Enhanced ChatGPT Clone: Features Agents, MCP, Skills, DeepSeek, Anthropic, AWS, OpenAI, Responses API, Azure, Groq, o1, GPT-5, Mistral, OpenRouter, Vertex AI, Gemini, Artifacts, AI model switching, message search, Code Interpreter, langchain, DALL-E-3, OpenAPI Actions, Functions, Secure Multi-User Auth, Presets, open-source for self-hosting. Active
TypeScript aianthropicartifactsawsazurechatgptchatgpt-cloneclaudeclonedeepseekgeminigooglegpt-5librechatmcpo1openairesponses-apivisionwebui
import { isSafeUrl, getMarkdownFiles, EMBEDDED_IS_SAFE_URL } from '../markdown';
describe('isSafeUrl', () => {
it('allows https URLs', () => {
expect(isSafeUrl('https://example.com')).toBe(true);
});
it('allows http URLs', () => {
expect(isSafeUrl('http://example.com/path')).toBe(true);
});
it('allows mailto links', () => {
expect(isSafeUrl('mailto:[email protected]')).toBe(true);
});
it('allows tel links', () => {
expect(isSafeUrl('tel:+1234567890')).toBe(true);
});
it('allows relative paths', () => {
expect(isSafeUrl('/path/to/page')).toBe(true);
expect(isSafeUrl('./relative')).toBe(true);
expect(isSafeUrl('../parent')).toBe(true);
});
it('allows anchor links', () => {
expect(isSafeUrl('#section')).toBe(true);
});
it('blocks javascript: protocol', () => {
expect(isSafeUrl('javascript:alert(1)')).toBe(false);
});
it('blocks javascript: with leading whitespace', () => {
expect(isSafeUrl(' javascript:alert(1)')).toBe(false);
});
it('blocks javascript: with mixed case', () => {
expect(isSafeUrl('JavaScript:alert(1)')).toBe(false);
});
it('blocks data: protocol', () => {
expect(isSafeUrl('data:text/html,<b>x</b>')).toBe(false);
});
it('blocks blob: protocol', () => {
expect(isSafeUrl('blob:http://example.com/uuid')).toBe(false);
});
it('blocks vbscript: protocol', () => {
expect(isSafeUrl('vbscript:MsgBox("xss")')).toBe(false);
});
it('blocks file: protocol', () => {
expect(isSafeUrl('file:///etc/passwd')).toBe(false);
});
it('blocks empty strings', () => {
expect(isSafeUrl('')).toBe(false);
});
it('blocks whitespace-only strings', () => {
expect(isSafeUrl(' ')).toBe(false);
});
it('blocks unknown/custom protocols', () => {
expect(isSafeUrl('custom:payload')).toBe(false);
});
});
describe('isSafeUrl sync verification', () => {
const embeddedFn = new Function('url', EMBEDDED_IS_SAFE_URL + '\nreturn isSafeUrl(url);') as (
url: string,
) => boolean;
const cases: [string, boolean][] = [
['https://example.com', true],
['http://example.com', true],
['mailto:[email protected]', true],
['tel:+1234567890', true],
['/relative', true],
['./relative', true],
['../up', true],
['#anchor', true],
['javascript:alert(1)', false],
[' javascript:void(0)', false],
['data:text/html,<b>x</b>', false],
['blob:http://x.com/uuid', false],
['vbscript:run', false],
['file:///etc/passwd', false],
['custom:payload', false],
['', false],
[' ', false],
];
it.each(cases)('embedded copy matches exported isSafeUrl for %j → %s', (url, expected) => {
expect(embeddedFn(url)).toBe(expected);
expect(isSafeUrl(url)).toBe(expected);
});
});
describe('markdown artifacts', () => {
describe('getMarkdownFiles', () => {
it('should return content.md with the original markdown content', () => {
const markdown = '# Hello World\n\nThis is a test.';
const files = getMarkdownFiles(markdown);
expect(files['content.md']).toBe(markdown);
});
it('should return default content when markdown is empty', () => {
const files = getMarkdownFiles('');
expect(files['content.md']).toBe('# No content provided');
});
it('should include index.html with static markdown rendering', () => {
const markdown = '# Test';
const files = getMarkdownFiles(markdown);
expect(files['index.html']).toContain('<!DOCTYPE html>');
expect(files['index.html']).toContain('marked.min.js');
expect(files['index.html']).toContain('marked.parse');
expect(files['index.html']).toContain('# Test');
});
it('should only produce content.md and index.html', () => {
const files = getMarkdownFiles('# Test');
expect(Object.keys(files).sort()).toEqual(['content.md', 'index.html']);
});
it('should include markdown CSS in index.html', () => {
const files = getMarkdownFiles('# Test');
expect(files['index.html']).toContain('.markdown-body');
expect(files['index.html']).toContain('list-style-type: disc');
expect(files['index.html']).toContain('prefers-color-scheme: dark');
});
/** The base sheet picks its palette from the iframe's own
* `prefers-color-scheme`, so the contrast block is appended unconditionally
* and therefore also overrides that media query. */
it('appends a contrast override that outranks the media query', () => {
const standard = getMarkdownFiles('# Test')['index.html'];
expect(standard).toContain('prefers-color-scheme: dark');
expect(standard).not.toContain('background-color: #000000');
const contrastDark = getMarkdownFiles('# Test', true, true)['index.html'];
const overrideIndex = contrastDark.lastIndexOf('.markdown-body { color: #ffffff');
expect(overrideIndex).toBeGreaterThan(contrastDark.indexOf('prefers-color-scheme: dark'));
expect(contrastDark).toContain('background-color: #000000');
expect(contrastDark).toContain('color: #8cc8ff');
const contrastLight = getMarkdownFiles('# Test', false, true)['index.html'];
expect(contrastLight).toContain('.markdown-body { color: #000000');
expect(contrastLight).toContain('color: #0000cc');
});
/** Anything the base sheet colours inside the media query has to be answered
* by the appended block, or an explicit contrast choice keeps a GitHub
* palette value wherever the override forgot to reach. */
it('answers every media-query colour in the contrast block', () => {
const styleOf = (html: string) =>
html.slice(html.indexOf('<style>') + '<style>'.length, html.indexOf('</style>'));
const baseStyle = styleOf(getMarkdownFiles('# Test')['index.html']);
const override = styleOf(getMarkdownFiles('# Test', true, true)['index.html']).slice(
baseStyle.length,
);
/** `selector -> property` pairs, one per selector in a comma-separated list. */
const declarations = (css: string): string[] =>
[...css.matchAll(/([^{}]+)\{([^{}]*)\}/g)].flatMap(([, selectors, body]) =>
selectors.split(',').flatMap((selector) =>
body
.split(';')
.map((declaration) => declaration.split(':')[0].trim())
.filter(Boolean)
.map((property) => `${selector.trim()} -> ${property}`),
),
);
const answered = new Set(declarations(override));
const mediaBlocks = [
...baseStyle.matchAll(/@media \(prefers-color-scheme: dark\) \{([\s\S]*?)\n\}/g),
];
expect(mediaBlocks.length).toBeGreaterThan(0);
const unanswered = mediaBlocks
.flatMap(([, block]) => declarations(block))
.filter((declaration) => !answered.has(declaration));
expect(unanswered).toEqual([]);
});
/** A CDN outage inside a contrast mode is exactly when the message matters,
* and an inline style would be the one colour the override cannot reach. */
it('themes the renderer failure message', () => {
const standard = getMarkdownFiles('# Test')['index.html'];
expect(standard).toContain('class="markdown-error"');
expect(standard).not.toMatch(/style="color:/);
expect(standard).toContain('color: #e53e3e');
expect(getMarkdownFiles('# Test', true, true)['index.html']).toContain(
'.markdown-error { color: #ff8f8f; }',
);
expect(getMarkdownFiles('# Test', false, true)['index.html']).toContain(
'.markdown-error { color: #a10000; }',
);
});
describe('content escaping', () => {
it('should escape backticks in markdown content', () => {
const markdown = 'Here is some `inline code`';
const files = getMarkdownFiles(markdown);
expect(files['index.html']).toContain('\\`');
});
it('should escape backslashes in markdown content', () => {
const markdown = 'Path: C:\\Users\\Test';
const files = getMarkdownFiles(markdown);
expect(files['index.html']).toContain('\\\\');
});
it('should escape dollar signs in markdown content', () => {
const markdown = 'Price: $100';
const files = getMarkdownFiles(markdown);
expect(files['index.html']).toContain('\\$');
});
it('should handle code blocks with backticks', () => {
const markdown = '```js\nconsole.log("test");\n```';
const files = getMarkdownFiles(markdown);
expect(files['index.html']).toContain('\\`\\`\\`');
});
it('should prevent </script> in content from breaking out of the script block', () => {
const markdown = 'Some content with </script><script>alert(1)</script>';
const files = getMarkdownFiles(markdown);
expect(files['index.html']).not.toContain('</script><script>alert(1)');
expect(files['index.html']).toContain('<\\/script');
});
});
describe('list indentation normalization', () => {
it('should normalize 2-space indented lists to 4-space', () => {
const markdown = '- Item 1\n - Subitem 1\n - Subitem 2';
const files = getMarkdownFiles(markdown);
expect(files['content.md']).toBe(markdown);
expect(files['index.html']).toContain('- Item 1');
expect(files['index.html']).toContain('Subitem 1');
});
it('should handle numbered lists with 2-space indents', () => {
const markdown = '1. First\n 2. Second nested';
const files = getMarkdownFiles(markdown);
expect(files['content.md']).toBe(markdown);
expect(files['index.html']).toContain('1. First');
expect(files['index.html']).toContain('2. Second nested');
});
it('should not affect already 4-space indented lists', () => {
const markdown = '- Item 1\n - Subitem 1';
const files = getMarkdownFiles(markdown);
expect(files['content.md']).toBe(markdown);
expect(files['index.html']).toContain('- Item 1');
expect(files['index.html']).toContain('Subitem 1');
});
});
describe('edge cases', () => {
it('should handle very long markdown content', () => {
const longMarkdown = '# Test\n\n' + 'Lorem ipsum '.repeat(1000);
const files = getMarkdownFiles(longMarkdown);
expect(files['content.md']).toBe(longMarkdown);
expect(files['index.html']).toContain('Lorem ipsum');
});
it('should handle markdown with special characters', () => {
const markdown = '# Test & < > " \'';
const files = getMarkdownFiles(markdown);
expect(files['content.md']).toBe(markdown);
});
it('should handle markdown with unicode characters', () => {
const markdown = '# 你好 世界 🌍';
const files = getMarkdownFiles(markdown);
expect(files['content.md']).toBe(markdown);
});
it('should handle markdown with only whitespace', () => {
const markdown = ' \n\n ';
const files = getMarkdownFiles(markdown);
expect(files['content.md']).toBe(markdown);
});
it('should handle markdown with mixed line endings', () => {
const markdown = '# Line 1\r\n## Line 2\n### Line 3';
const files = getMarkdownFiles(markdown);
expect(files['content.md']).toBe(markdown);
});
});
});
describe('static HTML structure', () => {
it('should generate a complete HTML document with marked.js', () => {
const files = getMarkdownFiles('# Test');
const html = files['index.html'];
expect(html).toContain('<!DOCTYPE html>');
expect(html).toContain('<html lang="en">');
expect(html).toContain('<title>Markdown Preview</title>');
expect(html).toContain('marked.min.js');
expect(html).toContain('marked.use(');
expect(html).toContain('marked.parse(');
});
it('should pin the CDN script to an exact version with SRI', () => {
const files = getMarkdownFiles('# Test');
const html = files['index.html'];
expect(html).toMatch(/marked@\d+\.\d+\.\d+/);
expect(html).toContain('integrity="sha384-');
expect(html).toContain('crossorigin="anonymous"');
});
it('should show an error message when marked fails to load', () => {
const files = getMarkdownFiles('# Test');
const html = files['index.html'];
expect(html).toContain("typeof marked === 'undefined'");
expect(html).toContain('failed to load');
expect(html).toContain('class="markdown-error"');
});
it('should strip raw HTML blocks via renderer override', () => {
const files = getMarkdownFiles('# Test');
const html = files['index.html'];
expect(html).toContain("html() { return ''; }");
});
it('should embed isSafeUrl logic in the HTML for link sanitization', () => {
const files = getMarkdownFiles('# Test');
const html = files['index.html'];
expect(html).toContain("new Set(['http:', 'https:', 'mailto:', 'tel:'])");
expect(html).toContain('isSafeUrl');
expect(html).toContain("trimmed.startsWith('/')");
expect(html).toContain("trimmed.startsWith('#')");
expect(html).toContain("trimmed.startsWith('.')");
});
it('should configure marked with GFM and line-break support', () => {
const files = getMarkdownFiles('# Test');
const html = files['index.html'];
expect(html).toContain('gfm: true');
expect(html).toContain('breaks: true');
});
it('should configure a custom renderer for link/image sanitization', () => {
const files = getMarkdownFiles('# Test');
const html = files['index.html'];
expect(html).toContain('renderer:');
expect(html).toContain('link(token)');
expect(html).toContain('image(token)');
});
it('should embed the markdown content in the HTML', () => {
const testContent = '# Heading\n- List item';
const files = getMarkdownFiles(testContent);
const html = files['index.html'];
expect(html).toContain('# Heading');
expect(html).toContain('- List item');
});
});
});