"
);
}
#[test]
fn test_vulnerable_libraries_reduce_security_score() {
use crate::best_practices::{
BestPracticesAnalysis, ConsoleErrorsAnalysis, VulnerableLibrariesAnalysis,
VulnerableLibrary,
};
use crate::security::{SecurityAnalysis, SecurityHeaders, SslInfo};
let mut report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
WcagResults::new(),
100,
);
report.security = Some(SecurityAnalysis {
score: 80,
grade: "B".to_string(),
headers: SecurityHeaders::default(),
ssl: SslInfo::default(),
issues: vec![],
protection: Default::default(),
sourcemap_leaks: Default::default(),
recommendations: vec![],
});
report.best_practices = Some(BestPracticesAnalysis {
console_errors: ConsoleErrorsAnalysis {
errors: vec![],
warnings: vec![],
error_count: 0,
warning_count: 0,
},
vulnerable_libraries: VulnerableLibrariesAnalysis {
detected: vec![],
vulnerable: vec![
VulnerableLibrary {
name: "jQuery".to_string(),
version: "1.11.3".to_string(),
severity: "high".to_string(),
description: "XSS".to_string(),
safe_version: "3.5.0+".to_string(),
},
VulnerableLibrary {
name: "Lodash".to_string(),
version: "4.17.20".to_string(),
severity: "medium".to_string(),
description: "Prototype pollution".to_string(),
safe_version: "4.17.21+".to_string(),
},
],
has_vulnerabilities: true,
duplicate_libraries: vec![],
},
score: 60,
});
let norm = normalize(&report);
let sec_entry = norm
.normalized
.module_scores
.iter()
.find(|m| m.name == "Security");
assert!(sec_entry.is_some());
assert_eq!(sec_entry.unwrap().score, 57);
}
#[test]
fn test_performance_results_new_fields_serialize() {
use crate::performance::{PerformanceGrade, PerformanceScore, WebVitals};
let perf = PerformanceResults {
vitals: WebVitals::default(),
score: PerformanceScore {
overall: 80,
grade: PerformanceGrade::Gold,
lcp_score: None,
fcp_score: None,
cls_score: None,
interactivity_score: None,
si_score: None,
metrics_available: 0,
size_penalty: None,
js_penalty: None,
request_penalty: None,
dom_penalty: None,
is_capped: None,
},
render_blocking: None,
content_weight: None,
third_party: None,
critical_chain: None,
minification: None,
animations: None,
coverage: None,
measurement_warnings: vec![],
};
let json = serde_json::to_string(&perf).expect("PerformanceResults must serialize");
assert!(!json.contains("\"third_party\""));
assert!(!json.contains("\"minification\""));
assert!(!json.contains("\"coverage\""));
assert!(!json.contains("\"animations\""));
assert!(!json.contains("\"measurement_warnings\""));
assert!(json.contains("\"score\""));
}
#[test]
fn test_unmeasured_performance_excluded_from_overall_score() {
use crate::performance::{PerformanceGrade, PerformanceScore, WebVitals};
let report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
WcagResults::new(),
100,
)
.with_performance(PerformanceResults {
vitals: WebVitals::default(),
score: PerformanceScore {
overall: 0,
grade: PerformanceGrade::Bronze,
lcp_score: None,
fcp_score: None,
cls_score: None,
interactivity_score: None,
si_score: None,
metrics_available: 0,
size_penalty: None,
js_penalty: None,
request_penalty: None,
dom_penalty: None,
is_capped: None,
},
render_blocking: None,
content_weight: None,
third_party: None,
critical_chain: None,
minification: None,
animations: None,
coverage: None,
measurement_warnings: vec![],
});
let norm = normalize(&report).normalized;
let perf_entry = norm
.module_scores
.iter()
.find(|m| m.name == "Performance")
.expect("performance module entry must be present");
assert!(!perf_entry.contributes_to_overall);
assert_eq!(perf_entry.measurement_type, "not_measured");
assert_eq!(norm.overall_score, norm.score);
}
#[test]
fn test_viewport_weighted_core_modules_contribute() {
use crate::audit::{ViewportScoreSet, ViewportScores};
use crate::{WcagLevel, WcagResults};
let mut report = AuditReport::new(
"https://example.com".to_string(),
WcagLevel::AA,
WcagResults::new(),
100,
);
report.viewport_scores = Some(ViewportScores {
desktop: ViewportScoreSet {
accessibility: 80,
performance: None,
overall: 80,
},
mobile: ViewportScoreSet {
accessibility: 20,
performance: None,
overall: 20,
},
weighted_overall: 38,
});
let norm = normalize(&report);
assert_eq!(norm.normalized.score, 38);
assert_eq!(
norm.normalized
.module_scores
.iter()
.find(|module| module.name == "Accessibility")
.map(|module| module.score),
Some(38)
);
assert_eq!(
norm.normalized.score_calculation_method,
"viewport_weighted"
);
assert!(norm.normalized.score_breakdown.as_ref().is_some_and(|b| {
b.calculation_note.contains("70% mobile") && b.viewport_blended_accessibility == 38
}));
let names_contributing: Vec<&str> = norm
.normalized
.module_scores
.iter()
.filter(|m| m.contributes_to_overall)
.map(|m| m.name.as_str())
.collect();
assert!(
names_contributing.contains(&"Accessibility"),
"Accessibility must contribute in viewport_weighted mode, got: {:?}",
names_contributing
);
}
}