tapps-agents 3.6.0__py3-none-any.whl → 3.6.1__py3-none-any.whl

tapps_agents/agents/tester/accessibility_auditor.py

@@ -1,309 +1,309 @@
-"""
-Accessibility Auditor using Playwright snapshots.
-
-Analyzes accessibility snapshots for WCAG 2.2 compliance and generates
-accessibility test assertions.
-"""
-
-from __future__ import annotations
-
-import logging
-import re
-from dataclasses import dataclass
-from typing import Any
-
-logger = logging.getLogger(__name__)
-
-
-@dataclass
-class AccessibilityIssue:
-    """Accessibility issue found in snapshot."""
-
-    severity: str  # "error", "warning", "info"
-    rule: str  # WCAG rule identifier
-    element: str  # Element description
-    message: str  # Issue description
-    suggestion: str  # How to fix
-
-
-@dataclass
-class AccessibilityAudit:
-    """Accessibility audit results."""
-
-    score: float  # 0-100 accessibility score
-    level: str  # "A", "AA", "AAA" compliance level
-    issues: list[AccessibilityIssue]
-    passed_checks: int
-    total_checks: int
-    summary: str
-
-
-class AccessibilityAuditor:
-    """
-    Audit accessibility using Playwright snapshots.
-
-    Analyzes accessibility snapshots for WCAG 2.2 compliance.
-    """
-
-    # WCAG 2.2 Level A requirements
-    WCAG_A_CHECKS = [
-        ("alt_text", "Images must have alt text", r'img.*alt=["\']([^"\']*)["\']'),
-        ("heading_structure", "Headings must be properly structured", r'<h([1-6])'),
-        ("form_labels", "Form inputs must have labels", r'<input.*(?:id|aria-labelledby)'),
-        ("link_text", "Links must have descriptive text", r'<a[^>]*>([^<]+)</a>'),
-        ("button_text", "Buttons must have accessible text", r'<button[^>]*>([^<]+)</button>'),
-    ]
-
-    # WCAG 2.2 Level AA requirements
-    WCAG_AA_CHECKS = [
-        ("color_contrast", "Text must meet color contrast requirements", None),
-        ("keyboard_navigation", "All interactive elements must be keyboard accessible", None),
-        ("focus_indicators", "Focus indicators must be visible", None),
-    ]
-
-    def audit(self, snapshot_content: str) -> AccessibilityAudit:
-        """
-        Analyze accessibility snapshot.
-
-        Args:
-            snapshot_content: Accessibility snapshot markdown content
-
-        Returns:
-            AccessibilityAudit with score and issues
-        """
-        issues: list[AccessibilityIssue] = []
-        passed_checks = 0
-        total_checks = 0
-
-        # Extract HTML content from snapshot
-        html_content = self._extract_html(snapshot_content)
-
-        # Run WCAG Level A checks
-        for check_id, description, pattern in self.WCAG_A_CHECKS:
-            total_checks += 1
-            issue = self._check_wcag_a(html_content, check_id, description, pattern)
-            if issue:
-                issues.append(issue)
-            else:
-                passed_checks += 1
-
-        # Run WCAG Level AA checks (basic checks)
-        for check_id, description, _ in self.WCAG_AA_CHECKS:
-            total_checks += 1
-            issue = self._check_wcag_aa(html_content, check_id, description)
-            if issue:
-                issues.append(issue)
-            else:
-                passed_checks += 1
-
-        # Calculate score
-        score = (passed_checks / total_checks * 100) if total_checks > 0 else 0.0
-
-        # Determine compliance level
-        level = self._determine_level(score, issues)
-
-        # Generate summary
-        summary = self._generate_summary(score, level, len(issues), total_checks)
-
-        return AccessibilityAudit(
-            score=score,
-            level=level,
-            issues=issues,
-            passed_checks=passed_checks,
-            total_checks=total_checks,
-            summary=summary,
-        )
-
-    def _extract_html(self, snapshot_content: str) -> str:
-        """Extract HTML content from snapshot markdown."""
-        # Snapshot may be markdown with HTML code blocks
-        html_match = re.search(r"```html\n(.*?)\n```", snapshot_content, re.DOTALL)
-        if html_match:
-            return html_match.group(1)
-
-        # Or it might be direct HTML
-        if "<html" in snapshot_content or "<div" in snapshot_content:
-            return snapshot_content
-
-        return snapshot_content
-
-    def _check_wcag_a(
-        self, html: str, check_id: str, description: str, pattern: str | None
-    ) -> AccessibilityIssue | None:
-        """Check WCAG Level A requirement."""
-        if check_id == "alt_text":
-            # Check for images without alt text
-            img_pattern = r'<img[^>]*>'
-            images = re.findall(img_pattern, html, re.IGNORECASE)
-            for img in images:
-                if 'alt=' not in img.lower() and 'aria-label=' not in img.lower():
-                    return AccessibilityIssue(
-                        severity="error",
-                        rule="WCAG 2.2.1.1 (Level A)",
-                        element=img[:50],
-                        message="Image missing alt text",
-                        suggestion="Add alt attribute to img tag: <img alt='description'>",
-                    )
-
-        elif check_id == "heading_structure":
-            # Check for proper heading hierarchy
-            headings = re.findall(r'<h([1-6])', html, re.IGNORECASE)
-            if headings:
-                levels = [int(h) for h in headings]
-                # Check for skipped levels (e.g., h1 -> h3)
-                for i in range(len(levels) - 1):
-                    if levels[i + 1] > levels[i] + 1:
-                        return AccessibilityIssue(
-                            severity="warning",
-                            rule="WCAG 2.4.6 (Level A)",
-                            element=f"h{levels[i]} -> h{levels[i+1]}",
-                            message="Heading levels should not be skipped",
-                            suggestion="Use sequential heading levels (h1, h2, h3, etc.)",
-                        )
-
-        elif check_id == "form_labels":
-            # Check for form inputs without labels
-            inputs = re.findall(r'<input[^>]*>', html, re.IGNORECASE)
-            for inp in inputs:
-                if 'id=' in inp or 'aria-labelledby=' in inp:
-                    continue
-                if 'aria-label=' in inp:
-                    continue
-                # Check if there's a label element nearby (simplified check)
-                return AccessibilityIssue(
-                    severity="error",
-                    rule="WCAG 2.4.6 (Level A)",
-                    element=inp[:50],
-                    message="Form input missing label",
-                    suggestion="Add <label> element or aria-label attribute",
-                )
-
-        elif check_id == "link_text":
-            # Check for links without descriptive text
-            links = re.findall(r'<a[^>]*>([^<]+)</a>', html, re.IGNORECASE)
-            for link_text in links:
-                link_text = link_text.strip()
-                if not link_text or len(link_text) < 3:
-                    return AccessibilityIssue(
-                        severity="error",
-                        rule="WCAG 2.4.4 (Level A)",
-                        element=f"Link: {link_text}",
-                        message="Link text is not descriptive",
-                        suggestion="Use descriptive link text (e.g., 'Read more about accessibility' not 'click here')",
-                    )
-
-        elif check_id == "button_text":
-            # Check for buttons without text
-            buttons = re.findall(r'<button[^>]*>([^<]*)</button>', html, re.IGNORECASE)
-            for button_text in buttons:
-                button_text = button_text.strip()
-                if not button_text:
-                    return AccessibilityIssue(
-                        severity="error",
-                        rule="WCAG 2.4.4 (Level A)",
-                        element="Button",
-                        message="Button missing accessible text",
-                        suggestion="Add text content or aria-label to button",
-                    )
-
-        return None
-
-    def _check_wcag_aa(self, html: str, check_id: str, description: str) -> AccessibilityIssue | None:
-        """Check WCAG Level AA requirement (basic checks)."""
-        # These are harder to check from HTML alone, so we do basic checks
-        if check_id == "color_contrast":
-            # Can't check color contrast from HTML alone
-            # This would require actual rendering
-            return None  # Pass for now (would need visual analysis)
-
-        elif check_id == "keyboard_navigation":
-            # Check for interactive elements that might not be keyboard accessible
-            # This is a simplified check
-            interactive_elements = re.findall(
-                r'<(button|a|input|select|textarea)[^>]*>', html, re.IGNORECASE
-            )
-            if not interactive_elements:
-                return None
-
-            # Check for tabindex="-1" which makes elements not keyboard accessible
-            if 'tabindex="-1"' in html:
-                return AccessibilityIssue(
-                    severity="warning",
-                    rule="WCAG 2.1.1 (Level A)",
-                    element="Interactive element",
-                    message="Some elements may not be keyboard accessible",
-                    suggestion="Ensure all interactive elements are keyboard accessible",
-                )
-
-        elif check_id == "focus_indicators":
-            # Check for CSS that might hide focus indicators
-            # This is a simplified check
-            if 'outline: none' in html or 'outline: 0' in html:
-                return AccessibilityIssue(
-                    severity="warning",
-                    rule="WCAG 2.4.7 (Level AA)",
-                    element="CSS",
-                    message="Focus indicators may be hidden",
-                    suggestion="Ensure focus indicators are visible (don't use outline: none without alternative)",
-                )
-
-        return None
-
-    def _determine_level(self, score: float, issues: list[AccessibilityIssue]) -> str:
-        """Determine WCAG compliance level."""
-        error_count = sum(1 for issue in issues if issue.severity == "error")
-        warning_count = sum(1 for issue in issues if issue.severity == "warning")
-
-        if error_count == 0 and warning_count == 0 and score >= 95:
-            return "AAA"
-        elif error_count == 0 and score >= 90:
-            return "AA"
-        elif error_count < 3 and score >= 80:
-            return "A"
-        else:
-            return "Not Compliant"
-
-    def _generate_summary(
-        self, score: float, level: str, issue_count: int, total_checks: int
-    ) -> str:
-        """Generate audit summary."""
-        return (
-            f"Accessibility Score: {score:.1f}/100\n"
-            f"WCAG Compliance: Level {level}\n"
-            f"Issues Found: {issue_count} ({total_checks - issue_count} passed)\n"
-            f"Total Checks: {total_checks}"
-        )
-
-    def generate_test_assertions(self, audit: AccessibilityAudit) -> list[str]:
-        """
-        Generate test assertions from audit results.
-
-        Args:
-            audit: AccessibilityAudit results
-
-        Returns:
-            List of test assertion code strings
-        """
-        assertions = []
-
-        # Overall score assertion
-        assertions.append(
-            f"assert accessibility_score >= {audit.score:.1f}, "
-            f"'Accessibility score {audit.score:.1f} is below threshold'"
-        )
-
-        # Compliance level assertion
-        assertions.append(
-            f"assert accessibility_level == '{audit.level}', "
-            f"'Expected WCAG {audit.level} compliance, got {audit.level}'"
-        )
-
-        # Issue-specific assertions
-        for issue in audit.issues:
-            if issue.severity == "error":
-                assertions.append(
-                    f"assert not has_accessibility_issue('{issue.rule}'), "
-                    f"'{issue.message}'"
-                )
-
-        return assertions
+"""
+Accessibility Auditor using Playwright snapshots.
+
+Analyzes accessibility snapshots for WCAG 2.2 compliance and generates
+accessibility test assertions.
+"""
+
+from __future__ import annotations
+
+import logging
+import re
+from dataclasses import dataclass
+from typing import Any
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class AccessibilityIssue:
+    """Accessibility issue found in snapshot."""
+
+    severity: str  # "error", "warning", "info"
+    rule: str  # WCAG rule identifier
+    element: str  # Element description
+    message: str  # Issue description
+    suggestion: str  # How to fix
+
+
+@dataclass
+class AccessibilityAudit:
+    """Accessibility audit results."""
+
+    score: float  # 0-100 accessibility score
+    level: str  # "A", "AA", "AAA" compliance level
+    issues: list[AccessibilityIssue]
+    passed_checks: int
+    total_checks: int
+    summary: str
+
+
+class AccessibilityAuditor:
+    """
+    Audit accessibility using Playwright snapshots.
+
+    Analyzes accessibility snapshots for WCAG 2.2 compliance.
+    """
+
+    # WCAG 2.2 Level A requirements
+    WCAG_A_CHECKS = [
+        ("alt_text", "Images must have alt text", r'img.*alt=["\']([^"\']*)["\']'),
+        ("heading_structure", "Headings must be properly structured", r'<h([1-6])'),
+        ("form_labels", "Form inputs must have labels", r'<input.*(?:id|aria-labelledby)'),
+        ("link_text", "Links must have descriptive text", r'<a[^>]*>([^<]+)</a>'),
+        ("button_text", "Buttons must have accessible text", r'<button[^>]*>([^<]+)</button>'),
+    ]
+
+    # WCAG 2.2 Level AA requirements
+    WCAG_AA_CHECKS = [
+        ("color_contrast", "Text must meet color contrast requirements", None),
+        ("keyboard_navigation", "All interactive elements must be keyboard accessible", None),
+        ("focus_indicators", "Focus indicators must be visible", None),
+    ]
+
+    def audit(self, snapshot_content: str) -> AccessibilityAudit:
+        """
+        Analyze accessibility snapshot.
+
+        Args:
+            snapshot_content: Accessibility snapshot markdown content
+
+        Returns:
+            AccessibilityAudit with score and issues
+        """
+        issues: list[AccessibilityIssue] = []
+        passed_checks = 0
+        total_checks = 0
+
+        # Extract HTML content from snapshot
+        html_content = self._extract_html(snapshot_content)
+
+        # Run WCAG Level A checks
+        for check_id, description, pattern in self.WCAG_A_CHECKS:
+            total_checks += 1
+            issue = self._check_wcag_a(html_content, check_id, description, pattern)
+            if issue:
+                issues.append(issue)
+            else:
+                passed_checks += 1
+
+        # Run WCAG Level AA checks (basic checks)
+        for check_id, description, _ in self.WCAG_AA_CHECKS:
+            total_checks += 1
+            issue = self._check_wcag_aa(html_content, check_id, description)
+            if issue:
+                issues.append(issue)
+            else:
+                passed_checks += 1
+
+        # Calculate score
+        score = (passed_checks / total_checks * 100) if total_checks > 0 else 0.0
+
+        # Determine compliance level
+        level = self._determine_level(score, issues)
+
+        # Generate summary
+        summary = self._generate_summary(score, level, len(issues), total_checks)
+
+        return AccessibilityAudit(
+            score=score,
+            level=level,
+            issues=issues,
+            passed_checks=passed_checks,
+            total_checks=total_checks,
+            summary=summary,
+        )
+
+    def _extract_html(self, snapshot_content: str) -> str:
+        """Extract HTML content from snapshot markdown."""
+        # Snapshot may be markdown with HTML code blocks
+        html_match = re.search(r"```html\n(.*?)\n```", snapshot_content, re.DOTALL)
+        if html_match:
+            return html_match.group(1)
+
+        # Or it might be direct HTML
+        if "<html" in snapshot_content or "<div" in snapshot_content:
+            return snapshot_content
+
+        return snapshot_content
+
+    def _check_wcag_a(
+        self, html: str, check_id: str, description: str, pattern: str | None
+    ) -> AccessibilityIssue | None:
+        """Check WCAG Level A requirement."""
+        if check_id == "alt_text":
+            # Check for images without alt text
+            img_pattern = r'<img[^>]*>'
+            images = re.findall(img_pattern, html, re.IGNORECASE)
+            for img in images:
+                if 'alt=' not in img.lower() and 'aria-label=' not in img.lower():
+                    return AccessibilityIssue(
+                        severity="error",
+                        rule="WCAG 2.2.1.1 (Level A)",
+                        element=img[:50],
+                        message="Image missing alt text",
+                        suggestion="Add alt attribute to img tag: <img alt='description'>",
+                    )
+
+        elif check_id == "heading_structure":
+            # Check for proper heading hierarchy
+            headings = re.findall(r'<h([1-6])', html, re.IGNORECASE)
+            if headings:
+                levels = [int(h) for h in headings]
+                # Check for skipped levels (e.g., h1 -> h3)
+                for i in range(len(levels) - 1):
+                    if levels[i + 1] > levels[i] + 1:
+                        return AccessibilityIssue(
+                            severity="warning",
+                            rule="WCAG 2.4.6 (Level A)",
+                            element=f"h{levels[i]} -> h{levels[i+1]}",
+                            message="Heading levels should not be skipped",
+                            suggestion="Use sequential heading levels (h1, h2, h3, etc.)",
+                        )
+
+        elif check_id == "form_labels":
+            # Check for form inputs without labels
+            inputs = re.findall(r'<input[^>]*>', html, re.IGNORECASE)
+            for inp in inputs:
+                if 'id=' in inp or 'aria-labelledby=' in inp:
+                    continue
+                if 'aria-label=' in inp:
+                    continue
+                # Check if there's a label element nearby (simplified check)
+                return AccessibilityIssue(
+                    severity="error",
+                    rule="WCAG 2.4.6 (Level A)",
+                    element=inp[:50],
+                    message="Form input missing label",
+                    suggestion="Add <label> element or aria-label attribute",
+                )
+
+        elif check_id == "link_text":
+            # Check for links without descriptive text
+            links = re.findall(r'<a[^>]*>([^<]+)</a>', html, re.IGNORECASE)
+            for link_text in links:
+                link_text = link_text.strip()
+                if not link_text or len(link_text) < 3:
+                    return AccessibilityIssue(
+                        severity="error",
+                        rule="WCAG 2.4.4 (Level A)",
+                        element=f"Link: {link_text}",
+                        message="Link text is not descriptive",
+                        suggestion="Use descriptive link text (e.g., 'Read more about accessibility' not 'click here')",
+                    )
+
+        elif check_id == "button_text":
+            # Check for buttons without text
+            buttons = re.findall(r'<button[^>]*>([^<]*)</button>', html, re.IGNORECASE)
+            for button_text in buttons:
+                button_text = button_text.strip()
+                if not button_text:
+                    return AccessibilityIssue(
+                        severity="error",
+                        rule="WCAG 2.4.4 (Level A)",
+                        element="Button",
+                        message="Button missing accessible text",
+                        suggestion="Add text content or aria-label to button",
+                    )
+
+        return None
+
+    def _check_wcag_aa(self, html: str, check_id: str, description: str) -> AccessibilityIssue | None:
+        """Check WCAG Level AA requirement (basic checks)."""
+        # These are harder to check from HTML alone, so we do basic checks
+        if check_id == "color_contrast":
+            # Can't check color contrast from HTML alone
+            # This would require actual rendering
+            return None  # Pass for now (would need visual analysis)
+
+        elif check_id == "keyboard_navigation":
+            # Check for interactive elements that might not be keyboard accessible
+            # This is a simplified check
+            interactive_elements = re.findall(
+                r'<(button|a|input|select|textarea)[^>]*>', html, re.IGNORECASE
+            )
+            if not interactive_elements:
+                return None
+
+            # Check for tabindex="-1" which makes elements not keyboard accessible
+            if 'tabindex="-1"' in html:
+                return AccessibilityIssue(
+                    severity="warning",
+                    rule="WCAG 2.1.1 (Level A)",
+                    element="Interactive element",
+                    message="Some elements may not be keyboard accessible",
+                    suggestion="Ensure all interactive elements are keyboard accessible",
+                )
+
+        elif check_id == "focus_indicators":
+            # Check for CSS that might hide focus indicators
+            # This is a simplified check
+            if 'outline: none' in html or 'outline: 0' in html:
+                return AccessibilityIssue(
+                    severity="warning",
+                    rule="WCAG 2.4.7 (Level AA)",
+                    element="CSS",
+                    message="Focus indicators may be hidden",
+                    suggestion="Ensure focus indicators are visible (don't use outline: none without alternative)",
+                )
+
+        return None
+
+    def _determine_level(self, score: float, issues: list[AccessibilityIssue]) -> str:
+        """Determine WCAG compliance level."""
+        error_count = sum(1 for issue in issues if issue.severity == "error")
+        warning_count = sum(1 for issue in issues if issue.severity == "warning")
+
+        if error_count == 0 and warning_count == 0 and score >= 95:
+            return "AAA"
+        elif error_count == 0 and score >= 90:
+            return "AA"
+        elif error_count < 3 and score >= 80:
+            return "A"
+        else:
+            return "Not Compliant"
+
+    def _generate_summary(
+        self, score: float, level: str, issue_count: int, total_checks: int
+    ) -> str:
+        """Generate audit summary."""
+        return (
+            f"Accessibility Score: {score:.1f}/100\n"
+            f"WCAG Compliance: Level {level}\n"
+            f"Issues Found: {issue_count} ({total_checks - issue_count} passed)\n"
+            f"Total Checks: {total_checks}"
+        )
+
+    def generate_test_assertions(self, audit: AccessibilityAudit) -> list[str]:
+        """
+        Generate test assertions from audit results.
+
+        Args:
+            audit: AccessibilityAudit results
+
+        Returns:
+            List of test assertion code strings
+        """
+        assertions = []
+
+        # Overall score assertion
+        assertions.append(
+            f"assert accessibility_score >= {audit.score:.1f}, "
+            f"'Accessibility score {audit.score:.1f} is below threshold'"
+        )
+
+        # Compliance level assertion
+        assertions.append(
+            f"assert accessibility_level == '{audit.level}', "
+            f"'Expected WCAG {audit.level} compliance, got {audit.level}'"
+        )
+
+        # Issue-specific assertions
+        for issue in audit.issues:
+            if issue.severity == "error":
+                assertions.append(
+                    f"assert not has_accessibility_issue('{issue.rule}'), "
+                    f"'{issue.message}'"
+                )
+
+        return assertions