import fs from 'node:fs' import path from 'node:path' import { execFileSync } from 'node:child_process' const root = process.cwd() const files = [...new Set(execFileSync('git', ['ls-files', '--cached', '--others', '--exclude-standard', '-z'], { encoding: 'utf8' }).split('\0'))].filter(file => file.endsWith('.md')) const stripCode = text => text.replace(/^```[^\n]*\n[\s\S]*?^```[^\n]*$/gm, '') const slug = text => text.toLowerCase().replace(/`/g, '').replace(/[^\p{Letter}\p{Number}\p{Mark}_\- ]/gu, '').replace(/ /g, '-') const anchors = file => { const source = stripCode(fs.readFileSync(file, 'utf8')) const used = new Map(), result = new Set() for (const [, heading] of source.matchAll(/^#{1,6}\s+(.+)$/gm)) { const name = slug(heading.trim()), count = used.get(name) ?? 0 result.add(count ? `${name}-${count}` : name); used.set(name, count + 1) } for (const [, name] of source.matchAll(/\b(?:id|name)=["']([^"']+)["']/g)) result.add(name) return result } const problems = []; let internalLinks = 0 for (const file of files) for (const [, raw] of stripCode(fs.readFileSync(file, 'utf8')).matchAll(/!?\[[^\]\n]*\]\(([^)\n]+)\)/g)) { const href = raw.trim().replace(/^<|>$/g, '') if (/^(?:[a-z][a-z\d+.-]*:|\/\/)/i.test(href)) continue internalLinks++ try { const [target, fragment] = href.split('#') const resolved = target ? path.resolve(path.dirname(file), decodeURIComponent(target)) : path.resolve(file) const relative = path.relative(root, resolved).replaceAll('\\', '/') let reason = relative.startsWith('_local/') || relative === '_local' ? 'local evidence is excluded from Git Clone' : !fs.existsSync(resolved) ? 'missing path' : null if (!reason && fragment && resolved.endsWith('.md') && !anchors(resolved).has(decodeURIComponent(fragment))) reason = 'missing heading' if (reason) problems.push({ file, href, reason }) } catch (error) { problems.push({ file, href, reason: error.message }) } } console.log(JSON.stringify({ files: files.length, internalLinks, problems }, null, 2)) if (problems.length) process.exitCode = 1