build-cran-binaries/stats/classify.test.ts

import { describe, expect, test } from 'bun:test';
import { classify, clientPlatform, parseNetworks, toEvent, tool, type LogEntry } from './classify';

const R_RHEL = 'R/4.5.3 R (4.5.3 x86_64-pc-linux-gnu x86_64 linux-gnu)';
const RENV_RHEL = 'renv (1.2.3); R/4.5.3 R (4.5.3 x86_64-pc-linux-gnu x86_64 linux-gnu)';
const R_ALPINE = 'R/4.6.1 (Alpine Linux 3.24.1) (aarch64-unknown-linux-musl aarch64 linux-musl)';
const R_43 = 'R (4.3.2 x86_64-pc-linux-gnu x86_64 linux-gnu)';

const base = { status: 200, user_agent: R_RHEL, asn: 3303, asn_org: 'Swisscom', own: false };

describe('classify', () => {
  test('flat tarball is a download with package, version and slot', () => {
    const d = classify({ ...base, path: '/amd64/rhel9/latest/src/contrib/data.table_1.17.8.tar.gz' });
    expect(d).toMatchObject({
      kind: 'download',
      outcome: 'served',
      slot: 'amd64/rhel9',
      arch: 'amd64',
      os: 'rhel9',
      package: 'data.table',
      version: '1.17.8',
      r_minor: '4.5',
      r_version: '4.5.3',
      tool: 'R',
      network: 'other',
    });
  });

  test('per-minor tarball and index keep the slot', () => {
    expect(classify({ ...base, path: '/arm64/alpine324/latest/src/contrib/4.6/Rcpp_1.1.0.tar.gz' })).toMatchObject({
      kind: 'download',
      slot: 'arm64/alpine324',
      package: 'Rcpp',
    });
    expect(classify({ ...base, path: '/amd64/rhel9/latest/src/contrib/4.5/PACKAGES.rds' }).kind).toBe('index');
  });

  test('index redirect with a known minor is internal minor routing', () => {
    const d = classify({ ...base, status: 302, path: '/amd64/rhel9/latest/src/contrib/PACKAGES.rds' });
    expect(d).toMatchObject({ kind: 'index', outcome: 'redirect', redirect_reason: 'minor_routing' });
  });

  test('redirect for an R minor we do not build is unsupported_r_minor', () => {
    const d = classify({ ...base, user_agent: R_43, status: 302, path: '/amd64/noble/latest/src/contrib/PACKAGES' });
    expect(d).toMatchObject({ redirect_reason: 'unsupported_r_minor', r_minor: '4.3' });
  });

  test('archived version without a binary falls back to CRAN', () => {
    const d = classify({
      ...base,
      status: 302,
      path: '/amd64/rhel9/latest/src/contrib/Archive/renv/renv_1.2.3.tar.gz',
    });
    expect(d).toMatchObject({ kind: 'download', redirect_reason: 'archive_fallback', package: 'renv' });
  });

  test('bare URL: identified distro resolves to a slot, stock R UA goes to CRAN', () => {
    expect(
      classify({ ...base, user_agent: R_ALPINE, status: 302, path: '/src/contrib/PACKAGES.rds' }).redirect_reason,
    ).toBe('slot_resolution');
    expect(classify({ ...base, status: 302, path: '/src/contrib/PACKAGES.rds' }).redirect_reason).toBe(
      'unknown_distro',
    );
  });

  test('renv and pak probing for Posit Package Manager is a probe', () => {
    expect(classify({ ...base, status: 404, path: '/amd64/__api__/status' }).kind).toBe('probe');
  });

  test('scanner paths are noise', () => {
    for (const path of ['/.env', '/wp-content/plugins/x.php', '/.git/config', '/xwpg.php']) {
      expect(classify({ ...base, status: 404, user_agent: 'Mozilla/5.0', path }).kind).toBe('noise');
    }
  });

  test('the stats file is not counted as repository use', () => {
    expect(classify({ ...base, path: '/stats/downloads.json' }).kind).toBe('stats');
  });

  test('cloud networks are labelled, own networks win over cloud', () => {
    expect(classify({ ...base, asn: 8075, asn_org: 'Microsoft Corporation', path: '/' }).network).toBe('cloud');
    expect(classify({ ...base, asn: null, asn_org: 'Some Hosting Ltd', path: '/' }).network).toBe('cloud');
    expect(classify({ ...base, asn: 24940, own: true, path: '/' }).network).toBe('own');
  });

  test('query strings and doubled slashes do not change the result', () => {
    expect(classify({ ...base, path: '/amd64//rhel9/latest/src/contrib/PACKAGES.rds?x=1' }).kind).toBe('index');
  });
});

describe('client identity', () => {
  test('renv and plain R from the same machine share one platform key', () => {
    expect(clientPlatform(RENV_RHEL)).toBe(clientPlatform(R_RHEL));
    expect(clientPlatform(R_ALPINE)).not.toBe(clientPlatform(R_RHEL));
  });

  test('tool detection', () => {
    expect(tool(RENV_RHEL)).toBe('renv');
    expect(tool(R_RHEL)).toBe('R');
    expect(tool('pak/0.9.0 R/4.5.1 (...)')).toBe('pak');
    expect(tool('curl/8.5.0')).toBe('curl');
    expect(tool('-')).toBe('none');
    expect(tool('Mozilla/5.0 (Windows NT 10.0)')).toBe('browser');
  });

  test('own networks match by prefix for IPv4 and IPv6', () => {
    const own = parseNetworks('203.0.113.0/24, 2001:db8:1:a100::/64');
    const matches = (ip: string) => own.some((m) => m(ip));
    expect(matches('203.0.113.0')).toBe(true);
    expect(matches('203.0.114.0')).toBe(false);
    expect(matches('2001:db8:1:a100::')).toBe(true);
    expect(matches('2001:db8:1:a101::')).toBe(false);
    expect(parseNetworks('')).toHaveLength(0);
  });

  test('toEvent hashes the client and never keeps the IP', () => {
    const entry: LogEntry = {
      timestamp: '2026-11-14T11:45:33.689+00:00',
      pullZoneId: 3857050,
      requestId: 'abc',
      cacheStatus: 'HIT',
      statusCode: 200,
      bytesSent: 1000,
      remoteIp: '203.0.113.0',
      countryCode: 'CH',
      host: 'cran.rpkgs.com',
      path: '/amd64/rhel9/latest/src/contrib/rlang_1.1.6.tar.gz',
      userAgent: RENV_RHEL,
      asn: 3303,
      asnOrganization: 'Swisscom',
    };
    const event = toEvent(entry, (key) => `h(${key.length})`, parseNetworks('203.0.113.0/24'));
    expect(event).not.toBeNull();
    expect(JSON.stringify(event)).not.toContain('203.0.113');
    expect(event!.own).toBe(true);
    expect(event!.network).toBe('own');
    expect(toEvent({ ...entry, requestId: null }, (k) => k, [])).toBeNull();
  });
});