feat(edge): route PACKAGES requests to the per-R-minor slot

- add edge/rpkgs-router.ts, which redirects PACKAGES, PACKAGES.gz and PACKAGES.rds into …/src/contrib/<x.y>/ for slots listed in UNION_SLOTS - leave tarballs alone: R keeps the pre-redirect contrib URL, so the union index steers per-minor tarballs with a Path field instead - stop resolving an unidentifiable distro to a phantom linux-gnu/linux-musl slot and send those clients to CRAN - mark every redirect no-store, since the target depends on the User-Agent - cover the routing matrix in edge/rpkgs-router.test.ts, run by just edge-test - correct the spec and plan: the union index carries Path: <x.y> on per-minor records, not Path: .. on flat ones
2026-08-07 14:00:14 +00:00 · 2026-08-07 14:00:14 +00:00 · 628d19d650
commit 628d19d650
parent 5a0a4fa200
2 changed files with 440 additions and 22 deletions
--- a/edge/rpkgs-router.test.ts
+++ b/edge/rpkgs-router.test.ts
@ -0,0 +1,160 @@
+/**
+ * Routing matrix for `edge/rpkgs-router.ts`.
+ *
+ * The script is exercised through the SDK's local server rather than by
+ * importing its internals, so what is tested is the artifact that gets
+ * deployed. Requests that the script passes through are proxied to the real
+ * origin, which keeps the "no redirect" cases honest: they assert that the
+ * client reached the flat slot, not merely that no `Location` was set.
+ *
+ * Run with `just edge-test`.
+ */
+import { assertEquals } from 'jsr:@std/assert@1';
+
+const SCRIPT = new URL('./rpkgs-router.ts', import.meta.url).pathname;
+const BASE = 'http://127.0.0.1:8080';
+const UNION_SLOTS = 'amd64/alpine324';
+
+const UA_R45_MUSL = 'R (4.5.3 x86_64-pc-linux-musl x86_64 linux-musl)';
+const UA_R46_MUSL = 'R (4.6.0 x86_64-pc-linux-musl x86_64 linux-musl)';
+const UA_R45_ALPINE = 'R/4.5.3 R (4.5.3 x86_64-pc-linux-musl x86_64 linux-musl) Alpine Linux 3.24';
+const UA_R45_DARWIN = 'R (4.5.1 aarch64-apple-darwin20 aarch64 darwin20)';
+const UA_CURL = 'curl/8.0.1';
+
+const SLOT = '/amd64/alpine324/latest/src/contrib';
+const OTHER_SLOT = '/amd64/noble/latest/src/contrib';
+
+interface Probe {
+  status: number;
+  location: string | null;
+  cacheControl: string | null;
+}
+
+async function probe(path: string, userAgent: string): Promise<Probe> {
+  const res = await fetch(BASE + path, {
+    headers: { 'User-Agent': userAgent },
+    redirect: 'manual',
+  });
+  await res.body?.cancel();
+  return {
+    status: res.status,
+    location: res.headers.get('location'),
+    cacheControl: res.headers.get('cache-control'),
+  };
+}
+
+/** Kill tolerantly: the child has already exited if the script failed to load. */
+async function stopServer(child: Deno.ChildProcess): Promise<void> {
+  try {
+    child.kill();
+  } catch {
+    // already gone
+  }
+  await child.status;
+}
+
+async function startServer(): Promise<Deno.ChildProcess> {
+  const child = new Deno.Command(Deno.execPath(), {
+    args: ['run', '-A', SCRIPT],
+    env: { UNION_SLOTS },
+    stdout: 'null',
+    stderr: 'inherit',
+  }).spawn();
+
+  for (let attempt = 0; attempt < 150; attempt++) {
+    try {
+      const res = await fetch(`${BASE}/`, {
+        headers: { 'User-Agent': UA_CURL },
+        redirect: 'manual',
+      });
+      await res.body?.cancel();
+      return child;
+    } catch {
+      await new Promise((resolve) => setTimeout(resolve, 200));
+    }
+  }
+
+  await stopServer(child);
+  throw new Error('edge script did not start listening on ' + BASE);
+}
+
+Deno.test('rpkgs-router', async (t) => {
+  const server = await startServer();
+
+  try {
+    await t.step("routes an index request to the client's R minor", async () => {
+      const res = await probe(`${SLOT}/PACKAGES.gz`, UA_R45_MUSL);
+      assertEquals(res.status, 302);
+      assertEquals(res.location, `https://cran.rpkgs.com${SLOT}/4.5/PACKAGES.gz`);
+    });
+
+    await t.step('routes R 4.6 to its own slot', async () => {
+      const res = await probe(`${SLOT}/PACKAGES.gz`, UA_R46_MUSL);
+      assertEquals(res.location, `https://cran.rpkgs.com${SLOT}/4.6/PACKAGES.gz`);
+    });
+
+    await t.step('routes PACKAGES and PACKAGES.rds too', async () => {
+      for (const file of ['PACKAGES', 'PACKAGES.rds']) {
+        const res = await probe(`${SLOT}/${file}`, UA_R45_MUSL);
+        assertEquals(res.location, `https://cran.rpkgs.com${SLOT}/4.5/${file}`, `expected ${file} to be routed`);
+      }
+    });
+
+    await t.step('marks the redirect uncacheable', async () => {
+      const res = await probe(`${SLOT}/PACKAGES.gz`, UA_R45_MUSL);
+      assertEquals(res.cacheControl, 'no-store');
+    });
+
+    await t.step('leaves a slot outside UNION_SLOTS alone', async () => {
+      const res = await probe(`${OTHER_SLOT}/PACKAGES.gz`, UA_R45_MUSL);
+      assertEquals(res.location, null);
+      assertEquals(res.status, 200);
+    });
+
+    await t.step('never routes a tarball', async () => {
+      const res = await probe(`${SLOT}/jsonlite_2.0.0.tar.gz`, UA_R45_MUSL);
+      assertEquals(res.location, null);
+      assertEquals(res.status, 200);
+    });
+
+    await t.step('does not redirect a path already under a minor', async () => {
+      const res = await probe(`${SLOT}/4.5/PACKAGES.gz`, UA_R45_MUSL);
+      assertEquals(res.location, null);
+      assertEquals(res.status, 200);
+    });
+
+    await t.step('leaves a client without an R version alone', async () => {
+      const res = await probe(`${SLOT}/PACKAGES.gz`, UA_CURL);
+      assertEquals(res.location, null);
+      assertEquals(res.status, 200);
+    });
+
+    await t.step('resolves the bare root to slot and minor', async () => {
+      const res = await probe('/src/contrib/PACKAGES.gz', UA_R45_ALPINE);
+      assertEquals(res.location, `https://cran.rpkgs.com${SLOT}/4.5/PACKAGES.gz`);
+    });
+
+    await t.step('sends an unidentifiable distro to CRAN', async () => {
+      const res = await probe('/src/contrib/PACKAGES.gz', UA_R45_MUSL);
+      assertEquals(res.location, 'https://cran.r-project.org/src/contrib/PACKAGES.gz');
+    });
+
+    await t.step('keeps the macOS rewrite', async () => {
+      const res = await probe('/src/contrib/foo_1.0.tar.gz', UA_R45_DARWIN);
+      assertEquals(res.location, 'https://cran.rpkgs.com/bin/macosx/big-sur-arm64/contrib/4.5/foo_1.0.tar.gz');
+    });
+
+    await t.step('keeps the macOS binary passthrough to CRAN', async () => {
+      const path = '/bin/macosx/big-sur-arm64/contrib/4.5/foo_1.0.tar.gz';
+      const res = await probe(path, UA_R45_DARWIN);
+      assertEquals(res.location, `https://cran.r-project.org${path}`);
+    });
+
+    await t.step('collapses duplicate slashes before matching', async () => {
+      const res = await probe(`/amd64/alpine324//latest/src/contrib//PACKAGES.gz`, UA_R45_MUSL);
+      assertEquals(res.location, `https://cran.rpkgs.com${SLOT}/4.5/PACKAGES.gz`);
+    });
+  } finally {
+    await stopServer(server);
+  }
+});
--- a/edge/rpkgs-router.ts
+++ b/edge/rpkgs-router.ts
@ -0,0 +1,235 @@
+/**
+ * Edge middleware for cran.rpkgs.com.
+ *
+ * Two jobs:
+ *
+ * 1. Resolve the bare `https://cran.rpkgs.com` form to a concrete
+ *    `<arch>/<os>` slot from the User-Agent, or send the client to CRAN when
+ *    the distro cannot be identified.
+ * 2. Route `PACKAGES*` requests to the per-R-minor slot
+ *    (`…/latest/src/contrib/<x.y>/`), so a stock `install.packages()` sees the
+ *    packages that only exist there.
+ *
+ * Only index files are routed. Tarballs are deliberately left alone: R keeps
+ * the contrib URL it asked for, not the one it was redirected to, so every
+ * tarball URL is resolved against the flat directory and the union index steers
+ * the per-minor ones with a `Path: <x.y>` field. Rewriting a tarball request
+ * here would send flat-slot packages into a directory that does not hold them.
+ *
+ * Routing is gated on UNION_SLOTS. The raw per-minor index holds only the
+ * ABI-sensitive subset of a slot; it is safe to route to it only once bincraft
+ * has republished it as a union of the per-minor and flat slots.
+ *
+ * Deployed by OpenTofu from this file (`bunnynet_compute_script.rpkgs_router`).
+ * Test with `just edge-test`.
+ */
+import * as BunnySDK from 'https://esm.sh/@bunny.net/edgescript-sdk@0.12';
+
+const PUBLIC_CDN_ORIGIN = 'https://cran.rpkgs.com';
+const CRAN_ORIGIN = 'https://cran.r-project.org';
+
+/** Slots ("<arch>/<os>", comma separated) whose per-minor index is a union. */
+const UNION_SLOTS = new Set(
+  (Deno.env.get('UNION_SLOTS') ?? '')
+    .split(',')
+    .map((slot) => slot.trim())
+    .filter((slot) => slot.length > 0),
+);
+
+/** `/<arch>/<os>/latest/src/contrib[/<rest>]` */
+const SLOT_PATH_REGEX = /^\/(amd64|arm64)\/([a-z0-9._-]+)\/latest\/src\/contrib\/?(.*)$/;
+
+/** A path that already sits in a per-minor slot, e.g. `4.5/PACKAGES.gz`. */
+const MINOR_DIR_REGEX = /^\d+\.\d+\//;
+
+/** The only files this script routes. */
+const INDEX_FILE_REGEX = /^PACKAGES(\.gz|\.rds)?$/;
+
+const SRC_CONTRIB_REGEX = /^\/src\/contrib\/(.+)$/;
+
+const MACOS_BIN_REGEX =
+  /^\/bin\/macosx\/(big-sur-arm64|big-sur-x86_64|monterey-arm64|monterey-x86_64)\/contrib\/([0-9.]+)\/(.+)$/;
+
+const RHEL_REGEX = /(almalinux|rocky)[^\d]*(\d+)/i;
+
+const UBUNTU_REGEX = /Ubuntu ([\d.]+)/i;
+const UBUNTU_CODENAMES: Record<string, string> = {
+  '24.04': 'noble',
+  '22.04': 'jammy',
+};
+
+const ALPINE_REGEX = /(?:Alpine Linux(?:\s+VERSION_ID=)?|alpine-)\s*(\d+)\.(\d+)/i;
+
+/**
+ * R's own User-Agent is `R (4.5.3 x86_64-pc-linux-musl …)`; the Posit-style one
+ * some sites configure is `R/4.5.3 R (…)`. Both carry the minor, which is why
+ * per-minor routing works without the distro being identifiable.
+ */
+const R_MINOR_REGEXES = [/\bR\/(\d+)\.(\d+)/, /\bR \((\d+)\.(\d+)/];
+
+function normalizePathname(pathname: string): string {
+  return pathname.replace(/\/{2,}/g, '/');
+}
+
+function redirectTo(location: string, status = 302): Response {
+  return new Response(null, {
+    status,
+    headers: {
+      Location: location,
+      // The target depends on the User-Agent, so the redirect itself must
+      // never be cached; only its target is a cacheable, UA-independent URL.
+      'Cache-Control': 'no-store',
+      'X-Via': 'MyMiddleware',
+      'X-Rewritten-By': 'rpkgs-edge-middleware',
+    },
+  });
+}
+
+function extractRMinor(userAgent: string): string | null {
+  for (const regex of R_MINOR_REGEXES) {
+    const match = userAgent.match(regex);
+    if (match) {
+      return `${match[1]}.${match[2]}`;
+    }
+  }
+  return null;
+}
+
+function mapArch(arch: string): string {
+  if (arch === 'aarch64') return 'arm64';
+  if (arch === 'x86_64') return 'amd64';
+  return arch;
+}
+
+function extractArch(userAgent: string): string {
+  const match = userAgent.match(/(x86_64|aarch64|arm64|i386|i686)/);
+  return match ? mapArch(match[1]) : '';
+}
+
+/**
+ * Identify the `<arch>/<os>` slot from the User-Agent, or null.
+ *
+ * A stock R User-Agent carries only `linux-gnu` / `linux-musl`, which are not
+ * slot names: returning them produced redirects into slots that do not exist
+ * (`/amd64/linux-musl/latest/…`, a guaranteed 404). An unidentifiable distro
+ * is reported as such so the caller can fall back to CRAN.
+ */
+function parseSlot(userAgent: string): string | null {
+  const arch = extractArch(userAgent);
+  if (!arch) {
+    return null;
+  }
+
+  const rhel = userAgent.match(RHEL_REGEX);
+  if (rhel) {
+    return `${arch}/rhel${rhel[2]}`;
+  }
+
+  const ubuntu = userAgent.match(UBUNTU_REGEX);
+  if (ubuntu) {
+    const codename = UBUNTU_CODENAMES[ubuntu[1]];
+    if (codename) {
+      return `${arch}/${codename}`;
+    }
+  }
+
+  const alpine = userAgent.match(ALPINE_REGEX);
+  if (alpine) {
+    return `${arch}/alpine${alpine[1]}${alpine[2]}`;
+  }
+
+  return null;
+}
+
+function parseMacUserAgent(userAgent: string): { os: string; arch: string; rver: string } | null {
+  const rverMatch = userAgent.match(/R \((\d+)\.(\d+)/);
+  const archMatch = userAgent.match(/(aarch64|arm64|x86_64)/);
+  const osMatch = userAgent.match(/darwin(\d+)/);
+
+  if (!rverMatch || !archMatch || !osMatch) {
+    return null;
+  }
+
+  const arch = archMatch[1] === 'aarch64' ? 'arm64' : archMatch[1];
+  const darwinVer = parseInt(osMatch[1], 10);
+  const os = darwinVer >= 21 && darwinVer < 22 ? `monterey-${arch}` : `big-sur-${arch}`;
+
+  return { os, arch, rver: `${rverMatch[1]}.${rverMatch[2]}` };
+}
+
+/**
+ * The contrib path a request should be served from, relative to the slot.
+ *
+ * Returns the per-minor path for an index file when the slot is known to carry
+ * a union index and the client's R minor is known; otherwise the flat path,
+ * which is what every client sees today.
+ */
+function contribPath(slot: string, rest: string, userAgent: string): string {
+  const flat = rest ? `/${slot}/latest/src/contrib/${rest}` : `/${slot}/latest/src/contrib`;
+
+  if (!INDEX_FILE_REGEX.test(rest) || !UNION_SLOTS.has(slot)) {
+    return flat;
+  }
+
+  const rMinor = extractRMinor(userAgent);
+  return rMinor ? `/${slot}/latest/src/contrib/${rMinor}/${rest}` : flat;
+}
+
+BunnySDK.net.http
+  .servePullZone({ url: 'https://cran.rpkgs.com/' })
+  .onOriginRequest((ctx) => {
+    const url = new URL(ctx.request.url);
+    const path = normalizePathname(url.pathname);
+    const userAgent = ctx.request.headers.get('User-Agent') || '';
+
+    // macOS clients are served from CRAN's own binary tree.
+    const srcContrib = path.match(SRC_CONTRIB_REGEX);
+    if (srcContrib && /darwin/.test(userAgent)) {
+      const mac = parseMacUserAgent(userAgent);
+      if (mac) {
+        return Promise.resolve(
+          redirectTo(`${PUBLIC_CDN_ORIGIN}/bin/macosx/${mac.os}/contrib/${mac.rver}/${srcContrib[1]}`),
+        );
+      }
+    }
+
+    if (MACOS_BIN_REGEX.test(path)) {
+      return Promise.resolve(redirectTo(`${CRAN_ORIGIN}${path}`));
+    }
+
+    // Already-qualified slot URLs: what the runtime images have baked in.
+    const slotPath = path.match(SLOT_PATH_REGEX);
+    if (slotPath) {
+      const slot = `${slotPath[1]}/${slotPath[2]}`;
+      const rest = slotPath[3];
+
+      // Never rewrite a request that is already in a per-minor slot, or the
+      // redirect would chase its own tail.
+      if (MINOR_DIR_REGEX.test(rest)) {
+        return Promise.resolve(ctx.request);
+      }
+
+      const target = contribPath(slot, rest, userAgent);
+      if (target === path) {
+        return Promise.resolve(ctx.request);
+      }
+      return Promise.resolve(redirectTo(`${PUBLIC_CDN_ORIGIN}${target}`));
+    }
+
+    // The bare `https://cran.rpkgs.com` form, resolved from the User-Agent.
+    if (path === '/' || path === '/src/contrib' || path.startsWith('/src/contrib/')) {
+      const slot = parseSlot(userAgent);
+      if (!slot) {
+        return Promise.resolve(redirectTo(`${CRAN_ORIGIN}${path}`));
+      }
+
+      const rest = srcContrib ? srcContrib[1] : '';
+      return Promise.resolve(redirectTo(`${PUBLIC_CDN_ORIGIN}${contribPath(slot, rest, userAgent)}`));
+    }
+
+    return Promise.resolve(ctx.request);
+  })
+  .onOriginResponse((ctx) => {
+    ctx.response.headers.append('X-Via', 'MyMiddleware');
+    return Promise.resolve(ctx.response);
+  });