how/data/run_single.py

"""
测试脚本：运行 待解构帖子.json（带历史帖子）

功能：
1. 加载最近3篇历史帖子（从早到晚排序）
2. 加载待解构帖子
3. 运行 WhatDeconstructionWorkflow
"""

import json
import sys
import os
import argparse
from pathlib import Path
from datetime import datetime

# 添加项目根目录到路径
project_root = Path(__file__).parent.parent
sys.path.insert(0, str(project_root))

# 手动加载.env文件
def load_env_file(env_path):
    """手动加载.env文件"""
    if not env_path.exists():
        return False

    with open(env_path, 'r') as f:
        for line in f:
            line = line.strip()
            # 跳过注释和空行
            if not line or line.startswith('#'):
                continue
            # 解析KEY=VALUE
            if '=' in line:
                key, value = line.split('=', 1)
                os.environ[key.strip()] = value.strip()

    return True

env_path = project_root / ".env"
if load_env_file(env_path):
    print(f"✅ 已加载环境变量从: {env_path}")
    # 验证API密钥
    api_key = os.environ.get("GEMINI_API_KEY", "")
    if api_key:
        print(f"   GEMINI_API_KEY: {api_key[:10]}...")
else:
    print(f"⚠️  未找到.env文件: {env_path}")

from src.workflows.what_deconstruction_workflow import WhatDeconstructionWorkflow
from src.utils.logger import get_logger

logger = get_logger(__name__)


def load_historical_posts(history_dir, target_timestamp=None, target_post_id=None, max_count=10):
    """
    加载历史帖子（根据publish_timestamp从新到旧排序）

    选择比目标帖子早发布，并且是最近发布的帖子，排除目标帖子本身

    Args:
        history_dir: 历史帖子目录
        target_timestamp: 目标帖子的发布时间戳（可选）
        target_post_id: 目标帖子的ID（用于过滤重复，可选）
        max_count: 最多加载的帖子数量

    Returns:
        list: 历史帖子列表（从新到旧排序）
    """
    history_path = Path(history_dir)

    if not history_path.exists():
        print(f"⚠️  历史帖子目录不存在: {history_path}")
        return []

    # 获取所有JSON文件
    json_files = list(history_path.glob("*.json"))

    if not json_files:
        print(f"⚠️  未找到历史帖子文件")
        return []

    print(f"\n📁 找到 {len(json_files)} 个历史帖子文件")

    # 读取所有帖子并提取publish_timestamp
    posts_with_timestamp = []
    for file_path in json_files:
        try:
            with open(file_path, 'r', encoding='utf-8') as f:
                post_data = json.load(f)

            # 获取发布时间戳，如果不存在则使用0
            timestamp = post_data.get("publish_timestamp", 0)
            post_id = post_data.get("channel_content_id", "")

            posts_with_timestamp.append({
                "file_path": file_path,
                "post_data": post_data,
                "timestamp": timestamp,
                "post_id": post_id
            })
        except Exception as e:
            print(f"   ⚠️  读取文件失败 {file_path.name}: {e}")
            continue

    if not posts_with_timestamp:
        print(f"⚠️  没有成功读取到任何帖子")
        return []

    # 过滤掉目标帖子本身
    if target_post_id is not None:
        original_count = len(posts_with_timestamp)
        posts_with_timestamp = [
            post for post in posts_with_timestamp
            if post["post_id"] != target_post_id
        ]
        filtered_count = original_count - len(posts_with_timestamp)
        if filtered_count > 0:
            print(f"🔍 过滤掉 {filtered_count} 个重复帖子（目标帖子本身）")

    # 如果提供了目标时间戳，只保留比目标帖子早的帖子
    if target_timestamp is not None:
        posts_with_timestamp = [
            post for post in posts_with_timestamp
            if post["timestamp"] < target_timestamp
        ]
        print(f"📊 筛选出 {len(posts_with_timestamp)} 个比目标帖子早的历史帖子")

        if not posts_with_timestamp:
            print(f"⚠️  没有找到比目标帖子早的历史帖子")
            return []

    # 按照publish_timestamp排序（从新到旧）
    posts_with_timestamp.sort(key=lambda x: x["timestamp"], reverse=True)

    # 选择最近的N篇（从新到旧）
    selected_posts = posts_with_timestamp[:max_count] if len(posts_with_timestamp) > max_count else posts_with_timestamp

    print(f"📋 选择最近 {len(selected_posts)} 篇历史帖子（按发布时间从新到旧）:")

    historical_posts = []
    for idx, post_info in enumerate(selected_posts, 1):
        post_data = post_info["post_data"]
        file_path = post_info["file_path"]
        timestamp = post_info["timestamp"]

        # 转换为需要的格式
        historical_post = {
            "text": {
                "title": post_data.get("title", ""),
                "body": post_data.get("body_text", ""),
                "hashtags": ""
            },
            "images": post_data.get("images", [])
        }

        historical_posts.append(historical_post)

        # 格式化时间显示
        publish_time = post_data.get("publish_time", "未知时间")
        print(f"   {idx}. {file_path.name}")
        print(f"      标题: {post_data.get('title', '无标题')}")
        print(f"      发布时间: {publish_time}")
        print(f"      图片数: {len(post_data.get('images', []))}")

    return historical_posts


def load_test_data(directory):
    """
    加载测试数据

    Args:
        directory: 帖子目录名（如"阿里多多酱"或"G88818"）
    """
    test_data_path = Path(__file__).parent / directory / "待解构帖子.json"

    with open(test_data_path, "r", encoding="utf-8") as f:
        data = json.load(f)

    return data


def convert_to_workflow_input(raw_data, historical_posts=None):
    """
    将原始数据转换为工作流输入格式

    Args:
        raw_data: 原始帖子数据
        historical_posts: 历史帖子列表（可选）
    """
    images = raw_data.get("images", [])

    input_data = {
        "multimedia_content": {
            "images": images,
            "video": raw_data.get("video", {}),
            "text": {
                "title": raw_data.get("title", ""),
                "body": raw_data.get("body_text", ""),
                "hashtags": ""
            }
        },
        "comments": raw_data.get("comments", []),
        "creator_info": {
            "nickname": raw_data.get("channel_account_name", ""),
            "account_id": raw_data.get("channel_account_id", "")
        }
    }

    # 如果有历史帖子，添加到输入数据中
    if historical_posts:
        input_data["historical_posts"] = historical_posts

    return input_data


def main():
    """主函数"""
    # 解析命令行参数
    parser = argparse.ArgumentParser(description='运行单个帖子的What解构工作流')
    parser.add_argument('directory', type=str, help='帖子目录名（如"阿里多多酱"或"G88818"）')
    args = parser.parse_args()

    directory = args.directory

    print("=" * 80)
    print(f"开始测试 What 解构工作流（带历史帖子）- 目录: {directory}")
    print("=" * 80)

    # 1. 加载测试数据（目标帖子）
    print("\n[1] 加载测试数据（目标帖子）...")
    try:
        raw_data = load_test_data(directory)
        target_timestamp = raw_data.get('publish_timestamp')
        target_post_id = raw_data.get('channel_content_id')
        target_publish_time = raw_data.get('publish_time', '未知时间')

        print(f"✅ 成功加载测试数据")
        print(f"   - 标题: {raw_data.get('title')}")
        print(f"   - 帖子ID: {target_post_id}")
        print(f"   - 发布时间: {target_publish_time}")
        print(f"   - 图片数: {len(raw_data.get('images', []))}")
        print(f"   - 点赞数: {raw_data.get('like_count')}")
        print(f"   - 评论数: {raw_data.get('comment_count')}")
    except Exception as e:
        print(f"❌ 加载测试数据失败: {e}")
        return

    # 2. 加载历史帖子（比目标帖子早的帖子，排除目标帖子本身）
    print("\n[2] 加载历史帖子...")
    history_dir = Path(__file__).parent / directory / "作者历史帖子"
    historical_posts = load_historical_posts(
        history_dir,
        target_timestamp=target_timestamp,
        target_post_id=target_post_id,
        max_count=15
    )

    if historical_posts:
        print(f"✅ 成功加载 {len(historical_posts)} 篇历史帖子")
    else:
        print(f"⚠️  未加载到历史帖子，将使用常规分析模式")

    # 3. 转换数据格式
    print("\n[3] 转换数据格式...")
    try:
        input_data = convert_to_workflow_input(raw_data, historical_posts)
        print(f"✅ 数据格式转换成功")
        print(f"   - 话题标签: {input_data['multimedia_content']['text']['hashtags']}")
        print(f"   - 历史帖子数: {len(input_data.get('historical_posts', []))}")
    except Exception as e:
        print(f"❌ 数据格式转换失败: {e}")
        return

    # 4. 初始化工作流
    print("\n[4] 初始化工作流...")
    try:
        workflow = WhatDeconstructionWorkflow(
            model_provider="google_genai",
            max_depth=10
        )
        print(f"✅ 工作流初始化成功")
    except Exception as e:
        print(f"❌ 工作流初始化失败: {e}")
        import traceback
        traceback.print_exc()
        return

    # 5. 执行工作流
    print("\n[5] 执行工作流...")
    print("   注意：这可能需要几分钟时间...")
    try:
        result = workflow.invoke(input_data)
        print(f"✅ 工作流执行成功")
    except Exception as e:
        print(f"❌ 工作流执行失败: {e}")
        import traceback
        traceback.print_exc()
        return

    # 6. 保存结果
    print("\n[6] 保存结果...")
    try:
        # 生成带时间戳的文件名
        timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
        output_filename = f"result_{timestamp}.json"
        output_path = Path(__file__).parent / directory / "output" / output_filename
        output_path.parent.mkdir(parents=True, exist_ok=True)

        with open(output_path, "w", encoding="utf-8") as f:
            json.dump(result, f, ensure_ascii=False, indent=2)

        print(f"✅ 结果已保存到: {output_path}")
        print(f"   文件名: {output_filename}")
    except Exception as e:
        print(f"❌ 保存结果失败: {e}")
        return

    # 7. 生成HTML可视化
    # print("\n[7] 生成HTML可视化...")
    # try:
    #     visualize_script = Path(__file__).parent / "visualize_result.py"
    #     if visualize_script.exists():
    #         import subprocess
    #         result_viz = subprocess.run(
    #             [sys.executable, str(visualize_script), str(output_path)],
    #             capture_output=True,
    #             text=True
    #         )
    #         if result_viz.returncode == 0:
    #             print(f"✅ HTML可视化生成成功")
    #             # 查找生成的HTML文件
    #             html_file = output_path.parent / f"{output_path.stem}_visualization.html"
    #             if html_file.exists():
    #                 print(f"   可视化文件: {html_file}")
    #         else:
    #             print(f"⚠️  HTML可视化生成失败: {result_viz.stderr}")
    #     else:
    #         print(f"⚠️  未找到可视化脚本: {visualize_script}")
    # except Exception as e:
    #     print(f"⚠️  生成HTML可视化失败: {e}")

    # 8. 显示结果摘要
    print("\n" + "=" * 80)
    print("结果摘要")
    print("=" * 80)

    if result:
        three_points = result.get("三点解构", {})
        inspiration_data = three_points.get("灵感点", {})
        keypoints_data = three_points.get("关键点", {})
        comments = result.get("评论分析", {}).get("解构维度", [])

        print(f"\n三点解构:")
        print(f"   - 灵感点数量: {inspiration_data.get('total_count', 0)}")
        print(f"   - 灵感点分析模式: {inspiration_data.get('analysis_mode', '未知')}")
        print(f"   - 目的点数量: 1")
        print(f"   - 关键点数量: {keypoints_data.get('total_count', 0)}")

        # 显示灵感点详情
        if inspiration_data.get('points'):
            print(f"\n灵感点列表:")
            for idx, point in enumerate(inspiration_data['points'], 1):
                print(f"   {idx}. {point.get('灵感点', '')}")

        print(f"\n评论分析:")
        print(f"   - 解构维度数: {len(comments)}")

        topic_understanding = result.get("选题理解", {})
        if topic_understanding:
            topic_theme = topic_understanding.get("topic_theme", "")
            print(f"\n选题理解:")
            print(f"   - 选题主题: {topic_theme}")

    print("\n" + "=" * 80)
    print("测试完成！")
    print("=" * 80)


if __name__ == "__main__":
    main()