hoytzhang/baidu-indexing-push.html

## baidu-indexing-push.html
<!DOCTYPE html>
<html lang="zh-CN">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>Sitemap/Feed 最新URL提取工具</title>
    <style>
        body {
            font-family: Arial, sans-serif;
            margin: 40px;
            background-color: #f5f5f5;
        }
        .container {
            max-width: 800px;
            margin: 0 auto;
            background: white;
            padding: 30px;
            border-radius: 8px;
            box-shadow: 0 2px 10px rgba(0,0,0,0.1);
        }
        h1 {
            color: #333;
            text-align: center;
        }
        .form-group {
            margin-bottom: 20px;
        }
        label {
            display: block;
            margin-bottom: 5px;
            font-weight: bold;
        }
        input[type="url"] {
            width: 100%;
            padding: 12px;
            border: 1px solid #ddd;
            border-radius: 4px;
            box-sizing: border-box;
        }
        button {
            background-color: #007cba;
            color: white;
            padding: 12px 24px;
            border: none;
            border-radius: 4px;
            cursor: pointer;
            font-size: 16px;
        }
        button:hover {
            background-color: #005a87;
        }
        button:disabled {
            background-color: #cccccc;
            cursor: not-allowed;
        }
        .result {
            margin-top: 30px;
        }
        .url-item {
            padding: 10px;
            border-bottom: 1px solid #eee;
        }
        .url-link {
            color: #007cba;
            text-decoration: none;
        }
        .url-link:hover {
            text-decoration: underline;
        }
        .lastmod {
            color: #666;
            font-size: 14px;
            margin-top: 5px;
        }
        .error {
            color: #d63638;
            background: #fcf0f1;
            padding: 15px;
            border-radius: 4px;
            margin: 20px 0;
        }
        .success {
            color: #008a20;
            background: #edfaef;
            padding: 15px;
            border-radius: 4px;
            margin: 20px 0;
        }
        .source-type {
            margin-bottom: 15px;
        }
        .source-type label {
            display: inline-block;
            margin-right: 20px;
            font-weight: normal;
        }
        .source-type input {
            margin-right: 5px;
        }
        .description {
            color: #666;
            font-size: 14px;
            margin-top: 5px;
        }
        .loading {
            text-align: center;
            padding: 20px;
        }
        .spinner {
            border: 4px solid rgba(0, 0, 0, 0.1);
            border-left-color: #007cba;
            border-radius: 50%;
            width: 30px;
            height: 30px;
            animation: spin 1s linear infinite;
            margin: 0 auto;
        }
        @keyframes spin {
            to { transform: rotate(360deg); }
        }
    </style>
</head>
<body>
    <div class="container">
        <h1>Sitemap/Feed 最新URL提取工具</h1>
        <form id="urlForm">
            <div class="source-type">
                <label><input type="radio" name="source_type" value="sitemap" checked> Sitemap</label>
                <label><input type="radio" name="source_type" value="feed"> RSS/Atom Feed</label>
            </div>

            <div class="form-group">
                <label for="source_url">请输入地址:</label>
                <input type="url" id="source_url" name="source_url" placeholder="https://example.com/sitemap.xml 或 https://example.com/feed" required>
            </div>
            <button type="submit">获取最新URL</button>
        </form>

        <div id="result"></div>

    </div>

    <script>
        document.getElementById('urlForm').addEventListener('submit', async function(e) {
            e.preventDefault();

            const sourceType = document.querySelector('input[name="source_type"]:checked').value;
            const sourceUrl = document.getElementById('source_url').value;
            const resultDiv = document.getElementById('result');

            if (!sourceUrl) {
                resultDiv.innerHTML = '<div class="error">请输入有效的地址</div>';
                return;
            }

            // 显示加载状态
            resultDiv.innerHTML = `
                <div class="loading">
                    <div class="spinner"></div>
                    <p>正在获取数据...</p>
                </div>
            `;

            try {
                const urls = await fetchUrls(sourceType, sourceUrl);
                displayResults(urls, sourceType, sourceUrl);
            } catch (error) {
                resultDiv.innerHTML = `<div class="error"><strong>处理过程中发生错误:</strong><br>${error.message}</div>`;
            }
        });

        async function fetchUrls(sourceType, sourceUrl) {
            // 使用代理解决跨域问题
            const proxyUrl = 'https://api.allorigins.win/raw?url=';
            const response = await fetch(proxyUrl + encodeURIComponent(sourceUrl));

            if (!response.ok) {
                throw new Error(`HTTP error! status: ${response.status}`);
            }

            const content = await response.text();

            if (sourceType === 'sitemap') {
                return parseSitemap(content);
            } else {
                return parseFeed(content);
            }
        }

        function parseSitemap(content) {
            const parser = new DOMParser();
            const xmlDoc = parser.parseFromString(content, 'text/xml');

            // 检查是否为索引sitemap
            const sitemaps = xmlDoc.getElementsByTagName('sitemap');
            if (sitemaps.length > 0) {
                throw new Error('此工具不支持索引sitemap，请提供具体的sitemap文件');
            }

            // 解析普通sitemap
            const urlElements = xmlDoc.getElementsByTagName('url');
            const urls = [];

            for (let i = 0; i < urlElements.length; i++) {
                const urlElement = urlElements[i];
                const loc = urlElement.getElementsByTagName('loc')[0]?.textContent || '';
                const lastmod = urlElement.getElementsByTagName('lastmod')[0]?.textContent || '';

                if (loc) {
                    urls.push({
                        loc: loc,
                        lastmod: lastmod
                    });
                }
            }

            // 按日期排序
            urls.sort((a, b) => {
                const dateA = a.lastmod ? new Date(a.lastmod) : new Date(0);
                const dateB = b.lastmod ? new Date(b.lastmod) : new Date(0);
                return dateB - dateA;
            });

            return urls.slice(0, 10);
        }

        function parseFeed(content) {
            const parser = new DOMParser();
            const xmlDoc = parser.parseFromString(content, 'text/xml');

            let items = [];
            let urls = [];

            // 尝试解析RSS
            const rssItems = xmlDoc.getElementsByTagName('item');
            if (rssItems.length > 0) {
                for (let i = 0; i < rssItems.length; i++) {
                    const item = rssItems[i];
                    const link = item.getElementsByTagName('link')[0]?.textContent || '';
                    const title = item.getElementsByTagName('title')[0]?.textContent || '';
                    const pubDate = item.getElementsByTagName('pubDate')[0]?.textContent || '';
                    const description = item.getElementsByTagName('description')[0]?.textContent || '';

                    if (link) {
                        urls.push({
                            loc: link,
                            title: title,
                            timestamp: pubDate ? new Date(pubDate).getTime() : Date.now(),
                            description: description,
                            pubDate: pubDate
                        });
                    }
                }
            }
            // 尝试解析Atom
            else {
                const atomEntries = xmlDoc.getElementsByTagName('entry');
                for (let i = 0; i < atomEntries.length; i++) {
                    const entry = atomEntries[i];
                    const link = entry.getElementsByTagName('link')[0]?.getAttribute('href') || '';
                    const title = entry.getElementsByTagName('title')[0]?.textContent || '';
                    const updated = entry.getElementsByTagName('updated')[0]?.textContent || '';
                    const published = entry.getElementsByTagName('published')[0]?.textContent || '';
                    const summary = entry.getElementsByTagName('summary')[0]?.textContent || '';

                    const date = updated || published;

                    if (link) {
                        urls.push({
                            loc: link,
                            title: title,
                            timestamp: date ? new Date(date).getTime() : Date.now(),
                            description: summary,
                            pubDate: date
                        });
                    }
                }
            }

            // 按日期排序
            urls.sort((a, b) => b.timestamp - a.timestamp);

            return urls.slice(0, 10);
        }

        function displayResults(urls, sourceType, sourceUrl) {
            const resultDiv = document.getElementById('result');

            if (urls.length === 0) {
                resultDiv.innerHTML = `
                    <div class="source-info">
                        来源类型: ${sourceType === 'sitemap' ? 'Sitemap' : 'RSS/Atom Feed'}<br>
                        来源地址: ${sourceUrl}
                    </div>
                    <div class="no-results">未找到任何URL</div>
                `;
                return;
            }

            let html = `
                <div class="source-info">
                    来源类型: ${sourceType === 'sitemap' ? 'Sitemap' : 'RSS/Atom Feed'}<br>
                    来源地址: ${sourceUrl}
                </div>
                <h1>最新更新的URL</h1>
                <div class="result">
            `;

            urls.forEach(url => {
                html += `
                    <div class="url-item">
                        <div>
                            <a href="${url.loc}" target="_blank" class="url-link">
                                ${url.title || url.loc}
                            </a>
                        </div>
                `;

                if (url.lastmod || url.pubDate) {
                    const dateStr = url.lastmod || url.pubDate;
                    const formattedDate = new Date(dateStr).toLocaleString('zh-CN');
                    html += `
                        <div class="lastmod">
                            ${url.lastmod ? '最后修改时间' : '发布时间'}: ${formattedDate}
                        </div>
                    `;
                }

                if (url.description) {
                    const desc = url.description.replace(/<[^>]*>/g, '').substring(0, 150) + '...';
                    html += `<div class="description">${desc}</div>`;
                }

                html += `</div>`;
            });

            html += `</div>`;
            resultDiv.innerHTML = html;
        }
    </script>
</body>
</html>
	<!DOCTYPE html>
	<html lang="zh-CN">
	<head>
	<meta charset="UTF-8">
	<meta name="viewport" content="width=device-width, initial-scale=1.0">
	<title>Sitemap/Feed 最新URL提取工具</title>
	<style>
	body {
	font-family: Arial, sans-serif;
	margin: 40px;
	background-color: #f5f5f5;
	}
	.container {
	max-width: 800px;
	margin: 0 auto;
	background: white;
	padding: 30px;
	border-radius: 8px;
	box-shadow: 0 2px 10px rgba(0,0,0,0.1);
	}
	h1 {
	color: #333;
	text-align: center;
	}
	.form-group {
	margin-bottom: 20px;
	}
	label {
	display: block;
	margin-bottom: 5px;
	font-weight: bold;
	}
	input[type="url"] {
	width: 100%;
	padding: 12px;
	border: 1px solid #ddd;
	border-radius: 4px;
	box-sizing: border-box;
	}
	button {
	background-color: #007cba;
	color: white;
	padding: 12px 24px;
	border: none;
	border-radius: 4px;
	cursor: pointer;
	font-size: 16px;
	}
	button:hover {
	background-color: #005a87;
	}
	button:disabled {
	background-color: #cccccc;
	cursor: not-allowed;
	}
	.result {
	margin-top: 30px;
	}
	.url-item {
	padding: 10px;
	border-bottom: 1px solid #eee;
	}
	.url-link {
	color: #007cba;
	text-decoration: none;
	}
	.url-link:hover {
	text-decoration: underline;
	}
	.lastmod {
	color: #666;
	font-size: 14px;
	margin-top: 5px;
	}
	.error {
	color: #d63638;
	background: #fcf0f1;
	padding: 15px;
	border-radius: 4px;
	margin: 20px 0;
	}
	.success {
	color: #008a20;
	background: #edfaef;
	padding: 15px;
	border-radius: 4px;
	margin: 20px 0;
	}
	.source-type {
	margin-bottom: 15px;
	}
	.source-type label {
	display: inline-block;
	margin-right: 20px;
	font-weight: normal;
	}
	.source-type input {
	margin-right: 5px;
	}
	.description {
	color: #666;
	font-size: 14px;
	margin-top: 5px;
	}
	.loading {
	text-align: center;
	padding: 20px;
	}
	.spinner {
	border: 4px solid rgba(0, 0, 0, 0.1);
	border-left-color: #007cba;
	border-radius: 50%;
	width: 30px;
	height: 30px;
	animation: spin 1s linear infinite;
	margin: 0 auto;
	}
	@keyframes spin {
	to { transform: rotate(360deg); }
	}
	</style>
	</head>
	<body>
	<div class="container">
	<h1>Sitemap/Feed 最新URL提取工具</h1>
	<form id="urlForm">
	<div class="source-type">
	<label><input type="radio" name="source_type" value="sitemap" checked> Sitemap</label>
	<label><input type="radio" name="source_type" value="feed"> RSS/Atom Feed</label>
	</div>

	<div class="form-group">
	<label for="source_url">请输入地址:</label>
	<input type="url" id="source_url" name="source_url" placeholder="https://example.com/sitemap.xml 或 https://example.com/feed" required>
	</div>
	<button type="submit">获取最新URL</button>
	</form>

	<div id="result"></div>

	</div>

	<script>
	document.getElementById('urlForm').addEventListener('submit', async function(e) {
	e.preventDefault();

	const sourceType = document.querySelector('input[name="source_type"]:checked').value;
	const sourceUrl = document.getElementById('source_url').value;
	const resultDiv = document.getElementById('result');

	if (!sourceUrl) {
	resultDiv.innerHTML = '<div class="error">请输入有效的地址</div>';
	return;
	}

	// 显示加载状态
	resultDiv.innerHTML = `
	<div class="loading">
	<div class="spinner"></div>
	<p>正在获取数据...</p>
	</div>
	`;

	try {
	const urls = await fetchUrls(sourceType, sourceUrl);
	displayResults(urls, sourceType, sourceUrl);
	} catch (error) {
	resultDiv.innerHTML = `<div class="error"><strong>处理过程中发生错误:</strong><br>${error.message}</div>`;
	}
	});

	async function fetchUrls(sourceType, sourceUrl) {
	// 使用代理解决跨域问题
	const proxyUrl = 'https://api.allorigins.win/raw?url=';
	const response = await fetch(proxyUrl + encodeURIComponent(sourceUrl));

	if (!response.ok) {
	throw new Error(`HTTP error! status: ${response.status}`);
	}

	const content = await response.text();

	if (sourceType === 'sitemap') {
	return parseSitemap(content);
	} else {
	return parseFeed(content);
	}
	}

	function parseSitemap(content) {
	const parser = new DOMParser();
	const xmlDoc = parser.parseFromString(content, 'text/xml');

	// 检查是否为索引sitemap
	const sitemaps = xmlDoc.getElementsByTagName('sitemap');
	if (sitemaps.length > 0) {
	throw new Error('此工具不支持索引sitemap，请提供具体的sitemap文件');
	}

	// 解析普通sitemap
	const urlElements = xmlDoc.getElementsByTagName('url');
	const urls = [];

	for (let i = 0; i < urlElements.length; i++) {
	const urlElement = urlElements[i];
	const loc = urlElement.getElementsByTagName('loc')[0]?.textContent \|\| '';
	const lastmod = urlElement.getElementsByTagName('lastmod')[0]?.textContent \|\| '';

	if (loc) {
	urls.push({
	loc: loc,
	lastmod: lastmod
	});
	}
	}

	// 按日期排序
	urls.sort((a, b) => {
	const dateA = a.lastmod ? new Date(a.lastmod) : new Date(0);
	const dateB = b.lastmod ? new Date(b.lastmod) : new Date(0);
	return dateB - dateA;
	});

	return urls.slice(0, 10);
	}

	function parseFeed(content) {
	const parser = new DOMParser();
	const xmlDoc = parser.parseFromString(content, 'text/xml');

	let items = [];
	let urls = [];

	// 尝试解析RSS
	const rssItems = xmlDoc.getElementsByTagName('item');
	if (rssItems.length > 0) {
	for (let i = 0; i < rssItems.length; i++) {
	const item = rssItems[i];
	const link = item.getElementsByTagName('link')[0]?.textContent \|\| '';
	const title = item.getElementsByTagName('title')[0]?.textContent \|\| '';
	const pubDate = item.getElementsByTagName('pubDate')[0]?.textContent \|\| '';
	const description = item.getElementsByTagName('description')[0]?.textContent \|\| '';

	if (link) {
	urls.push({
	loc: link,
	title: title,
	timestamp: pubDate ? new Date(pubDate).getTime() : Date.now(),
	description: description,
	pubDate: pubDate
	});
	}
	}
	}
	// 尝试解析Atom
	else {
	const atomEntries = xmlDoc.getElementsByTagName('entry');
	for (let i = 0; i < atomEntries.length; i++) {
	const entry = atomEntries[i];
	const link = entry.getElementsByTagName('link')[0]?.getAttribute('href') \|\| '';
	const title = entry.getElementsByTagName('title')[0]?.textContent \|\| '';
	const updated = entry.getElementsByTagName('updated')[0]?.textContent \|\| '';
	const published = entry.getElementsByTagName('published')[0]?.textContent \|\| '';
	const summary = entry.getElementsByTagName('summary')[0]?.textContent \|\| '';

	const date = updated \|\| published;

	if (link) {
	urls.push({
	loc: link,
	title: title,
	timestamp: date ? new Date(date).getTime() : Date.now(),
	description: summary,
	pubDate: date
	});
	}
	}
	}

	// 按日期排序
	urls.sort((a, b) => b.timestamp - a.timestamp);

	return urls.slice(0, 10);
	}

	function displayResults(urls, sourceType, sourceUrl) {
	const resultDiv = document.getElementById('result');

	if (urls.length === 0) {
	resultDiv.innerHTML = `
	<div class="source-info">
	来源类型: ${sourceType === 'sitemap' ? 'Sitemap' : 'RSS/Atom Feed'}<br>
	来源地址: ${sourceUrl}
	</div>
	<div class="no-results">未找到任何URL</div>
	`;
	return;
	}

	let html = `
	<div class="source-info">
	来源类型: ${sourceType === 'sitemap' ? 'Sitemap' : 'RSS/Atom Feed'}<br>
	来源地址: ${sourceUrl}
	</div>
	<h1>最新更新的URL</h1>
	<div class="result">
	`;

	urls.forEach(url => {
	html += `
	<div class="url-item">
	<div>
	<a href="${url.loc}" target="_blank" class="url-link">
	${url.title \|\| url.loc}
	</a>
	</div>
	`;

	if (url.lastmod \|\| url.pubDate) {
	const dateStr = url.lastmod \|\| url.pubDate;
	const formattedDate = new Date(dateStr).toLocaleString('zh-CN');
	html += `
	<div class="lastmod">
	${url.lastmod ? '最后修改时间' : '发布时间'}: ${formattedDate}
	</div>
	`;
	}

	if (url.description) {
	const desc = url.description.replace(/<[^>]*>/g, '').substring(0, 150) + '...';
	html += `<div class="description">${desc}</div>`;
	}

	html += `</div>`;
	});

	html += `</div>`;
	resultDiv.innerHTML = html;
	}
	</script>
	</body>
	</html>
No results found