solace-agent-mesh 0.2.4__py3-none-any.whl → 1.0.1__py3-none-any.whl

solace_agent_mesh/evaluation/report_data_processor.py

@@ -0,0 +1,972 @@
+"""
+Refactored report data processor with improved structure and readability.
+This module extracts and processes evaluation data for HTML report generation.
+"""
+
+import json
+import logging
+from datetime import datetime
+from dataclasses import dataclass, field
+from typing import Dict, List, Optional, Any, Set, Tuple
+from pathlib import Path
+from collections import defaultdict, Counter
+import random
+
+# Import test case loader
+from .test_case_loader import load_test_case
+
+# Set up logging
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class EvaluationMetrics:
+    """Core evaluation data structure."""
+
+    models: List[str] = field(default_factory=list)
+    total_execution_time: Optional[float] = None
+    total_execution_time_formatted: str = "Not available"
+    generation_time: str = field(
+        default_factory=lambda: datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+    )
+    timestamp: str = field(
+        default_factory=lambda: datetime.now().strftime("%B %d, %Y at %I:%M %p")
+    )
+    runs: str = "Not available"
+    total_tests: int = 0
+    duration: str = "Not available"
+    test_case_names: List[str] = field(default_factory=list)
+
+
+@dataclass
+class ModelPerformance:
+    """Individual model performance data."""
+
+    model_name: str
+    average_score: float = 0.0
+    success_rate: float = 0.0
+    test_count: int = 0
+    estimated_cost: float = 0.0
+    scores: List[float] = field(default_factory=list)
+
+
+@dataclass
+class TestCaseResult:
+    """Test case specific results."""
+
+    test_case_id: str
+    category: str
+    description: str = ""
+    model_results: Dict[str, Any] = field(default_factory=dict)
+    average_score: float = 0.0
+
+
+@dataclass
+class ChartConfiguration:
+    """Chart and visualization data."""
+
+    categories: List[str] = field(default_factory=list)
+    datasets: List[Dict[str, Any]] = field(default_factory=list)
+    category_scores: Dict[str, Dict[str, float]] = field(default_factory=dict)
+
+
+@dataclass
+class CategoryStatistics:
+    """Category-based statistics."""
+
+    category_name: str
+    test_cases: List[str] = field(default_factory=list)
+    model_scores: Dict[str, float] = field(default_factory=dict)
+
+
+class FileService:
+    """Handles file I/O operations with proper error handling."""
+
+    @staticmethod
+    def load_json(filepath: Path) -> Any:
+        """Load JSON data from file."""
+        try:
+            with open(filepath, "r") as f:
+                return json.load(f)
+        except FileNotFoundError:
+            logger.warning(f"File not found: {filepath}")
+            return None
+        except json.JSONDecodeError as e:
+            logger.error(f"Invalid JSON in file {filepath}: {e}")
+            return None
+        except Exception as e:
+            logger.error(f"Error reading file {filepath}: {e}")
+            return None
+
+    @staticmethod
+    def file_exists(filepath: Path) -> bool:
+        """Check if file exists."""
+        return filepath.exists() and filepath.is_file()
+
+    @staticmethod
+    def list_directories(path: Path) -> List[str]:
+        """List directories in the given path."""
+        try:
+            return [
+                item.name
+                for item in path.iterdir()
+                if item.is_dir() and not item.name.startswith(".")
+            ]
+        except Exception as e:
+            logger.error(f"Error listing directories in {path}: {e}")
+            return []
+
+
+class ResultsExtractionService:
+    """Extracts raw data from results directories."""
+
+    def __init__(self, file_service: FileService):
+        self.file_service = file_service
+
+    def extract_model_results(self, results_dir: Path) -> Dict[str, Any]:
+        """Extract results for all models."""
+        model_results = {}
+
+        for model_name in self.file_service.list_directories(results_dir):
+            model_path = results_dir / model_name
+            results_file = model_path / "results.json"
+
+            if self.file_service.file_exists(results_file):
+                results_data = self.file_service.load_json(results_file)
+                if results_data:
+                    model_results[model_name] = results_data
+                    logger.debug(f"Loaded results for model: {model_name}")
+
+        logger.info(f"Extracted results for {len(model_results)} models")
+        return model_results
+
+    def extract_execution_stats(self, results_dir: Path) -> Optional[Dict[str, Any]]:
+        """Extract execution statistics."""
+        stats_file = results_dir / "stats.json"
+
+        if self.file_service.file_exists(stats_file):
+            stats_data = self.file_service.load_json(stats_file)
+            if stats_data:
+                logger.debug("Loaded execution statistics")
+                return stats_data
+
+        logger.warning("No execution statistics found")
+        return None
+
+
+class MetricsCalculationService:
+    """Calculates performance statistics and aggregations."""
+
+    @staticmethod
+    def calculate_model_performance(
+        model_name: str, results_data: Dict[str, Any]
+    ) -> ModelPerformance:
+        """Calculate performance metrics for a single model."""
+        performance = ModelPerformance(model_name=model_name)
+
+        scores = []
+        test_count = 0
+
+        # Handle new data structure with test_cases array
+        if "test_cases" in results_data:
+            for test_case in results_data["test_cases"]:
+                if "runs" in test_case:
+                    for run_data in test_case["runs"]:
+                        if isinstance(run_data, dict):
+                            # Use llm_eval score if available, otherwise use response_match
+                            score = run_data.get("llm_eval", {}).get("score")
+                            if score is None:
+                                score = run_data.get("response_match", 0)
+
+                            # Ensure score is valid
+                            if score is not None and isinstance(score, (int, float)):
+                                scores.append(score)
+
+                            test_count += 1
+
+        # Calculate metrics
+        if scores:
+            performance.scores = scores
+            performance.average_score = sum(scores) / len(scores)
+            performance.success_rate = (
+                len([s for s in scores if s >= 0.7]) / len(scores) * 100
+            )
+
+        performance.test_count = test_count
+        performance.estimated_cost = test_count * 0.02  # Rough estimate
+
+        return performance
+
+    @staticmethod
+    def format_execution_time(total_time: float) -> Tuple[str, str]:
+        """Format execution time into readable strings."""
+        minutes = int(total_time // 60)
+        seconds = int(total_time % 60)
+        formatted = f"{minutes}m {seconds}s"
+        duration = f"{minutes} minutes {seconds} seconds"
+        return formatted, duration
+
+    @staticmethod
+    def calculate_run_statistics(model_results: Dict[str, Any]) -> Tuple[int, str]:
+        """Calculate run statistics from model results."""
+        test_cases = set()
+        all_run_counts = []
+
+        for model_name, results in model_results.items():
+            if "test_cases" in results:
+                for test_case in results["test_cases"]:
+                    test_case_id = test_case.get("test_case_id")
+                    if test_case_id:
+                        test_cases.add(test_case_id)
+
+                        # Count unique runs
+                        if "runs" in test_case:
+                            unique_runs = set()
+                            for run in test_case["runs"]:
+                                if isinstance(run, dict) and "run" in run:
+                                    unique_runs.add(run["run"])
+                            if unique_runs:
+                                all_run_counts.append(len(unique_runs))
+
+        total_tests = len(test_cases)
+
+        # Determine run count description
+        if all_run_counts:
+            run_count_mode = Counter(all_run_counts).most_common(1)[0][0]
+            runs_description = f"{run_count_mode} run{'s' if run_count_mode != 1 else ''} per test case"
+        else:
+            runs_description = "Not available"
+
+        return total_tests, runs_description
+
+
+class ChartDataService:
+    """Generates chart and visualization data."""
+
+    def __init__(self, file_service: FileService):
+        self.file_service = file_service
+
+    def generate_chart_configuration(
+        self, model_results: Dict[str, Any], test_cases: Dict[str, Dict[str, Any]]
+    ) -> ChartConfiguration:
+        """Generate chart configuration data."""
+        chart_config = ChartConfiguration()
+
+        # Extract categories and organize test cases
+        category_test_mapping = self._extract_category_mapping(model_results)
+
+        # Calculate category scores for each model
+        category_scores = self._calculate_category_scores(
+            category_test_mapping, test_cases, model_results
+        )
+
+        # Prepare chart data
+        if category_scores:
+            chart_config.categories = sorted(list(category_scores.keys()))
+            chart_config.category_scores = category_scores
+            chart_config.datasets = self._generate_chart_datasets(
+                category_scores, model_results
+            )
+
+        return chart_config
+
+    def _extract_category_mapping(
+        self, model_results: Dict[str, Any]
+    ) -> Dict[str, Set[str]]:
+        """Extract category to test case mapping."""
+        category_test_mapping = defaultdict(set)
+
+        for model_name, results in model_results.items():
+            if "test_cases" in results:
+                for test_case in results["test_cases"]:
+                    test_id = test_case.get("test_case_id")
+                    category = test_case.get("category")
+                    if test_id and category:
+                        category_test_mapping[category].add(test_id)
+
+        # Convert sets to sorted lists
+        return {
+            cat: sorted(list(tests)) for cat, tests in category_test_mapping.items()
+        }
+
+    def _calculate_category_scores(
+        self,
+        category_test_mapping: Dict[str, List[str]],
+        test_cases: Dict[str, Dict[str, Any]],
+        model_results: Dict[str, Any],
+    ) -> Dict[str, Dict[str, float]]:
+        """Calculate average scores by category for each model."""
+        category_scores = {}
+
+        for category, test_names in category_test_mapping.items():
+            category_scores[category] = {}
+
+            for model_name in model_results.keys():
+                scores = []
+
+                # Collect scores for this category and model
+                for test_name in test_names:
+                    if test_name in test_cases and model_name in test_cases[test_name]:
+                        test_data = test_cases[test_name][model_name]
+                        if isinstance(test_data, dict) and "runs" in test_data:
+                            for run in test_data["runs"]:
+                                if isinstance(run, dict):
+                                    # Prioritize llm_eval score over response_match
+                                    score = run.get("llm_eval", {}).get("score")
+                                    if score is not None and isinstance(
+                                        score, (int, float)
+                                    ):
+                                        scores.append(score)
+                                    else:
+                                        score = run.get("response_match", 0)
+                                        if score is not None and isinstance(
+                                            score, (int, float)
+                                        ):
+                                            scores.append(score)
+
+                # Calculate average score for this category and model
+                category_scores[category][model_name] = (
+                    sum(scores) / len(scores) if scores else 0
+                )
+
+        return category_scores
+
+    def _generate_chart_datasets(
+        self,
+        category_scores: Dict[str, Dict[str, float]],
+        model_results: Dict[str, Any],
+    ) -> List[Dict[str, Any]]:
+        """Generate chart datasets for visualization."""
+        # Enhanced model colors with better contrast
+        model_colors = {
+            "gpt-4": "#059669",
+            "gpt-4-1": "#10b981",
+            "claude-3-sonnet": "#0ea5e9",
+            "gemini-pro": "#f59e0b",
+            "gemini-2.5-pro": "#f59e0b",
+            "gemini-flash": "#8b5cf6",
+            "gemini-2.5-flash": "#a855f7",
+            "gpt-3.5-turbo": "#ef4444",
+            "claude-3-haiku": "#84cc16",
+        }
+
+        chart_datasets = []
+        categories = sorted(list(category_scores.keys()))
+
+        for model_name in sorted(model_results.keys()):
+            model_data = []
+            for category in categories:
+                score = category_scores[category].get(model_name, 0)
+                model_data.append(round(score, 3))
+
+            color = model_colors.get(model_name)
+            if color is None:
+                # Generate a random color if not in the predefined list
+                r = lambda: random.randint(0, 255)
+                color = f"#{r():02x}{r():02x}{r():02x}"
+
+            chart_datasets.append(
+                {
+                    "label": model_name,
+                    "data": model_data,
+                    "backgroundColor": color,
+                    "borderColor": color,
+                    "borderWidth": 1,
+                    "borderRadius": 4,
+                    "borderSkipped": False,
+                }
+            )
+
+        return chart_datasets
+
+
+class ModalDataService:
+    """Generates data specifically for modal functionality."""
+
+    def __init__(self, file_service: FileService):
+        self.file_service = file_service
+
+    def generate_modal_test_data(
+        self, test_case_id: str, model_results: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """Generate test data for modal JavaScript consumption."""
+        modal_data = {"model_scores": {}, "tool_scores": {}, "individual_runs": {}}
+
+        # Extract data for this specific test case
+        for model_name, results in model_results.items():
+            if "test_cases" in results:
+                for test_case in results["test_cases"]:
+                    if test_case.get("test_case_id") == test_case_id:
+                        # Extract model scores
+                        runs = test_case.get("runs", [])
+                        if runs:
+                            response_scores = []
+                            tool_scores = []
+                            individual_runs = []
+
+                            for run_data in runs:
+                                if isinstance(run_data, dict):
+                                    # Get response score (prioritize llm_eval)
+                                    response_score = run_data.get("response_match", 0)
+                                    llm_score = run_data.get("llm_eval", {}).get(
+                                        "score"
+                                    )
+
+                                    # Get tool score
+                                    tool_score = run_data.get("tool_match", 1.0)
+
+                                    # Get other data
+                                    duration = run_data.get("duration_seconds", 0)
+                                    run_number = run_data.get("run", 1)
+                                    reasoning = run_data.get("llm_eval", {}).get(
+                                        "reasoning", "No reasoning provided"
+                                    )
+
+                                    if response_score is not None:
+                                        response_scores.append(response_score)
+                                        tool_scores.append(tool_score)
+
+                                        individual_runs.append(
+                                            {
+                                                "run_number": run_number,
+                                                "response_score": response_score,
+                                                "tool_score": tool_score,
+                                                "llm_eval": llm_score,
+                                                "llm_reasoning": reasoning,
+                                                "execution_time": duration,
+                                                "query": "",
+                                                "actual_response": "",
+                                                "expected_response": "",
+                                            }
+                                        )
+
+                            # Calculate averages
+                            if response_scores:
+                                modal_data["model_scores"][model_name] = sum(
+                                    response_scores
+                                ) / len(response_scores)
+                                modal_data["tool_scores"][model_name] = sum(
+                                    tool_scores
+                                ) / len(tool_scores)
+                                modal_data["individual_runs"][
+                                    model_name
+                                ] = individual_runs
+
+        return modal_data
+
+
+class TemplateDataService:
+    """Formats data for template consumption."""
+
+    def __init__(self, file_service: FileService):
+        self.file_service = file_service
+        self.modal_service = ModalDataService(file_service)
+
+    def generate_performance_metrics_table(
+        self, model_performances: Dict[str, ModelPerformance]
+    ) -> str:
+        """Generate HTML table rows for performance metrics."""
+        metrics_rows = []
+
+        for model_name, performance in model_performances.items():
+            if performance.scores:
+                score_class = self._get_score_class(performance.average_score)
+
+                metrics_rows.append(
+                    f"""
+                    <tr>
+                        <td class="model-name">{model_name}</td>
+                        <td class="metric-value {score_class}">{performance.average_score:.2f}</td>
+                        <td class="metric-value {score_class}">{performance.success_rate:.0f}%</td>
+                        <td class="metric-value">{performance.test_count}</td>
+                        <td class="estimated-cost">${performance.estimated_cost:.2f}</td>
+                    </tr>
+                """
+                )
+
+        return "".join(metrics_rows)
+
+    def generate_breakdown_content(
+        self,
+        test_case_results: List[TestCaseResult],
+        model_performances: Dict[str, ModelPerformance],
+        model_results: Dict[str, Any] = None,
+    ) -> str:
+        """Generate detailed breakdown content by category with modal support."""
+        # Group test cases by category
+        categories_with_tests = defaultdict(list)
+        for test_result in test_case_results:
+            categories_with_tests[test_result.category].append(test_result)
+
+        breakdown_sections = []
+
+        for category, test_results in sorted(categories_with_tests.items()):
+            category_tests = []
+
+            for test_result in test_results:
+                test_scores = []
+
+                for model_name, performance in model_performances.items():
+                    if test_result.test_case_id in test_result.model_results:
+                        model_data = test_result.model_results[
+                            test_result.test_case_id
+                        ].get(model_name, {})
+
+                        if isinstance(model_data, dict) and "runs" in model_data:
+                            scores = []
+                            durations = []
+                            success_count = 0
+
+                            for run in model_data["runs"]:
+                                if isinstance(run, dict):
+                                    # Use llm_eval score if available, otherwise use response_match
+                                    score = run.get("llm_eval", {}).get("score")
+                                    if score is not None and isinstance(
+                                        score, (int, float)
+                                    ):
+                                        scores.append(score)
+                                    else:
+                                        score = run.get("response_match", 0)
+                                        if score is not None and isinstance(
+                                            score, (int, float)
+                                        ):
+                                            scores.append(score)
+
+                                    # Track duration and success
+                                    duration = run.get("duration_seconds", 0)
+                                    if duration is not None and isinstance(
+                                        duration, (int, float)
+                                    ):
+                                        durations.append(duration)
+                                    if (
+                                        score is not None
+                                        and isinstance(score, (int, float))
+                                        and score >= 0.7
+                                    ):
+                                        success_count += 1
+
+                            if scores:
+                                avg_score = sum(scores) / len(scores)
+                                avg_duration = (
+                                    sum(durations) / len(durations) if durations else 0
+                                )
+                                score_class = self._get_score_class(avg_score)
+
+                                test_scores.append(
+                                    f"""
+                                    <div class="model-result {score_class}">
+                                        <span class="model-score">{model_name}</span>
+                                        <span class="score-value">LLM Eval: {avg_score:.3f}</span>
+                                        <span class="avg-duration">Avg time: {avg_duration:.1f}s</span>
+                                    </div>
+                                """
+                                )
+
+                if test_scores:
+                    # Generate modal data for this test case
+                    modal_data = {}
+                    if model_results:
+                        modal_data = self.modal_service.generate_modal_test_data(
+                            test_result.test_case_id, model_results
+                        )
+
+                    # Escape JSON for HTML attribute
+                    modal_data_json = json.dumps(modal_data).replace('"', "&quot;")
+
+                    category_tests.append(
+                        f"""
+                        <div class="test-item" 
+                             data-test-name="{test_result.test_case_id}"
+                             data-test-description="{test_result.description}"
+                             data-test-data="{modal_data_json}">
+                            <div class="test-header">
+                                <span class="test-name">{test_result.test_case_id}</span>
+                                <span class="test-description">{test_result.description}</span>
+                            </div>
+                            <div class="model-results">
+                                {''.join(test_scores)}
+                            </div>
+                        </div>
+                    """
+                    )
+
+            if category_tests:
+                test_count = len(category_tests)
+                breakdown_sections.append(
+                    f"""
+                    <div class="category-section">
+                        <div class="category-header">
+                            <h3 class="category-title">{category} ({test_count} test case{'s' if test_count != 1 else ''})</h3>
+                            <span class="category-toggle">▶</span>
+                        </div>
+                        <div class="category-content">
+                            {''.join(category_tests)}
+                        </div>
+                    </div>
+                """
+                )
+
+        return "".join(breakdown_sections)
+
+    def generate_model_execution_times(self, model_results: Dict[str, Any]) -> str:
+        """Generate model execution times HTML."""
+        execution_times_html = []
+
+        # Model colors for consistency
+        model_colors = {
+            "gpt-4": "#059669",
+            "gpt-4-1": "#10b981",
+            "claude-3-sonnet": "#0ea5e9",
+            "gemini-pro": "#f59e0b",
+            "gemini-2.5-pro": "#f59e0b",
+            "gemini-flash": "#8b5cf6",
+            "gpt-3.5-turbo": "#ef4444",
+            "claude-3-haiku": "#84cc16",
+        }
+
+        for model_name in sorted(model_results.keys()):
+            results = model_results[model_name]
+            execution_time = results.get("total_execution_time")
+
+            if execution_time is not None:
+                # Format time as minutes and seconds
+                minutes = int(execution_time // 60)
+                seconds = int(execution_time % 60)
+                time_formatted = f"{minutes}m {seconds}s"
+
+                color = model_colors.get(model_name, "#6b7280")
+                execution_times_html.append(
+                    f"""
+                    <div style="background: #f9fafb; padding: 15px; border-radius: 8px; border-left: 4px solid {color}; box-shadow: 0 1px 3px rgba(0,0,0,0.1);">
+                        <div style="font-weight: 600; color: #1f2937; margin-bottom: 5px; font-size: 1rem;">{model_name}</div>
+                        <div style="color: {color}; font-weight: 700; font-size: 1.25rem; font-family: 'SF Mono', Monaco, 'Cascadia Code', 'Roboto Mono', Consolas, 'Courier New', monospace;">{time_formatted}</div>
+                        <div style="color: #6b7280; font-size: 0.85rem; margin-top: 2px;">Total execution time</div>
+                    </div>
+                """
+                )
+
+        return "".join(execution_times_html)
+
+    def calculate_best_worst_tests(
+        self, test_case_results: List[TestCaseResult]
+    ) -> Tuple[str, str]:
+        """Calculate best and worst performing tests."""
+        test_averages = {}
+
+        for test_result in test_case_results:
+            if test_result.average_score > 0:
+                test_averages[test_result.test_case_id] = test_result.average_score
+
+        if test_averages:
+            best_test = max(test_averages, key=test_averages.get)
+            worst_test = min(test_averages, key=test_averages.get)
+
+            best_result = f"{best_test} (avg: {test_averages[best_test]:.2f})"
+            worst_result = f"{worst_test} (avg: {test_averages[worst_test]:.2f})"
+
+            return best_result, worst_result
+
+        return "Not available", "Not available"
+
+    def calculate_average_time(
+        self, model_performances: Dict[str, ModelPerformance]
+    ) -> str:
+        """Calculate overall average time."""
+        all_durations = []
+
+        # Extract durations from model results if available
+        for performance in model_performances.values():
+            if hasattr(performance, "durations") and performance.durations:
+                all_durations.extend(performance.durations)
+
+        if all_durations:
+            avg_duration = sum(all_durations) / len(all_durations)
+            return f"{avg_duration:.1f}s"
+
+        return "Not available"
+
+    def _get_score_class(self, score: float) -> str:
+        """Get CSS class based on score value."""
+        if score >= 0.7:
+            return "score-high"
+        elif score >= 0.4:
+            return "score-medium"
+        else:
+            return "score-low"
+
+
+class ModelResultsProcessor:
+    """Processes individual model results."""
+
+    def __init__(self, file_service: FileService):
+        self.file_service = file_service
+
+    def organize_test_cases(
+        self, model_results: Dict[str, Any]
+    ) -> Dict[str, Dict[str, Any]]:
+        """Organize test cases by test case ID and model."""
+        test_cases = {}
+
+        for model_name, results in model_results.items():
+            if "test_cases" in results:
+                for test_case in results["test_cases"]:
+                    test_case_id = test_case.get("test_case_id")
+                    if test_case_id:
+                        if test_case_id not in test_cases:
+                            test_cases[test_case_id] = {}
+                        test_cases[test_case_id][model_name] = test_case
+
+        return test_cases
+
+    def create_test_case_results(
+        self, test_cases: Dict[str, Dict[str, Any]]
+    ) -> List[TestCaseResult]:
+        """Create TestCaseResult objects from organized test cases."""
+        test_case_results = []
+
+        for test_case_id, models_data in test_cases.items():
+            # Get test case metadata
+            try:
+                # Get path from the first available run data
+                first_model_data = next(iter(models_data.values()))
+                test_case_path = first_model_data["runs"][0]["test_case_path"]
+                test_case_data = load_test_case(test_case_path)
+                category = test_case_data.get("category", "Uncategorized")
+                description = test_case_data.get(
+                    "description", f"Test case: {test_case_id}"
+                )
+            except (StopIteration, IndexError, KeyError):
+                category = "Uncategorized"
+                description = f"Test case: {test_case_id}"
+
+            # Calculate average score across all models and runs
+            all_scores = []
+            for model_data in models_data.values():
+                if isinstance(model_data, dict) and "runs" in model_data:
+                    for run in model_data["runs"]:
+                        if isinstance(run, dict):
+                            score = run.get("llm_eval", {}).get("score")
+                            if score is None:
+                                score = run.get("response_match", 0)
+                            if score is not None and isinstance(score, (int, float)):
+                                all_scores.append(score)
+
+            average_score = sum(all_scores) / len(all_scores) if all_scores else 0
+
+            test_result = TestCaseResult(
+                test_case_id=test_case_id,
+                category=category,
+                description=description,
+                model_results={test_case_id: models_data},
+                average_score=average_score,
+            )
+
+            test_case_results.append(test_result)
+
+        return test_case_results
+
+
+class ReportDataProcessor:
+    """Main processor that coordinates the entire data processing pipeline."""
+
+    def __init__(self):
+        self.file_service = FileService()
+        self.extraction_service = ResultsExtractionService(self.file_service)
+        self.metrics_service = MetricsCalculationService()
+        self.chart_service = ChartDataService(self.file_service)
+        self.template_service = TemplateDataService(self.file_service)
+        self.processor = ModelResultsProcessor(self.file_service)
+
+    def get_evaluation_data(self, results_dir: Path) -> Dict[str, Any]:
+        """Extract and process basic evaluation data."""
+        logger.info("Processing evaluation data...")
+
+        # Initialize metrics
+        metrics = EvaluationMetrics()
+
+        # Extract model results
+        model_results = self.extraction_service.extract_model_results(results_dir)
+        if not model_results:
+            logger.warning("No model results found")
+            return self._metrics_to_dict(metrics)
+
+        # Set basic model information
+        metrics.models = list(model_results.keys())
+
+        # Calculate test statistics
+        total_tests, runs_description = self.metrics_service.calculate_run_statistics(
+            model_results
+        )
+        metrics.total_tests = total_tests
+        metrics.runs = runs_description
+        metrics.test_case_names = self._extract_test_case_names(model_results)
+
+        # Extract execution statistics
+        stats_data = self.extraction_service.extract_execution_stats(results_dir)
+        if stats_data and "total_execution_time" in stats_data:
+            total_time = stats_data["total_execution_time"]
+            metrics.total_execution_time = total_time
+            formatted_time, duration = self.metrics_service.format_execution_time(
+                total_time
+            )
+            metrics.total_execution_time_formatted = formatted_time
+            metrics.duration = duration
+
+        logger.info(f"Processed evaluation data for {len(metrics.models)} models")
+        return self._metrics_to_dict(metrics)
+
+    def get_detailed_evaluation_data(self, results_dir: Path) -> Dict[str, Any]:
+        """Extract and process detailed evaluation data for charts and breakdowns."""
+        logger.info("Processing detailed evaluation data...")
+
+        # Extract model results
+        model_results = self.extraction_service.extract_model_results(results_dir)
+        if not model_results:
+            logger.warning("No model results found for detailed data")
+            return self._empty_detailed_data()
+
+        # Calculate model performances
+        model_performances = {}
+        total_evaluations = 0
+
+        for model_name, results_data in model_results.items():
+            performance = self.metrics_service.calculate_model_performance(
+                model_name, results_data
+            )
+            model_performances[model_name] = performance
+            total_evaluations += performance.test_count
+
+        # Organize test cases
+        test_cases = self.processor.organize_test_cases(model_results)
+        test_case_results = self.processor.create_test_case_results(test_cases)
+
+        # Generate chart configuration
+        chart_config = self.chart_service.generate_chart_configuration(
+            model_results, test_cases
+        )
+
+        # Generate template data
+        performance_metrics_rows = (
+            self.template_service.generate_performance_metrics_table(model_performances)
+        )
+        breakdown_content = self.template_service.generate_breakdown_content(
+            test_case_results, model_performances, model_results
+        )
+        model_execution_times = self.template_service.generate_model_execution_times(
+            model_results
+        )
+
+        # Calculate best/worst tests and average time
+        best_test, worst_test = self.template_service.calculate_best_worst_tests(
+            test_case_results
+        )
+        avg_time = self.template_service.calculate_average_time(model_performances)
+
+        detailed_data = {
+            "performance_metrics_rows": performance_metrics_rows,
+            "breakdown_content": breakdown_content,
+            "best_test": best_test,
+            "worst_test": worst_test,
+            "avg_time": avg_time,
+            "total_evaluations": total_evaluations,
+            "categories_data": json.dumps(chart_config.categories),
+            "chart_datasets_data": json.dumps(chart_config.datasets),
+            "model_execution_times": model_execution_times,
+        }
+
+        logger.info("Processed detailed evaluation data successfully")
+        return detailed_data
+
+    def _extract_test_case_names(self, model_results: Dict[str, Any]) -> List[str]:
+        """Extract unique test case names from model results."""
+        test_case_names = set()
+
+        for results in model_results.values():
+            if "test_cases" in results:
+                for test_case in results["test_cases"]:
+                    test_case_id = test_case.get("test_case_id")
+                    if test_case_id:
+                        test_case_names.add(test_case_id)
+
+        return sorted(list(test_case_names))
+
+    def _metrics_to_dict(self, metrics: EvaluationMetrics) -> Dict[str, Any]:
+        """Convert EvaluationMetrics to dictionary."""
+        # Generate model tags HTML
+        model_tags = ""
+        if metrics.models:
+            model_tags = "".join(
+                [f'<span class="model-tag">{model}</span>' for model in metrics.models]
+            )
+
+        # Generate test cases list HTML
+        test_cases_list = ""
+        if metrics.test_case_names:
+            test_cases_list = "".join(
+                [
+                    f"<li>{test_case}.test.json</li>"
+                    for test_case in metrics.test_case_names
+                ]
+            )
+        elif metrics.total_tests > 0:
+            test_cases_list = "<li>Test cases available (names not loaded)</li>"
+
+        return {
+            "models": metrics.models,
+            "total_execution_time": metrics.total_execution_time,
+            "total_execution_time_formatted": metrics.total_execution_time_formatted,
+            "generation_time": metrics.generation_time,
+            "timestamp": metrics.timestamp,
+            "runs": metrics.runs,
+            "total_tests": metrics.total_tests,
+            "duration": metrics.duration,
+            "test_case_names": metrics.test_case_names,
+            # Template-specific keys
+            "total_models": str(len(metrics.models)),
+            "model_tags": model_tags,
+            "test_cases_list": test_cases_list,
+        }
+
+    def _empty_detailed_data(self) -> Dict[str, Any]:
+        """Return empty detailed data structure."""
+        return {
+            "performance_metrics_rows": "",
+            "breakdown_content": "",
+            "best_test": "Not available",
+            "worst_test": "Not available",
+            "avg_time": "Not available",
+            "total_evaluations": 0,
+            "categories_data": "[]",
+            "chart_datasets_data": "[]",
+            "model_execution_times": "",
+        }
+
+
+def main():
+    """Main entry point for testing the report data processor."""
+    import sys
+    from pathlib import Path
+
+    if len(sys.argv) > 1:
+        results_dir = Path(sys.argv[1])
+    else:
+        # Default to test results directory
+        script_dir = Path(__file__).parent
+        results_dir = script_dir / "results" / "tests"
+
+    processor = ReportDataProcessor()
+
+    print("Testing evaluation data extraction...")
+    eval_data = processor.get_evaluation_data(results_dir)
+    print(f"Found {len(eval_data.get('models', []))} models")
+
+    print("Testing detailed evaluation data extraction...")
+    detailed_data = processor.get_detailed_evaluation_data(results_dir)
+    print(f"Total evaluations: {detailed_data.get('total_evaluations', 0)}")
+
+    print("Report data processing completed successfully!")
+
+
+if __name__ == "__main__":
+    main()