Support Using Only Report Engine
This commit is contained in:
马一丁
@@ -0,0 +1,476 @@
|
||||
#!/usr/bin/env python
|
||||
"""
|
||||
Report Engine 命令行版本
|
||||
|
||||
这是一个不需要前端的命令行报告生成程序。
|
||||
主要流程：
|
||||
1. 检查PDF依赖
|
||||
2. 获取最新的log、md文件
|
||||
3. 直接调用Report Engine生成报告（跳过文件增加审核）
|
||||
4. 自动保存HTML和PDF（如果有依赖）到final_reports/
|
||||
|
||||
使用方法：
|
||||
python report_engine_only.py [选项]
|
||||
|
||||
选项：
|
||||
--query QUERY 指定报告主题（可选，默认从文件名提取）
|
||||
--skip-pdf 跳过PDF生成（即使有依赖）
|
||||
--verbose 显示详细日志
|
||||
--help 显示帮助信息
|
||||
"""
|
||||
|
||||
import os
|
||||
import sys
|
||||
import json
|
||||
import argparse
|
||||
from pathlib import Path
|
||||
from datetime import datetime
|
||||
from typing import Dict, Any, Optional
|
||||
|
||||
from loguru import logger
|
||||
|
||||
# 全局配置
|
||||
VERBOSE = False
|
||||
|
||||
# 配置日志
|
||||
def setup_logger(verbose: bool = False):
|
||||
"""设置日志配置"""
|
||||
global VERBOSE
|
||||
VERBOSE = verbose
|
||||
|
||||
logger.remove() # 移除默认处理器
|
||||
logger.add(
|
||||
sys.stdout,
|
||||
format="<green>{time:YYYY-MM-DD HH:mm:ss}</green> | <level>{level: <8}</level> | <level>{message}</level>",
|
||||
level="DEBUG" if verbose else "INFO"
|
||||
)
|
||||
|
||||
|
||||
def check_dependencies() -> tuple[bool, Optional[str]]:
|
||||
"""
|
||||
检查PDF生成所需的系统依赖
|
||||
|
||||
Returns:
|
||||
tuple: (is_available: bool, message: str)
|
||||
- is_available: PDF功能是否可用
|
||||
- message: 依赖检查结果消息
|
||||
"""
|
||||
logger.info("=" * 70)
|
||||
logger.info("步骤 1/4: 检查系统依赖")
|
||||
logger.info("=" * 70)
|
||||
|
||||
try:
|
||||
from ReportEngine.utils.dependency_check import check_pango_available
|
||||
is_available, message = check_pango_available()
|
||||
|
||||
if is_available:
|
||||
logger.success("✓ PDF 依赖检测通过，将同时生成 HTML 和 PDF 文件")
|
||||
else:
|
||||
logger.warning("⚠ PDF 依赖缺失，仅生成 HTML 文件")
|
||||
logger.info("\n" + message)
|
||||
|
||||
return is_available, message
|
||||
except Exception as e:
|
||||
logger.error(f"依赖检查失败: {e}")
|
||||
return False, str(e)
|
||||
|
||||
|
||||
def get_latest_engine_reports() -> Dict[str, str]:
|
||||
"""
|
||||
获取三个引擎目录中的最新报告文件
|
||||
|
||||
Returns:
|
||||
Dict[str, str]: 引擎名称到文件路径的映射
|
||||
"""
|
||||
logger.info("\n" + "=" * 70)
|
||||
logger.info("步骤 2/4: 获取最新的分析引擎报告")
|
||||
logger.info("=" * 70)
|
||||
|
||||
# 定义三个引擎的目录
|
||||
directories = {
|
||||
'insight': 'insight_engine_streamlit_reports',
|
||||
'media': 'media_engine_streamlit_reports',
|
||||
'query': 'query_engine_streamlit_reports'
|
||||
}
|
||||
|
||||
latest_files = {}
|
||||
|
||||
for engine, directory in directories.items():
|
||||
if not os.path.exists(directory):
|
||||
logger.warning(f"⚠ {engine.capitalize()} Engine 目录不存在: {directory}")
|
||||
continue
|
||||
|
||||
# 获取所有 .md 文件
|
||||
md_files = [f for f in os.listdir(directory) if f.endswith('.md')]
|
||||
|
||||
if not md_files:
|
||||
logger.warning(f"⚠ {engine.capitalize()} Engine 目录中没有找到 .md 文件")
|
||||
continue
|
||||
|
||||
# 获取最新文件
|
||||
latest_file = max(
|
||||
md_files,
|
||||
key=lambda x: os.path.getmtime(os.path.join(directory, x))
|
||||
)
|
||||
latest_path = os.path.join(directory, latest_file)
|
||||
latest_files[engine] = latest_path
|
||||
|
||||
logger.info(f"✓ 找到 {engine.capitalize()} Engine 最新报告")
|
||||
|
||||
if not latest_files:
|
||||
logger.error("❌ 未找到任何引擎报告文件，请先运行分析引擎生成报告")
|
||||
sys.exit(1)
|
||||
|
||||
logger.info(f"\n共找到 {len(latest_files)} 个引擎的最新报告")
|
||||
|
||||
return latest_files
|
||||
|
||||
|
||||
def confirm_file_selection(latest_files: Dict[str, str]) -> bool:
|
||||
"""
|
||||
向用户确认选择的文件是否正确
|
||||
|
||||
Args:
|
||||
latest_files: 引擎名称到文件路径的映射
|
||||
|
||||
Returns:
|
||||
bool: 用户确认则返回True，否则返回False
|
||||
"""
|
||||
logger.info("\n" + "=" * 70)
|
||||
logger.info("请确认以下选择的文件：")
|
||||
logger.info("=" * 70)
|
||||
|
||||
for engine, file_path in latest_files.items():
|
||||
filename = os.path.basename(file_path)
|
||||
# 获取文件修改时间
|
||||
mtime = os.path.getmtime(file_path)
|
||||
mtime_str = datetime.fromtimestamp(mtime).strftime('%Y-%m-%d %H:%M:%S')
|
||||
|
||||
logger.info(f" {engine.capitalize()} Engine:")
|
||||
logger.info(f" 文件名: {filename}")
|
||||
logger.info(f" 路径: {file_path}")
|
||||
logger.info(f" 修改时间: {mtime_str}")
|
||||
logger.info("")
|
||||
|
||||
logger.info("=" * 70)
|
||||
|
||||
# 提示用户确认
|
||||
try:
|
||||
response = input("是否使用以上文件生成报告? [Y/n]: ").strip().lower()
|
||||
|
||||
# 默认是y，所以空输入或y都表示确认
|
||||
if response == '' or response == 'y' or response == 'yes':
|
||||
logger.success("✓ 用户确认，继续生成报告")
|
||||
return True
|
||||
else:
|
||||
logger.warning("✗ 用户取消操作")
|
||||
return False
|
||||
except (KeyboardInterrupt, EOFError):
|
||||
logger.warning("\n✗ 用户取消操作")
|
||||
return False
|
||||
|
||||
|
||||
def load_engine_reports(latest_files: Dict[str, str]) -> list[str]:
|
||||
"""
|
||||
加载引擎报告内容
|
||||
|
||||
Args:
|
||||
latest_files: 引擎名称到文件路径的映射
|
||||
|
||||
Returns:
|
||||
list[str]: 报告内容列表
|
||||
"""
|
||||
reports = []
|
||||
|
||||
for engine, file_path in latest_files.items():
|
||||
try:
|
||||
with open(file_path, 'r', encoding='utf-8') as f:
|
||||
content = f.read()
|
||||
reports.append(content)
|
||||
logger.debug(f"已加载 {engine} 报告，长度: {len(content)} 字符")
|
||||
except Exception as e:
|
||||
logger.error(f"加载 {engine} 报告失败: {e}")
|
||||
|
||||
return reports
|
||||
|
||||
|
||||
def extract_query_from_reports(latest_files: Dict[str, str]) -> str:
|
||||
"""
|
||||
从报告文件名中提取查询主题
|
||||
|
||||
Args:
|
||||
latest_files: 引擎名称到文件路径的映射
|
||||
|
||||
Returns:
|
||||
str: 提取的查询主题
|
||||
"""
|
||||
# 尝试从文件名中提取主题
|
||||
for engine, file_path in latest_files.items():
|
||||
filename = os.path.basename(file_path)
|
||||
# 假设文件名格式为: report_主题_时间戳.md
|
||||
if '_' in filename:
|
||||
parts = filename.replace('.md', '').split('_')
|
||||
if len(parts) >= 2:
|
||||
# 提取中间部分作为主题
|
||||
topic = '_'.join(parts[1:-1]) if len(parts) > 2 else parts[1]
|
||||
if topic:
|
||||
return topic
|
||||
|
||||
# 如果无法提取，返回默认值
|
||||
return "综合分析报告"
|
||||
|
||||
|
||||
def generate_report(reports: list[str], query: str, pdf_available: bool) -> Dict[str, Any]:
|
||||
"""
|
||||
调用Report Engine生成报告
|
||||
|
||||
Args:
|
||||
reports: 报告内容列表
|
||||
query: 报告主题
|
||||
pdf_available: PDF功能是否可用
|
||||
|
||||
Returns:
|
||||
Dict[str, Any]: 包含生成结果的字典
|
||||
"""
|
||||
logger.info("\n" + "=" * 70)
|
||||
logger.info("步骤 3/4: 生成综合报告")
|
||||
logger.info("=" * 70)
|
||||
logger.info(f"报告主题: {query}")
|
||||
logger.info(f"输入报告数量: {len(reports)}")
|
||||
|
||||
try:
|
||||
from ReportEngine.agent import ReportAgent
|
||||
|
||||
# 初始化Report Agent
|
||||
logger.info("正在初始化 Report Engine...")
|
||||
agent = ReportAgent()
|
||||
|
||||
# 定义流式事件处理器
|
||||
def stream_handler(event_type: str, payload: Dict[str, Any]):
|
||||
"""处理Report Engine的流式事件"""
|
||||
if event_type == 'stage':
|
||||
stage = payload.get('stage', '')
|
||||
if stage == 'agent_start':
|
||||
logger.info(f"开始生成报告: {payload.get('report_id', '')}")
|
||||
elif stage == 'template_selected':
|
||||
logger.info(f"✓ 已选择模板: {payload.get('template', '')}")
|
||||
elif stage == 'template_sliced':
|
||||
logger.info(f"✓ 模板解析完成，共 {payload.get('section_count', 0)} 个章节")
|
||||
elif stage == 'layout_designed':
|
||||
logger.info(f"✓ 文档布局设计完成")
|
||||
logger.info(f" 标题: {payload.get('title', '')}")
|
||||
elif stage == 'word_plan_ready':
|
||||
logger.info(f"✓ 篇幅规划完成，目标章节数: {payload.get('chapter_targets', 0)}")
|
||||
elif stage == 'chapters_compiled':
|
||||
logger.info(f"✓ 章节生成完成，共 {payload.get('chapter_count', 0)} 个章节")
|
||||
elif stage == 'html_rendered':
|
||||
logger.info(f"✓ HTML 渲染完成")
|
||||
elif stage == 'report_saved':
|
||||
logger.info(f"✓ 报告已保存")
|
||||
elif event_type == 'chapter_status':
|
||||
chapter_id = payload.get('chapterId', '')
|
||||
title = payload.get('title', '')
|
||||
status = payload.get('status', '')
|
||||
if status == 'generating':
|
||||
logger.info(f" 正在生成章节: {title}")
|
||||
elif status == 'completed':
|
||||
attempt = payload.get('attempt', 1)
|
||||
warning = payload.get('warning', '')
|
||||
if warning:
|
||||
logger.warning(f" ✓ 章节完成: {title} (第 {attempt} 次尝试，{payload.get('warningMessage', '')})")
|
||||
else:
|
||||
logger.success(f" ✓ 章节完成: {title}")
|
||||
elif event_type == 'error':
|
||||
logger.error(f"错误: {payload.get('message', '')}")
|
||||
|
||||
# 生成报告
|
||||
logger.info("开始生成报告，这可能需要几分钟时间...")
|
||||
result = agent.generate_report(
|
||||
query=query,
|
||||
reports=reports,
|
||||
forum_logs="", # 不使用论坛日志
|
||||
custom_template="", # 使用自动模板选择
|
||||
save_report=True, # 自动保存报告
|
||||
stream_handler=stream_handler
|
||||
)
|
||||
|
||||
logger.success("✓ 报告生成成功！")
|
||||
return result
|
||||
|
||||
except Exception as e:
|
||||
logger.exception(f"❌ 报告生成失败: {e}")
|
||||
sys.exit(1)
|
||||
|
||||
|
||||
def save_pdf(document_ir_path: str, query: str) -> Optional[str]:
|
||||
"""
|
||||
从IR文件生成并保存PDF
|
||||
|
||||
Args:
|
||||
document_ir_path: Document IR文件路径
|
||||
query: 报告主题
|
||||
|
||||
Returns:
|
||||
Optional[str]: PDF文件路径，如果失败则返回None
|
||||
"""
|
||||
logger.info("\n正在生成 PDF 文件...")
|
||||
|
||||
try:
|
||||
# 读取IR数据
|
||||
with open(document_ir_path, 'r', encoding='utf-8') as f:
|
||||
document_ir = json.load(f)
|
||||
|
||||
# 创建PDF渲染器
|
||||
from ReportEngine.renderers import PDFRenderer
|
||||
renderer = PDFRenderer()
|
||||
|
||||
# 生成PDF字节流
|
||||
pdf_bytes = renderer.render_to_bytes(document_ir, optimize_layout=True)
|
||||
|
||||
# 保存PDF文件
|
||||
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
|
||||
query_safe = "".join(
|
||||
c for c in query if c.isalnum() or c in (" ", "-", "_")
|
||||
).rstrip()
|
||||
query_safe = query_safe.replace(" ", "_")[:30] or "report"
|
||||
|
||||
pdf_dir = Path("final_reports") / "pdf"
|
||||
pdf_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
pdf_filename = f"final_report_{query_safe}_{timestamp}.pdf"
|
||||
pdf_path = pdf_dir / pdf_filename
|
||||
|
||||
pdf_path.write_bytes(pdf_bytes)
|
||||
|
||||
logger.success(f"✓ PDF 已保存: {pdf_path}")
|
||||
return str(pdf_path)
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"❌ PDF 生成失败: {e}")
|
||||
return None
|
||||
|
||||
|
||||
def parse_arguments():
|
||||
"""解析命令行参数"""
|
||||
parser = argparse.ArgumentParser(
|
||||
description="Report Engine 命令行版本 - 无需前端的报告生成工具",
|
||||
formatter_class=argparse.RawDescriptionHelpFormatter,
|
||||
epilog="""
|
||||
示例:
|
||||
python report_engine_only.py
|
||||
python report_engine_only.py --query "土木工程行业分析"
|
||||
python report_engine_only.py --skip-pdf --verbose
|
||||
|
||||
注意:
|
||||
程序会自动获取三个引擎目录中的最新报告文件，
|
||||
不进行文件增加审核，直接生成综合报告。
|
||||
"""
|
||||
)
|
||||
|
||||
parser.add_argument(
|
||||
'--query',
|
||||
type=str,
|
||||
default=None,
|
||||
help='指定报告主题（默认从文件名自动提取）'
|
||||
)
|
||||
|
||||
parser.add_argument(
|
||||
'--skip-pdf',
|
||||
action='store_true',
|
||||
help='跳过PDF生成（即使系统支持）'
|
||||
)
|
||||
|
||||
parser.add_argument(
|
||||
'--verbose',
|
||||
action='store_true',
|
||||
help='显示详细日志信息'
|
||||
)
|
||||
|
||||
return parser.parse_args()
|
||||
|
||||
|
||||
def main():
|
||||
"""主函数"""
|
||||
# 解析命令行参数
|
||||
args = parse_arguments()
|
||||
|
||||
# 设置日志
|
||||
setup_logger(verbose=args.verbose)
|
||||
|
||||
logger.info("\n")
|
||||
logger.info("╔" + "═" * 68 + "╗")
|
||||
logger.info("║" + " " * 20 + "Report Engine 命令行版本" + " " * 24 + "║")
|
||||
logger.info("╚" + "═" * 68 + "╝")
|
||||
logger.info("\n")
|
||||
|
||||
# 步骤 1: 检查依赖
|
||||
pdf_available, _ = check_dependencies()
|
||||
|
||||
# 如果用户指定跳过PDF，则禁用PDF生成
|
||||
if args.skip_pdf:
|
||||
logger.info("用户指定 --skip-pdf，将跳过 PDF 生成")
|
||||
pdf_available = False
|
||||
|
||||
# 步骤 2: 获取最新文件
|
||||
latest_files = get_latest_engine_reports()
|
||||
|
||||
# 确认文件选择
|
||||
if not confirm_file_selection(latest_files):
|
||||
logger.info("\n程序已退出")
|
||||
sys.exit(0)
|
||||
|
||||
# 加载报告内容
|
||||
reports = load_engine_reports(latest_files)
|
||||
|
||||
if not reports:
|
||||
logger.error("❌ 未能加载任何报告内容")
|
||||
sys.exit(1)
|
||||
|
||||
# 提取或使用指定的查询主题
|
||||
query = args.query if args.query else extract_query_from_reports(latest_files)
|
||||
logger.info(f"使用报告主题: {query}")
|
||||
|
||||
# 步骤 3: 生成报告
|
||||
result = generate_report(reports, query, pdf_available)
|
||||
|
||||
# 步骤 4: 保存文件
|
||||
logger.info("\n" + "=" * 70)
|
||||
logger.info("步骤 4/4: 保存生成的文件")
|
||||
logger.info("=" * 70)
|
||||
|
||||
# HTML 已经在 generate_report 中自动保存
|
||||
html_path = result.get('report_filepath', '')
|
||||
if html_path:
|
||||
logger.success(f"✓ HTML 已保存: {result.get('report_relative_path', html_path)}")
|
||||
|
||||
# 如果有PDF依赖，生成并保存PDF
|
||||
if pdf_available:
|
||||
ir_path = result.get('ir_filepath', '')
|
||||
if ir_path and os.path.exists(ir_path):
|
||||
pdf_path = save_pdf(ir_path, query)
|
||||
else:
|
||||
logger.warning("⚠ 未找到 IR 文件，无法生成 PDF")
|
||||
else:
|
||||
logger.info("⚠ 跳过 PDF 生成（缺少系统依赖或用户指定跳过）")
|
||||
|
||||
# 总结
|
||||
logger.info("\n" + "=" * 70)
|
||||
logger.success("✓ 报告生成完成！")
|
||||
logger.info("=" * 70)
|
||||
logger.info(f"报告 ID: {result.get('report_id', 'N/A')}")
|
||||
logger.info(f"HTML 文件: {result.get('report_relative_path', 'N/A')}")
|
||||
if pdf_available:
|
||||
logger.info(f"PDF 文件: final_reports/pdf/ 目录下")
|
||||
logger.info("=" * 70)
|
||||
logger.info("\n程序结束")
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
try:
|
||||
main()
|
||||
except KeyboardInterrupt:
|
||||
logger.warning("\n\n用户中断程序")
|
||||
sys.exit(0)
|
||||
except Exception as e:
|
||||
logger.exception(f"\n程序异常退出: {e}")
|
||||
sys.exit(1)
|
||||
Reference in New Issue
Block a user