Merge pull request opendatalab#37 from darkrush/main_html

Main html

Author: e06084

examples/main_html_eval.py

@@ -0,0 +1,142 @@
+#!/usr/bin/env python3
+"""
+WebMainBench 基本使用示例
+"""
+
+import json
+from pathlib import Path
+
+# 导入 WebMainBench 模块
+from webmainbench import (
+    DataLoader, DataSaver, BenchmarkDataset, DataSample,
+    ExtractorFactory, MainHTMLEvaluator,
+    format_results, setup_logging
+)
+
+
+def load_benchdata(dataset_path: str) -> BenchmarkDataset:
+    dataset_path = Path(dataset_path)
+    print(f"📂 数据集文件: {dataset_path}")
+    
+    if not dataset_path.exists():
+        print(f"❌ 数据文件不存在: {dataset_path}")
+        print("请确保已运行数据提取命令创建样本数据集")
+        return
+    
+    # 加载数据集
+    dataset = DataLoader.load_jsonl(dataset_path, include_results=False)
+    dataset.name = "real_preprocessed_html_test"
+    dataset.description = "基于真实数据的预处理HTML功能测试"
+    return dataset
+
+
+def load_extractor(model_path: str):
+    extractor = ExtractorFactory.create("dripper", config={"model_path": model_path})
+    return extractor
+
+
+def save_results(result_file: Path, results: list[dict]):
+    with result_file.open("w", encoding="utf-8") as f:
+        for res in results:
+            f.write(json.dumps(res, ensure_ascii=False) + "\n")
+    
+    
+
+def demo_llm_webkit_with_preprocessed_html_evaluation(model_path: str):
+    """演示LLM-WebKit预处理HTML功能的评测"""
+    
+    print("\n=== LLM-WebKit 预处理HTML功能演示 ===\n")
+    
+    # 设置日志
+    setup_logging(level="INFO")
+    
+    # 1. 从真实数据集加载包含预处理HTML的数据
+    print("1. 从真实数据集加载预处理HTML数据...")
+    
+    # 使用DataLoader加载真实的样本数据
+   
+    dataset = load_benchdata("data/WebMainBench_llm-webkit_v1_WebMainBench_1827_v1_WebMainBench_dataset_merge_with_llm_webkit.jsonl")
+    print(f"✅ 真实数据集加载成功，包含 {len(dataset)} 个样本")
+    
+
+    
+    # 2. 创建预处理HTML模式的LLM-WebKit抽取器
+    print("2. 创建预处理HTML模式的LLM-WebKit抽取器...")
+    
+    extractor = load_extractor(model_path)
+    print(f"✅ 抽取器创建成功")
+    print(f"📋 配置信息:")
+    print(f"  - 跳过LLM推理: 是（直接处理预处理HTML）")
+    print()
+    
+    # 4. 运行评测
+    print("4. 开始评测...")
+    print("=" * 50)
+    
+    evaluator = MainHTMLEvaluator()
+    result = evaluator.evaluate(
+        dataset=dataset,
+        extractor=extractor,
+        max_samples=None
+    )
+    
+    # 5. 显示评测结果
+    print("\n5. 📊 预处理HTML模式评测结果:")
+    print("=" * 50)
+    
+    results_dict = result.to_dict()
+    metrics = results_dict.get('overall_metrics', {})
+    
+    # 显示关键指标
+    print(f"\n🏆 综合指标:")
+    for key in metrics.keys():
+        print(f"  {key}: {metrics[key]:.4f}")
+    
+    print(f"\n⚡ 性能统计:")
+    sample_results = results_dict.get('sample_results', [])
+    if sample_results:
+        extraction_times = [s.get('extraction_time', 0) for s in sample_results if s.get('extraction_success')]
+        if extraction_times:
+            avg_time = sum(extraction_times) / len(extraction_times)
+            print(f"  平均提取时间: {avg_time:.3f}秒")
+            print(f"  处理速度: {1/avg_time:.1f}样本/秒")
+    
+    success_count = len([s for s in sample_results if s.get('extraction_success', False)])
+    print(f"  成功样本数: {success_count}/{len(dataset)}")
+    
+    # 7. 保存结果
+    print(f"\n6. 💾 保存评测结果...")
+    
+    results_dir = Path("results")
+    results_dir.mkdir(exist_ok=True)
+    # 新增：保存带抽取结果的增强数据集（JSONL格式）
+    jsonl_dataset_path = results_dir / f"{extractor.name}_preprocessed_html_dataset_with_results.jsonl"
+    save_results(jsonl_dataset_path, result.sample_results)
+    print(f"✅ 结果已保存到: {jsonl_dataset_path}")
+    
+    
+    print(f"✅ 带抽取结果的JSONL数据集已保存到: {jsonl_dataset_path}")
+    results_path = results_dir / f"{extractor.name}_preprocessed_html_evaluation_results.json"
+    report_path = results_dir / f"{extractor.name}_preprocessed_html_evaluation_report.csv"
+    
+    DataSaver.save_evaluation_results(result, results_path)
+    DataSaver.save_summary_report(result, report_path)
+    
+    print(f"✅ 详细结果已保存到: {results_path}")
+    print(f"✅ CSV报告已保存到: {report_path}")
+    
+
+
+if __name__ == "__main__":
+    import argparse
+    parser = argparse.ArgumentParser(description="WebMainBench 基本使用示例")
+    parser.add_argument("--model_path", required=True, help="LLM model路径")
+    args = parser.parse_args()
+    try:
+        demo_llm_webkit_with_preprocessed_html_evaluation(args.model_path)
+        print("\n✅ 示例运行完成！")
+        
+    except Exception as e:
+        print(f"\n❌ 运行出错: {e}")
+        import traceback
+        traceback.print_exc() 

requirements.txt

@@ -10,4 +10,5 @@ trafilatura
 # llm-web-kit==3.2.0
 https://github.com/opendatalab/magic-html/releases/download/magic_html-0.1.5-released/magic_html-0.1.5-py3-none-any.whl
 streamlit
-markdown
+markdown
+jieba

webmainbench/__init__.py

@@ -11,7 +11,7 @@
 from .data import DataLoader, DataSaver, BenchmarkDataset, DataSample
 from .extractors import BaseExtractor, ExtractorFactory, ExtractionResult
 from .metrics import BaseMetric, MetricCalculator, MetricResult
-from .evaluator import Evaluator, EvaluationResult
+from .evaluator import Evaluator, EvaluationResult, MainHTMLEvaluator
 from .utils import setup_logging, format_results
 
 __all__ = [
@@ -29,4 +29,5 @@
     "EvaluationResult",
     "setup_logging",
     "format_results",
+    "MainHTMLEvaluator"
 ] 

webmainbench/evaluator/__init__.py

@@ -5,8 +5,9 @@
 """
 
 from .evaluator import Evaluator, EvaluationResult
-
+from .main_html_evaluator import MainHTMLEvaluator
 __all__ = [
     "Evaluator",
     "EvaluationResult",
+    "MainHTMLEvaluator"
 ]