data/processed/RQ1/RecruitmentAssistant-H_A2A/analyze_retry_patterns.py

#!/usr/bin/env python3
"""
Analyze RETRY patterns in the RecruitmentAssistant-H_A2A project.
Summarize error locations, retry counts, and retry rates.
"""

import os
import re
from pathlib import Path
from collections import defaultdict
import json
import csv

# Base path and model list
BASE_DIR = Path("/Users/wzr/TOSEM-2025/RESULTS")
MODELS = [
    "DeepSeek-R1",
    "DeepSeek-V3-1",
    "GPT-4o-mini",
    "GPT-5",
    "Gemini-2.5-flash",
    "Gemini-2.5-flash-nothinking",
    "Qwen3-235b",
]
PROJECT_NAME = "RecruitmentAssistant-H_A2A"


def extract_error_info(line: str) -> dict:
    """Extract error information from an error line.

    Returns:
        {'has_error': bool, 'node_type': str, 'node_name': str, 'error_msg': str}
    """
    # Remove tree-structure characters
    clean = re.sub(r"^[│├└\-\s]+", "", line).strip()

    # Check error marker
    if "❌" not in clean:
        return {"has_error": False}

    # Remove error marker
    clean = clean.split("❌", 1)[1].lstrip()

    # Extract node type and name
    # Improvement: handle ERROR info better to avoid including trailing token stats
    node_match = re.match(
        r"\[(SPAN|Chain|AGENT|Tool|LLM)\]\s+([^\[]+?)(?:\s+\[ERROR:\s*(.+?)\])?(?:\s+\[.*)?$",
        clean,
    )
    if not node_match:
        return {
            "has_error": True,
            "node_type": "Unknown",
            "node_name": "Unknown",
            "error_msg": "",
        }

    node_type = node_match.group(1)
    node_name = node_match.group(2).strip()
    error_msg = node_match.group(3).strip() if node_match.group(3) else ""

    # Handle the truncated-error placeholder
    if "[error too long, truncated]" in error_msg:
        error_msg = error_msg.replace("[error too long, truncated]", "(truncated)")

    # Normalize node name
    if node_type == "AGENT":
        node_name = re.sub(r"\._execute_core$", "", node_name)
    elif node_type == "Tool":
        node_name = re.sub(r"\._use$", "", node_name)
    elif node_type == "Chain":
        node_name = re.sub(r"Crew_[a-f0-9\-]+\.kickoff", "Crew***.kickoff", node_name)

    return {
        "has_error": True,
        "node_type": node_type,
        "node_name": node_name,
        "error_msg": error_msg,
    }


def extract_retry_info(line: str) -> dict:
    """Extract RETRY information from a retry line.

    Returns:
        {'is_retry': bool, 'retry_number': int, 'node_type': str, 'node_name': str}
    """
    # Remove tree-structure characters
    clean = re.sub(r"^[│├└\-\s]+", "", line).strip()

    # Detect retry markers: (retry N) or [RETRYN]
    retry_match = re.search(r"\(retry\s+(\d+)\)", clean)
    if not retry_match:
        retry_match = re.search(r"\[RETRY(\d+)\]", clean)

    if not retry_match:
        return {"is_retry": False}

    retry_number = int(retry_match.group(1))

    # Extract node type and name
    # Improvement: match node name more accurately
    node_match = re.match(
        r"\[(SPAN|Chain|AGENT)\]\s+([^\[\(]+?)(?:\s+\(retry\s+\d+\))?(?:\s+\[RETRY\d+\])?.*$",
        clean,
    )
    if not node_match:
        return {
            "is_retry": True,
            "retry_number": retry_number,
            "node_type": "Unknown",
            "node_name": "Unknown",
        }

    node_type = node_match.group(1)
    node_name = node_match.group(2).strip()

    return {
        "is_retry": True,
        "retry_number": retry_number,
        "node_type": node_type,
        "node_name": node_name,
    }


def analyze_session(md_file: str) -> dict:
    """Analyze one session's execution_path.md.

    Returns:
        {
            'has_error': bool,
            'has_retry': bool,
            'max_retry_number': int,
            'total_retries': int,
            'errors': [{'node_type': str, 'node_name': str, 'error_msg': str}, ...],
            'retries': [{'retry_number': int, 'node_type': str, 'node_name': str}, ...]
        }
    """
    if not os.path.exists(md_file):
        return None

    with open(md_file, "r", encoding="utf-8") as f:
        content = f.read()

    # Extract the "Execution Path Tree" section
    tree_match = re.search(
        r"## Execution Path Tree.*?```\n(.*?)```", content, re.DOTALL
    )
    if not tree_match:
        return None

    tree_content = tree_match.group(1)

    errors = []
    retries = []

    for line in tree_content.split("\n"):
        if not line.strip():
            continue

        # Detect errors
        error_info = extract_error_info(line)
        if error_info["has_error"]:
            errors.append(
                {
                    "node_type": error_info.get("node_type", "Unknown"),
                    "node_name": error_info.get("node_name", "Unknown"),
                    "error_msg": error_info.get("error_msg", ""),
                }
            )

        # Detect retries
        retry_info = extract_retry_info(line)
        if retry_info["is_retry"]:
            retries.append(
                {
                    "retry_number": retry_info["retry_number"],
                    "node_type": retry_info["node_type"],
                    "node_name": retry_info["node_name"],
                }
            )

    max_retry = max([r["retry_number"] for r in retries]) if retries else 0

    return {
        "has_error": len(errors) > 0,
        "has_retry": len(retries) > 0,
        "max_retry_number": max_retry,
        "total_retries": len(retries),
        "errors": errors,
        "retries": retries,
    }


def collect_model_stats(model_name: str) -> dict:
    """Collect RETRY statistics for a single model.

    Returns:
        {
            'model': str,
            'total_sessions': int,
            'sessions_with_error': int,
            'sessions_with_retry': int,
            'total_retry_attempts': int,
            'retry_rate': float,
            'error_by_agent': {agent_name: count},
            'error_types': {error_msg: count},
            'max_retry_number': int,
            'session_details': [...]
        }
    """
    test_results_dir = BASE_DIR / model_name / PROJECT_NAME / "test_results"

    if not test_results_dir.exists():
        return None

    stats = {
        "model": model_name,
        "total_sessions": 0,
        "sessions_with_error": 0,
        "sessions_with_retry": 0,
        "total_retry_attempts": 0,
        "error_by_agent": defaultdict(int),
        "error_by_node_type": defaultdict(int),
        "error_types": defaultdict(int),
        "max_retry_number": 0,
        "session_details": [],
    }

    for session_dir in sorted(test_results_dir.iterdir()):
        if not session_dir.is_dir():
            continue

        exec_path_file = session_dir / "execution_path.md"
        if not exec_path_file.exists():
            continue

        stats["total_sessions"] += 1

        analysis = analyze_session(str(exec_path_file))
        if not analysis:
            continue

        # Aggregate errors and retries
        if analysis["has_error"]:
            stats["sessions_with_error"] += 1

        if analysis["has_retry"]:
            stats["sessions_with_retry"] += 1
            stats["total_retry_attempts"] += analysis["total_retries"]
            stats["max_retry_number"] = max(
                stats["max_retry_number"], analysis["max_retry_number"]
            )

        # Error locations
        for error in analysis["errors"]:
            if error["node_type"] == "AGENT":
                stats["error_by_agent"][error["node_name"]] += 1
            stats["error_by_node_type"][error["node_type"]] += 1

            # Build a short signature for error types
            error_msg = error["error_msg"]
            if error_msg:
                # Use the first 100 characters as an error-type key
                error_type = (
                    error_msg[:100]
                    if len(error_msg) <= 100
                    else error_msg[:100] + "..."
                )
                stats["error_types"][error_type] += 1

        # Save session details
        stats["session_details"].append(
            {
                "session": session_dir.name,
                "has_error": analysis["has_error"],
                "has_retry": analysis["has_retry"],
                "retry_count": analysis["total_retries"],
                "errors": analysis["errors"],
                "retries": analysis["retries"],
            }
        )

    # Compute retry rate
    stats["retry_rate"] = (
        (stats["sessions_with_retry"] / stats["total_sessions"] * 100)
        if stats["total_sessions"] > 0
        else 0
    )
    stats["error_rate"] = (
        (stats["sessions_with_error"] / stats["total_sessions"] * 100)
        if stats["total_sessions"] > 0
        else 0
    )

    # Convert defaultdict to dict
    stats["error_by_agent"] = dict(stats["error_by_agent"])
    stats["error_by_node_type"] = dict(stats["error_by_node_type"])
    stats["error_types"] = dict(stats["error_types"])

    return stats


def print_summary(all_stats):
    """Print summary statistics."""
    print("\n" + "=" * 100)
    print(f"RETRY pattern analysis summary - {PROJECT_NAME}")
    print("=" * 100 + "\n")

    # Overall table
    print("## Per-model summary\n")
    print(
        f"{'Model':<35} {'Total_Sessions':<14} {'Error_Rate':<12} {'Retry_Rate':<12} {'Total_Retries':<14} {'Max_Retry':<10}"
    )
    print("-" * 100)

    for stats in all_stats:
        if stats:
            print(
                f"{stats['model']:<35} {stats['total_sessions']:<12} "
                f"{stats['error_rate']:>10.1f}% {stats['retry_rate']:>10.1f}% "
                f"{stats['total_retry_attempts']:<12} {stats['max_retry_number']:<10}"
            )

    print("\n" + "=" * 100)

    # Per-model details
    for stats in all_stats:
        if not stats or stats["sessions_with_error"] == 0:
            continue

        print(f"\n### 📊 {stats['model']}\n")
        print(f"- **Total sessions**: {stats['total_sessions']}")
        print(
            f"- **Sessions with errors**: {stats['sessions_with_error']} ({stats['error_rate']:.1f}%)"
        )
        print(
            f"- **Sessions with retries**: {stats['sessions_with_retry']} ({stats['retry_rate']:.1f}%)"
        )
        print(f"- **Total retry attempts**: {stats['total_retry_attempts']}")
        print(f"- **Max retry number**: {stats['max_retry_number']}")

        if stats["error_by_agent"]:
            print("\n**Agents where errors occurred**:")
            for agent, count in sorted(
                stats["error_by_agent"].items(), key=lambda x: x[1], reverse=True
            ):
                print(f"  - {agent}: {count}")

        if stats["error_by_node_type"]:
            print("\n**Node types where errors occurred**:")
            for node_type, count in sorted(
                stats["error_by_node_type"].items(), key=lambda x: x[1], reverse=True
            ):
                print(f"  - {node_type}: {count}")

        if stats["error_types"]:
            print("\n**Error types (top 5)**:")
            for error_type, count in sorted(
                stats["error_types"].items(), key=lambda x: x[1], reverse=True
            )[:5]:
                print(f"  - [{count}] {error_type}")

        print("\n" + "-" * 100)


def save_results(all_stats):
    """Save results to files."""
    output_dir = Path(__file__).parent

    # Save detailed JSON
    json_file = output_dir / "retry_analysis.json"
    json_data = []
    for stats in all_stats:
        if stats:
            json_data.append(stats)

    with open(json_file, "w", encoding="utf-8") as f:
        json.dump(json_data, f, indent=2, ensure_ascii=False)
    print(f"\n✅ Detailed JSON saved: {json_file}")

    # Save CSV summary
    csv_file = output_dir / "retry_summary.csv"
    with open(csv_file, "w", newline="", encoding="utf-8") as f:
        writer = csv.writer(f)
        writer.writerow(
            [
                "Model",
                "Total_Sessions",
                "Sessions_With_Error",
                "Error_Rate(%)",
                "Sessions_With_Retry",
                "Retry_Rate(%)",
                "Total_Retry_Attempts",
                "Max_Retry_Number",
            ]
        )

        for stats in all_stats:
            if stats:
                writer.writerow(
                    [
                        stats["model"],
                        stats["total_sessions"],
                        stats["sessions_with_error"],
                        f"{stats['error_rate']:.2f}",
                        stats["sessions_with_retry"],
                        f"{stats['retry_rate']:.2f}",
                        stats["total_retry_attempts"],
                        stats["max_retry_number"],
                    ]
                )

    print(f"✅ CSV summary saved: {csv_file}")

    # Save error-location stats
    error_csv_file = output_dir / "error_by_agent.csv"
    with open(error_csv_file, "w", newline="", encoding="utf-8") as f:
        writer = csv.writer(f)
        writer.writerow(["Model", "Agent_Name", "Error_Count"])

        for stats in all_stats:
            if stats and stats["error_by_agent"]:
                for agent, count in sorted(
                    stats["error_by_agent"].items(), key=lambda x: x[1], reverse=True
                ):
                    writer.writerow([stats["model"], agent, count])

    print(f"✅ Error location stats saved: {error_csv_file}")


if __name__ == "__main__":
    print(f"Starting RETRY pattern analysis - {PROJECT_NAME}...")

    all_stats = []
    for model in MODELS:
        print(f"\n📊 Analyzing model: {model}")
        stats = collect_model_stats(model)
        if stats:
            all_stats.append(stats)
            print(
                f"  ✅ Done: {stats['total_sessions']} sessions, "
                f"{stats['sessions_with_error']} errors, "
                f"{stats['sessions_with_retry']} retries"
            )
        else:
            print("  ⚠️  Skipped (directory does not exist)")

    print_summary(all_stats)
    save_results(all_stats)

    print("\n" + "=" * 100)
    print("✅ Done!")
    print("=" * 100)