ocr_verify/compare_use_unified_diff.py

import difflib
import json
from typing import List, Dict
from compare_ocr_results import OCRResultComparator

class OCRResultComparatorUnifiedDiff(OCRResultComparator):
    # 继承自OCRResultComparator以复用部分功能
    def __init__(self):
        super().__init__()

    def generate_unified_diff(self, paras1: List[str], paras2: List[str], file1_path: str, file2_path: str) -> Dict:
        """
        生成类似git diff的统一差异格式，并返回结构化数据
        """
        # 直接调用进行预处理
        file1_lines = paras1
        file2_lines = paras2

        # 使用unified_diff生成差异
        diff = difflib.unified_diff(
            file1_lines, 
            file2_lines, 
            fromfile=file1_path, 
            tofile=file2_path,
            lineterm=''  # 确保每行末尾不添加额外字符
        )
        
        # 将差异生成器转换为列表
        diff_output = list(diff)
        
        # 解析diff输出并生成结构化数据
        structured_diff = self._parse_unified_diff(diff_output, file1_lines, file2_lines, file1_path, file2_path)
        
        return structured_diff

    def _parse_unified_diff(self, diff_lines: List[str], file1_lines: List[str], file2_lines: List[str], 
                       file1_path: str, file2_path: str) -> Dict:
        """解析unified diff输出并生成结构化数据"""
        differences = []
        current_hunk = None
        file1_line_num = 0
        file2_line_num = 0
        
        for line in diff_lines:
            if line.startswith('---') or line.startswith('+++'):
                continue
            elif line.startswith('@@'):
                # 解析hunk头部，例如: @@ -1,5 +1,4 @@
                import re
                match = re.match(r'@@ -(\d+)(?:,(\d+))? \+(\d+)(?:,(\d+))? @@', line)
                if match:
                    file1_start = int(match.group(1))
                    file1_count = int(match.group(2)) if match.group(2) else 1
                    file2_start = int(match.group(3))
                    file2_count = int(match.group(4)) if match.group(4) else 1
                    
                    current_hunk = {
                        'file1_start': file1_start,
                        'file1_count': file1_count,
                        'file2_start': file2_start,
                        'file2_count': file2_count
                    }
                    file1_line_num = file1_start - 1  # 转为0基索引
                    file2_line_num = file2_start - 1
            elif line.startswith(' '):
                # 未改变的行
                file1_line_num += 1
                file2_line_num += 1
            elif line.startswith('-'):
                # 文件1中删除的行
                content = line[1:]  # 去掉'-'前缀
                differences.append({
                    'type': 'paragraph',
                    'position': f'段落{file1_line_num + 1}',
                    'file1_value': content,
                    'file2_value': "",
                    'description': '文件1中独有的段落',
                    'similarity': 0.0,
                    'severity': 'medium',
                    'line_number': file1_line_num + 1,
                    'change_type': 'deletion'
                })
                file1_line_num += 1
            elif line.startswith('+'):
                # 文件2中添加的行
                content = line[1:]  # 去掉'+'前缀
                differences.append({
                    'type': 'paragraph',
                    'position': f'段落{file2_line_num + 1}',
                    'file1_value': "",
                    'file2_value': content,
                    'description': '文件2中独有的段落',
                    'similarity': 0.0,
                    'severity': 'medium',
                    'line_number': file2_line_num + 1,
                    'change_type': 'addition'
                })
                file2_line_num += 1
    
        # 计算统计信息
        stats = {
            'total_differences': len(differences),
            'table_differences': 0,  # diff不包含表格差异
            'paragraph_differences': len(differences),
            'amount_differences': 0,
            'high_severity': len([d for d in differences if d.get('severity') == 'high']),
            'medium_severity': len([d for d in differences if d.get('severity') == 'medium']),
            'low_severity': len([d for d in differences if d.get('severity') == 'low']),
            'deletions': len([d for d in differences if d.get('change_type') == 'deletion']),
            'additions': len([d for d in differences if d.get('change_type') == 'addition'])
        }
        
        return {
            'differences': differences,
            'statistics': stats,
            'file1_tables': 0,
            'file2_tables': 0,
            'file1_paragraphs': len(file1_lines),
            'file2_paragraphs': len(file2_lines),
            'file1_path': file1_path,
            'file2_path': file2_path,
            'diff_type': 'unified_diff'
        }

    def generate_unified_diff_report(self, paras1: List[str], paras2: List[str], file1_path: str, file2_path: str, output_file: str):
        """生成unified diff的JSON和Markdown报告"""
        # 生成结构化diff数据
        diff_data = self.generate_unified_diff(paras1, paras2, file1_path, file2_path)
        
        # 添加时间戳
        import datetime
        diff_data['timestamp'] = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')
        
        # 生成JSON报告
        json_file = f"{output_file}_unified_diff.json"
        with open(json_file, 'w', encoding='utf-8') as f:
            json.dump(diff_data, f, ensure_ascii=False, indent=2)
        
        # 生成Markdown报告
        md_file = f"{output_file}_unified_diff.md"
        self._generate_unified_diff_markdown(diff_data, md_file)
        
        print(f"📄 Unified Diff JSON报告: {json_file}")
        print(f"📝 Unified Diff Markdown报告: {md_file}")
        
        return diff_data

    def _generate_unified_diff_markdown(self, diff_data: Dict, output_file: str):
        """生成unified diff的Markdown报告"""
        with open(output_file, 'w', encoding='utf-8') as f:
            f.write("# OCR结果Unified Diff对比报告\n\n")
            
            # 基本信息
            f.write("## 基本信息\n\n")
            f.write(f"- **文件1**: `{diff_data['file1_path']}`\n")
            f.write(f"- **文件2**: `{diff_data['file2_path']}`\n")
            f.write(f"- **比较时间**: {diff_data.get('timestamp', 'N/A')}\n")
            f.write(f"- **对比方式**: Unified Diff\n\n")
            
            # 统计信息
            stats = diff_data['statistics']
            f.write("## 统计信息\n\n")
            f.write(f"- 总差异数量: **{stats['total_differences']}**\n")
            f.write(f"- 删除行数: **{stats['deletions']}**\n")
            f.write(f"- 添加行数: **{stats['additions']}**\n")
            f.write(f"- 文件1段落数: {diff_data['file1_paragraphs']}\n")
            f.write(f"- 文件2段落数: {diff_data['file2_paragraphs']}\n\n")
            
            # 差异详情
            if diff_data['differences']:
                f.write("## 差异详情\n\n")
                
                # 按变更类型分组
                deletions = [d for d in diff_data['differences'] if d['change_type'] == 'deletion']
                additions = [d for d in diff_data['differences'] if d['change_type'] == 'addition']
                
                if deletions:
                    f.write(f"### 🗑️ 删除内容 ({len(deletions)}项)\n\n")
                    for i, diff in enumerate(deletions, 1):
                        f.write(f"**{i}. 第{diff['line_number']}行**\n")
                        f.write(f"```\n{diff['file1_value']}\n```\n\n")
                
                if additions:
                    f.write(f"### ➕ 新增内容 ({len(additions)}项)\n\n")
                    for i, diff in enumerate(additions, 1):
                        f.write(f"**{i}. 第{diff['line_number']}行**\n")
                        f.write(f"```\n{diff['file2_value']}\n```\n\n")
                
                # 详细差异表格
                f.write("## 详细差异列表\n\n")
                f.write("| 序号 | 类型 | 行号 | 变更类型 | 内容 | 描述 |\n")
                f.write("| --- | --- | --- | --- | --- | --- |\n")
                
                for i, diff in enumerate(diff_data['differences'], 1):
                    change_icon = "🗑️" if diff['change_type'] == 'deletion' else "➕"
                    content = diff['file1_value'] if diff['change_type'] == 'deletion' else diff['file2_value']
                    f.write(f"| {i} | {change_icon} | {diff['line_number']} | {diff['change_type']} | ")
                    f.write(f"`{content[:50]}{'...' if len(content) > 50 else ''}` | ")
                    f.write(f"{diff['description']} |\n")
            else:
                f.write("## 结论\n\n")
                f.write("🎉 **完美匹配！没有发现任何差异。**\n\n")