@just-every/mcp-read-website-fast
Version:
Markdown Content Preprocessor - Fetch web pages, extract content, convert to clean Markdown
104 lines (103 loc) • 3.98 kB
JavaScript
import { Command } from 'commander';
import { fetchMarkdown } from './internal/fetchMarkdown.js';
import { secureCrawl } from './internal/secureCrawl.js';
import { readFileSync } from 'fs';
import { fileURLToPath } from 'url';
import { dirname, join } from 'path';
const __filename = fileURLToPath(import.meta.url);
const __dirname = dirname(__filename);
const packageJson = JSON.parse(readFileSync(join(__dirname, '../package.json'), 'utf-8'));
const program = new Command();
program
.name('mcp')
.description('Markdown Content Preprocessor - Extract and convert web content to clean Markdown')
.version(packageJson.version);
program
.command('fetch <url>')
.description('Fetch a URL and convert to Markdown')
.option('-p, --pages <number>', 'Maximum number of pages to crawl', '1')
.option('-c, --concurrency <number>', 'Max concurrent requests', '3')
.option('--no-robots', 'Ignore robots.txt')
.option('--all-origins', 'Allow cross-origin crawling')
.option('-u, --user-agent <string>', 'Custom user agent')
.option('--cache-dir <path>', 'Cache directory', '.cache')
.option('-t, --timeout <ms>', 'Request timeout in milliseconds', '30000')
.option('--cookies-file <path>', 'Path to Netscape cookie file for authenticated pages')
.option('-o, --output <format>', 'Output format: json, markdown, or both', 'markdown')
.action(async (url, options) => {
try {
const pages = parseInt(options.pages, 10);
const depth = pages > 1 ? 1 : 0;
const crawlOptions = {
depth: depth,
maxConcurrency: parseInt(options.concurrency, 10),
respectRobots: options.robots,
sameOriginOnly: !options.allOrigins,
userAgent: options.userAgent,
cacheDir: options.cacheDir,
timeout: parseInt(options.timeout, 10),
};
if (options.cookiesFile) {
crawlOptions.cookiesFile = options.cookiesFile;
}
console.error(`Fetching ${url}...`);
if (options.output === 'json') {
const results = await secureCrawl(url, crawlOptions);
console.log(JSON.stringify(results, null, 2));
}
else if (options.output === 'markdown') {
const result = await fetchMarkdown(url, {
...crawlOptions,
maxPages: pages,
});
if (result.markdown) {
console.log(result.markdown);
}
if (result.error) {
console.error(`Error: ${result.error}`);
if (!result.markdown) {
process.exit(1);
}
}
}
else if (options.output === 'both') {
const results = await secureCrawl(url, crawlOptions);
results.forEach((result) => {
console.log(`\n## URL: ${result.url}\n`);
if (result.markdown) {
console.log(result.markdown);
}
if (result.error) {
console.error(`${result.markdown ? 'Warning' : 'Error'}: ${result.error}`);
}
});
}
}
catch (error) {
console.error('Error:', error instanceof Error ? error.message : error);
process.exit(1);
}
});
program
.command('clear-cache')
.description('Clear the cache directory')
.option('--cache-dir <path>', 'Cache directory', '.cache')
.action(async (options) => {
try {
const { rm } = await import('fs/promises');
await rm(options.cacheDir, { recursive: true, force: true });
console.log(`Cache cleared: ${options.cacheDir}`);
}
catch (error) {
console.error('Error clearing cache:', error);
process.exit(1);
}
});
program
.command('serve')
.description('Run as an MCP server')
.action(async () => {
await import('./serve.js');
});
program.parse();