124 lines
4.8 KiB
HTML
124 lines
4.8 KiB
HTML
|
<!DOCTYPE html>
|
|||
|
|
<html lang="zh-CN">
|
|||
|
|
<head>
|
|||
|
|
<meta charset="UTF-8">
|
|||
|
|
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
|||
|
|
<title>表格检测调试</title>
|
|||
|
|
<style>
|
|||
|
|
body { font-family: Arial, sans-serif; margin: 20px; }
|
|||
|
|
.test-case { margin: 20px 0; padding: 15px; border: 1px solid #ddd; border-radius: 5px; }
|
|||
|
|
.content { background: #f8f9fa; padding: 10px; font-family: monospace; white-space: pre-wrap; }
|
|||
|
|
.result { margin-top: 10px; padding: 10px; border-radius: 3px; }
|
|||
|
|
.success { background: #d4edda; color: #155724; }
|
|||
|
|
.error { background: #f8d7da; color: #721c24; }
|
|||
|
|
</style>
|
|||
|
|
</head>
|
|||
|
|
<body>
|
|||
|
|
<h1>表格检测调试</h1>
|
|||
|
|
<div id="results"></div>
|
|||
|
|
|
|||
|
|
<script>
|
|||
|
|
// 复制hasTable函数
|
|||
|
|
function hasTable(content) {
|
|||
|
|
if (!content || typeof content !== 'string') {
|
|||
|
|
console.log('🔍 hasTable: 内容为空或不是字符串');
|
|||
|
|
return false;
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
console.log('🔍 hasTable: 检查内容:', content);
|
|||
|
|
|
|||
|
|
// 检查是否包含表格语法
|
|||
|
|
const lines = content.split('\n');
|
|||
|
|
let hasTableRow = false;
|
|||
|
|
let hasSeparator = false;
|
|||
|
|
|
|||
|
|
console.log('🔍 hasTable: 分割后的行数:', lines.length);
|
|||
|
|
|
|||
|
|
for (const line of lines) {
|
|||
|
|
const trimmedLine = line.trim();
|
|||
|
|
console.log('🔍 hasTable: 检查行:', trimmedLine);
|
|||
|
|
|
|||
|
|
// 检查表格行（包含|字符且至少有3个部分）
|
|||
|
|
if (trimmedLine.includes('|') && trimmedLine.split('|').length >= 3) {
|
|||
|
|
hasTableRow = true;
|
|||
|
|
console.log('✅ hasTable: 找到表格行');
|
|||
|
|
}
|
|||
|
|
// 检查分隔符行（包含-和|，且主要由这些字符组成）
|
|||
|
|
if (trimmedLine.includes('|') && trimmedLine.includes('-') && /^[\s\|\-\:]+$/.test(trimmedLine)) {
|
|||
|
|
hasSeparator = true;
|
|||
|
|
console.log('✅ hasTable: 找到分隔符行');
|
|||
|
|
}
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// 如果只有表格行但没有分隔符，也可能是表格
|
|||
|
|
// 或者如果内容包含多个|字符，也可能是表格
|
|||
|
|
const pipeCount = (content.match(/\|/g) || []).length;
|
|||
|
|
const hasMultiplePipes = pipeCount >= 6; // 至少3行，每行2个|
|
|||
|
|
|
|||
|
|
console.log('🔍 hasTable: 结果 - hasTableRow:', hasTableRow, 'hasSeparator:', hasSeparator, 'pipeCount:', pipeCount, 'hasMultiplePipes:', hasMultiplePipes);
|
|||
|
|
|
|||
|
|
const result = (hasTableRow && hasSeparator) || (hasTableRow && hasMultiplePipes);
|
|||
|
|
console.log('🔍 hasTable: 最终结果:', result);
|
|||
|
|
|
|||
|
|
return result;
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// 测试用例
|
|||
|
|
const testCases = [
|
|||
|
|
{
|
|||
|
|
name: "数据质量评估矩阵",
|
|||
|
|
content: `| 评估维度 | 权重 | 评分标准 | 计算公式 |
|
|||
|
|
|---------|------|---------|---------|
|
|||
|
|
| 完整性 | 0.3 | 缺失值比例 < 5% | $\\text{完整性} = 1 - \\frac{\\text{缺失值数量}}{\\text{总数据量}}$ |
|
|||
|
|
| 准确性 | 0.3 | 误差率 < 3% | $\\text{准确性} = 1 - \\frac{\\text{错误记录数}}{\\text{总记录数}}$ |`
|
|||
|
|
},
|
|||
|
|
{
|
|||
|
|
name: "产品价格表",
|
|||
|
|
content: `产品价格表
|
|||
|
|
| 产品 | 价格 |
|
|||
|
|
|------|------|
|
|||
|
|
| 苹果 | 4元 |
|
|||
|
|
| 香蕉 | 2元 |`
|
|||
|
|
},
|
|||
|
|
{
|
|||
|
|
name: "简单表格",
|
|||
|
|
content: `| 姓名 | 年龄 | 城市 |
|
|||
|
|
|------|------|------|
|
|||
|
|
| 张三 | 25 | 北京 |
|
|||
|
|
| 李四 | 30 | 上海 |`
|
|||
|
|
},
|
|||
|
|
{
|
|||
|
|
name: "只有表格行，没有分隔符",
|
|||
|
|
content: `| 姓名 | 年龄 | 城市 |
|
|||
|
|
| 张三 | 25 | 北京 |
|
|||
|
|
| 李四 | 30 | 上海 |`
|
|||
|
|
},
|
|||
|
|
{
|
|||
|
|
name: "包含表格字符但不是表格",
|
|||
|
|
content: `这是一个包含|字符的文本，但不是表格`
|
|||
|
|
}
|
|||
|
|
];
|
|||
|
|
|
|||
|
|
// 运行测试
|
|||
|
|
const resultsDiv = document.getElementById('results');
|
|||
|
|
|
|||
|
|
testCases.forEach((testCase, index) => {
|
|||
|
|
const testDiv = document.createElement('div');
|
|||
|
|
testDiv.className = 'test-case';
|
|||
|
|
|
|||
|
|
const result = hasTable(testCase.content);
|
|||
|
|
|
|||
|
|
testDiv.innerHTML = `
|
|||
|
|
<h3>测试用例 ${index + 1}: ${testCase.name}</h3>
|
|||
|
|
<div class="content">${testCase.content}</div>
|
|||
|
|
<div class="result ${result ? 'success' : 'error'}">
|
|||
|
|
检测结果: ${result ? '✅ 检测到表格' : '❌ 未检测到表格'}
|
|||
|
|
</div>
|
|||
|
|
`;
|
|||
|
|
|
|||
|
|
resultsDiv.appendChild(testDiv);
|
|||
|
|
});
|
|||
|
|
</script>
|
|||
|
|
</body>
|
|||
|
|
</html>
|