/** * Vite dev middleware exposing the local doocus tree index to the app. * * Reads `docs-output/index.json` (produced by process_tree.py) — a replica of * the whole local drive tree — plus the extracted `.doocus/content.md` * sidecars, and streams the ORIGINAL files by relative path (resolved against * the index's recorded root). All reads are local; nothing leaves the machine. * * Routes: * GET /api/tree the index.json tree * GET /api/detail?path= node + extracted content + meta + inline text * GET /api/original?path= stream the original file (native viewer / packaging) */ import type { Plugin } from 'vite' import type { ServerResponse } from 'node:http' import fs from 'node:fs' import fsp from 'node:fs/promises' import path from 'node:path' interface Options { outputDir: string } interface Node { path: string name: string ext: string family: string mode: 'extracted' | 'meeting' | 'link' bytes: number modified: string url: string | null out?: string } interface Index { root: string generatedAt: string counts: Record files: Node[] } // Text-native files: readable as-is, so the "original" doubles as its own text. const TEXT_EXTS = new Set(['md', 'markdown', 'txt', 'json', 'yaml', 'yml', 'csv', 'html', 'htm']) const MIME: Record = { '.pdf': 'application/pdf', '.png': 'image/png', '.jpg': 'image/jpeg', '.jpeg': 'image/jpeg', '.html': 'text/html', '.htm': 'text/html', '.md': 'text/markdown', '.txt': 'text/plain', '.csv': 'text/csv', '.json': 'application/json', '.yaml': 'text/yaml', '.yml': 'text/yaml', '.docx': 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', '.pptx': 'application/vnd.openxmlformats-officedocument.presentationml.presentation', '.xlsx': 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet', '.mp4': 'video/mp4', } export function doocusApi(opts: Options): Plugin { const outputDir = path.resolve(opts.outputDir) const indexPath = path.join(outputDir, 'index.json') function readIndex(): Index | null { try { return JSON.parse(fs.readFileSync(indexPath, 'utf-8')) } catch { return null } } /** Resolve a node's original file, rejecting traversal outside the index root. */ function originalAbs(index: Index, rel: string): string | null { const root = path.resolve(index.root) const abs = path.resolve(root, rel) if (abs !== root && !abs.startsWith(root + path.sep)) return null return abs } return { name: 'doocus-api', configureServer(server) { server.config.logger.info(`[doocus-api] serving tree from: ${indexPath}`) server.middlewares.use('/api', (req, res, next) => { const url = new URL(req.url ?? '/', 'http://localhost') const parts = url.pathname.split('/').filter(Boolean) handle(res, parts, url.searchParams, req.headers.range).catch((err) => { server.config.logger.error(`[doocus-api] ${String(err)}`) sendJson(res, 500, { error: String(err) }) }).then((handled) => { if (!handled) next() }) }) }, } async function handle(res: ServerResponse, parts: string[], q: URLSearchParams, range?: string): Promise { if (parts[0] === 'tree') { const index = readIndex() if (!index) { sendJson(res, 404, { error: `index.json not found in ${outputDir}` }); return true } sendJson(res, 200, index) return true } if (parts[0] === 'detail') { await getDetail(res, q.get('path') ?? '') return true } if (parts[0] === 'original') { await serveOriginal(res, q.get('path') ?? '', range) return true } // GET /api/meeting?path= meetus review data for a meeting // GET /api/meeting/frame?path=&file= stream a frame from its .meetus/ if (parts[0] === 'meeting' && parts[1] === 'frame') { serveMeetingFrame(res, q.get('path') ?? '', q.get('file') ?? '') return true } if (parts[0] === 'meeting') { await getMeeting(res, q.get('path') ?? '') return true } return false } /** The .meetus sidecar dir + the video stem for a meeting node's rel path. */ function meetusDirFor(rel: string): { dir: string; stem: string } { return { dir: path.join(outputDir, rel + '.meetus'), stem: path.parse(rel).name } } async function getMeeting(res: ServerResponse, rel: string): Promise { const { dir, stem } = meetusDirFor(rel) // Whisper segments (may be null-flagged in manifest but present on disk). const segments: Array<{ start: number; end: number; text: string; speaker: string | null }> = [] const whisper = path.join(dir, `${stem}.json`) if (fs.existsSync(whisper)) { try { const data = JSON.parse(await fsp.readFile(whisper, 'utf-8')) for (const s of (Array.isArray(data) ? data : data.segments ?? [])) { segments.push({ start: Number(s.start ?? 0), end: Number(s.end ?? s.start ?? 0), text: String(s.text ?? '').trim(), speaker: s.speaker ?? null, }) } } catch { /* none */ } } // Frame → time from enhanced.txt ([MM:SS] before each `Frame: frames/…`). const enhanced = path.join(dir, `${stem}_enhanced.txt`) const times = new Map() let enhancedAvailable = false if (fs.existsSync(enhanced)) { try { enhancedAvailable = true let last: number | null = null for (const line of (await fsp.readFile(enhanced, 'utf-8')).split('\n')) { const ts = line.match(/^\[(\d+):(\d+)\]/) if (ts) last = Number(ts[1]) * 60 + Number(ts[2]) const fm = line.match(/Frame:\s*\S*?([^/\\\s]+\.jpg)/i) if (fm && last != null) times.set(fm[1], last) } } catch { /* none */ } } const frames: Array<{ file: string; url: string; time: number | null }> = [] try { for (const file of (await fsp.readdir(path.join(dir, 'frames'))).filter((f) => f.toLowerCase().endsWith('.jpg')).sort()) { const fny = file.match(/_(\d+(?:\.\d+)?)s\.jpg$/i) frames.push({ file, url: `/api/meeting/frame?path=${encodeURIComponent(rel)}&file=${encodeURIComponent(file)}`, time: fny ? Number(fny[1]) : times.get(file) ?? null, }) } } catch { /* no frames */ } sendJson(res, 200, { path: rel, hasOutput: fs.existsSync(dir), segments, frames, enhancedAvailable, videoUrl: `/api/original?path=${encodeURIComponent(rel)}`, }) } function serveMeetingFrame(res: ServerResponse, rel: string, file: string): void { const { dir } = meetusDirFor(rel) const safe = path.basename(file) const full = path.join(dir, 'frames', safe) if (safe !== file || !fs.existsSync(full)) { sendJson(res, 404, { error: 'frame not found' }); return } res.setHeader('Content-Type', 'image/jpeg') res.setHeader('Cache-Control', 'no-cache') fs.createReadStream(full).pipe(res) } async function getDetail(res: ServerResponse, rel: string): Promise { const index = readIndex() if (!index) { sendJson(res, 404, { error: 'no index' }); return } const node = index.files.find((f) => f.path === rel) if (!node) { sendJson(res, 404, { error: 'node not found' }); return } let content: string | null = null // extracted search text (heavy formats) let meta: unknown = null let text: string | null = null // inline text for text-native originals if (node.mode === 'extracted' && node.out) { try { content = await fsp.readFile(path.join(outputDir, node.out, 'content.md'), 'utf-8') } catch { /* */ } try { meta = JSON.parse(await fsp.readFile(path.join(outputDir, node.out, 'meta.json'), 'utf-8')) } catch { /* */ } } else if (node.mode === 'link' && TEXT_EXTS.has(node.ext)) { const abs = originalAbs(index, rel) if (abs) { try { text = await fsp.readFile(abs, 'utf-8') } catch { /* */ } } } sendJson(res, 200, { node, content, meta, text, mime: MIME[path.extname(node.name).toLowerCase()] ?? 'application/octet-stream', originalUrl: `/api/original?path=${encodeURIComponent(rel)}`, }) } async function serveOriginal(res: ServerResponse, rel: string, range?: string): Promise { const index = readIndex() if (!index) { sendJson(res, 404, { error: 'no index' }); return } const abs = originalAbs(index, rel) if (!abs || !isFile(abs)) { sendJson(res, 404, { error: 'original not reachable' }); return } const mime = MIME[path.extname(abs).toLowerCase()] ?? 'application/octet-stream' const size = fs.statSync(abs).size res.setHeader('Content-Type', mime) res.setHeader('Accept-Ranges', 'bytes') // Range support — needed so large meeting videos can seek/stream. const m = range?.match(/bytes=(\d*)-(\d*)/) if (m) { const start = m[1] ? parseInt(m[1], 10) : 0 const end = m[2] ? parseInt(m[2], 10) : size - 1 if (start >= size || end >= size || start > end) { res.statusCode = 416 res.setHeader('Content-Range', `bytes */${size}`) res.end() return } res.statusCode = 206 res.setHeader('Content-Range', `bytes ${start}-${end}/${size}`) res.setHeader('Content-Length', String(end - start + 1)) fs.createReadStream(abs, { start, end }).pipe(res) return } res.setHeader('Content-Length', String(size)) fs.createReadStream(abs).pipe(res) } } function sendJson(res: ServerResponse, status: number, obj: unknown): void { const body = JSON.stringify(obj) res.statusCode = status res.setHeader('Content-Type', 'application/json') res.setHeader('Content-Length', Buffer.byteLength(body)) res.end(body) } function isFile(p: string): boolean { try { return fs.statSync(p).isFile() } catch { return false } }