theseus/bundled-addons/pithos/core/s3.js

// A small S3 client for s3d: AWS SigV4 (path-style, us-east-1, UNSIGNED-PAYLOAD
// for bodies so uploads stream straight through) plus the handful of bucket
// and object calls the console needs.

import crypto from 'node:crypto';
import http from 'node:http';
import { Readable } from 'node:stream';

const REGION = 'us-east-1';
const SERVICE = 's3';

const sha256 = (s) => crypto.createHash('sha256').update(s).digest('hex');
const hmac = (k, s) => crypto.createHmac('sha256', k).update(s).digest();

// RFC 3986 encoding as SigV4 wants it; '/' kept for object keys in paths.
function uriEncode(str, keepSlash) {
  return encodeURIComponent(str)
    .replace(/[!'()*]/g, (c) => '%' + c.charCodeAt(0).toString(16).toUpperCase())
    .replace(keepSlash ? /%2F/g : /$^/, '/');
}

function canonicalQuery(query) {
  return Object.keys(query).sort()
    .map((k) => `${uriEncode(k)}=${uriEncode(String(query[k] ?? ''))}`)
    .join('&');
}

function amzDate(d = new Date()) {
  return d.toISOString().replace(/[-:]/g, '').replace(/\.\d{3}/, '');
}

function signingKey(secret, date) {
  return hmac(hmac(hmac(hmac('AWS4' + secret, date), REGION), SERVICE), 'aws4_request');
}

export function objectPath(bucket, key) {
  return '/' + uriEncode(bucket) + (key != null ? '/' + uriEncode(key, true) : '');
}

export class S3Client {
  // endpoint: 'http://127.0.0.1:8000'
  constructor({ endpoint, accessKeyId, secretKey }) {
    this.endpoint = new URL(endpoint);
    this.accessKeyId = accessKeyId;
    this.secretKey = secretKey;
  }

  sign(method, path, query, headers, { now = amzDate(), payloadHash = 'UNSIGNED-PAYLOAD' } = {}) {
    const date = now.slice(0, 8);
    const h = { ...headers, host: this.endpoint.host, 'x-amz-date': now, 'x-amz-content-sha256': payloadHash };
    const names = Object.keys(h).map((k) => k.toLowerCase()).sort();
    const lower = Object.fromEntries(Object.entries(h).map(([k, v]) => [k.toLowerCase(), String(v).trim()]));
    const signed = names.join(';');
    const canonical = [method, path, canonicalQuery(query), names.map((n) => `${n}:${lower[n]}\n`).join(''), signed, payloadHash].join('\n');
    const scope = `${date}/${REGION}/${SERVICE}/aws4_request`;
    const toSign = ['AWS4-HMAC-SHA256', now, scope, sha256(canonical)].join('\n');
    const sig = crypto.createHmac('sha256', signingKey(this.secretKey, date)).update(toSign).digest('hex');
    lower.authorization = `AWS4-HMAC-SHA256 Credential=${this.accessKeyId}/${scope}, SignedHeaders=${signed}, Signature=${sig}`;
    delete lower.host;
    return lower;
  }

  // Presigned GET for share links. publicEndpoint lets the link name the
  // address readers will use, which may differ from the loopback the GUI uses.
  presign(bucket, key, expiresSec = 3600, publicEndpoint, now = amzDate()) {
    const base = publicEndpoint ? new URL(publicEndpoint) : this.endpoint;
    const date = now.slice(0, 8);
    const scope = `${date}/${REGION}/${SERVICE}/aws4_request`;
    const path = objectPath(bucket, key);
    const query = {
      'X-Amz-Algorithm': 'AWS4-HMAC-SHA256',
      'X-Amz-Credential': `${this.accessKeyId}/${scope}`,
      'X-Amz-Date': now,
      'X-Amz-Expires': String(Math.min(Math.max(1, expiresSec | 0), 604800)),
      'X-Amz-SignedHeaders': 'host',
    };
    const canonical = ['GET', path, canonicalQuery(query), `host:${base.host}\n`, 'host', 'UNSIGNED-PAYLOAD'].join('\n');
    const toSign = ['AWS4-HMAC-SHA256', now, scope, sha256(canonical)].join('\n');
    query['X-Amz-Signature'] = crypto.createHmac('sha256', signingKey(this.secretKey, date)).update(toSign).digest('hex');
    return `${base.origin}${path}?${canonicalQuery(query)}`;
  }

  // Low-level request. body: Buffer | string | Readable | undefined.
  // Resolves the raw http.IncomingMessage so callers can stream it.
  request(method, path, { query = {}, headers = {}, body } = {}) {
    const signedHeaders = this.sign(method, path, query, headers);
    const qs = canonicalQuery(query);
    return new Promise((resolve, reject) => {
      const req = http.request({
        protocol: this.endpoint.protocol,
        hostname: this.endpoint.hostname,
        port: this.endpoint.port,
        method,
        path: path + (qs ? '?' + qs : ''),
        headers: signedHeaders,
      }, resolve);
      req.on('error', reject);
      if (body instanceof Readable) body.pipe(req);
      else req.end(body);
    });
  }

  async call(method, path, opts) {
    const res = await this.request(method, path, opts);
    const text = await readBody(res);
    if (res.statusCode >= 300) throw s3Error(res.statusCode, text);
    return { status: res.statusCode, headers: res.headers, text };
  }

  async listBuckets() {
    const { text } = await this.call('GET', '/');
    return blocks(text, 'Bucket').map((b) => ({ name: tag(b, 'Name'), created: tag(b, 'CreationDate') }));
  }

  createBucket(bucket) { return this.call('PUT', objectPath(bucket)); }
  deleteBucket(bucket) { return this.call('DELETE', objectPath(bucket)); }

  async listObjects(bucket, { prefix = '', delimiter = '/', token, max = 1000 } = {}) {
    const query = { 'list-type': '2', prefix, delimiter, 'max-keys': String(max) };
    if (token) query['continuation-token'] = token;
    const { text } = await this.call('GET', objectPath(bucket), { query });
    return {
      folders: blocks(text, 'CommonPrefixes').map((b) => tag(b, 'Prefix')),
      objects: blocks(text, 'Contents').map((b) => ({
        key: tag(b, 'Key'), size: Number(tag(b, 'Size')), modified: tag(b, 'LastModified'), etag: tag(b, 'ETag'),
      })),
      truncated: tag(text, 'IsTruncated') === 'true',
      nextToken: tag(text, 'NextContinuationToken'),
    };
  }

  async deleteObject(bucket, key) { return this.call('DELETE', objectPath(bucket, key)); }

  // Deletes every key under prefix. Returns the number deleted.
  async deletePrefix(bucket, prefix) {
    let n = 0;
    let token;
    do {
      const page = await this.listObjects(bucket, { prefix, delimiter: '', token });
      for (const o of page.objects) { await this.deleteObject(bucket, o.key); n++; }
      token = page.truncated ? page.nextToken : null;
    } while (token);
    return n;
  }

  putObject(bucket, key, body, { contentType, contentLength } = {}) {
    const headers = {};
    if (contentType) headers['content-type'] = contentType;
    if (contentLength != null) headers['content-length'] = String(contentLength);
    return this.call('PUT', objectPath(bucket, key), { headers, body });
  }

  getObject(bucket, key, { range } = {}) {
    return this.request('GET', objectPath(bucket, key), { headers: range ? { range } : {} });
  }

  async getPolicy(bucket) {
    const res = await this.request('GET', objectPath(bucket), { query: { policy: '' } });
    const text = await readBody(res);
    if (res.statusCode === 404) return null;
    if (res.statusCode >= 300) throw s3Error(res.statusCode, text);
    return JSON.parse(text);
  }

  putPolicy(bucket, policy) {
    const body = JSON.stringify(policy);
    return this.call('PUT', objectPath(bucket), { query: { policy: '' }, body, headers: { 'content-type': 'application/json', 'content-length': Buffer.byteLength(body) } });
  }

  deletePolicy(bucket) { return this.call('DELETE', objectPath(bucket), { query: { policy: '' } }); }
}

// The only policy shape s3d accepts: anonymous read of objects and listing.
export function publicReadPolicy(bucket, { list = false } = {}) {
  const statements = [{
    Effect: 'Allow', Principal: '*', Action: ['s3:GetObject'], Resource: [`arn:aws:s3:::${bucket}/*`],
  }];
  if (list) statements.push({ Effect: 'Allow', Principal: '*', Action: ['s3:ListBucket'], Resource: [`arn:aws:s3:::${bucket}`] });
  return { Version: '2012-10-17', Statement: statements };
}

export function readBody(res) {
  return new Promise((resolve, reject) => {
    const chunks = [];
    res.on('data', (c) => chunks.push(c));
    res.on('end', () => resolve(Buffer.concat(chunks).toString('utf8')));
    res.on('error', reject);
  });
}

export class S3Error extends Error {
  constructor(status, code, message) { super(message || code); this.status = status; this.code = code; }
}

function s3Error(status, text) {
  return new S3Error(status, tag(text, 'Code') || `HTTP ${status}`, tag(text, 'Message') || text.slice(0, 200) || `HTTP ${status}`);
}

// S3 responses are flat enough that tag extraction is all we need.
function blocks(xml, name) {
  const re = new RegExp(`<${name}>([\\s\\S]*?)</${name}>`, 'g');
  const out = [];
  let m;
  while ((m = re.exec(xml))) out.push(m[1]);
  return out;
}

function tag(xml, name) {
  const m = xml.match(new RegExp(`<${name}>([\\s\\S]*?)</${name}>`));
  return m ? unescapeXml(m[1]) : null;
}

function unescapeXml(s) {
  return s.replace(/&(lt|gt|amp|quot|apos|#\d+|#x[0-9a-f]+);/gi, (_, e) => {
    switch (e.toLowerCase()) {
      case 'lt': return '<'; case 'gt': return '>'; case 'amp': return '&';
      case 'quot': return '"'; case 'apos': return "'";
      default: return String.fromCodePoint(e[1].toLowerCase() === 'x' ? parseInt(e.slice(2), 16) : parseInt(e.slice(1), 10));
    }
  });
}