Add SSRF URL validator for clean-pdf

src/lib/clean-pdf/__tests__/ssrf.test.ts

@@ -0,0 +1,58 @@
+import { describe, expect, it } from "vitest";
+import { BlockedUrlError, assertPublicUrl, isPrivateAddress } from "@/lib/clean-pdf/ssrf";
+
+describe("isPrivateAddress", () => {
+  const privateIps = [
+    "127.0.0.1", "0.0.0.0", "10.1.2.3", "100.64.0.1", "169.254.169.254",
+    "172.16.0.1", "172.31.255.255", "192.168.1.1", "192.0.0.8", "198.18.0.1",
+    "224.0.0.1", "255.255.255.255",
+    "::1", "::", "fc00::1", "fd12:3456::1", "fe80::1", "::ffff:10.0.0.1", "::ffff:127.0.0.1",
+  ];
+  const publicIps = ["93.184.216.34", "8.8.8.8", "172.32.0.1", "2606:2800:220:1::1", "::ffff:8.8.8.8"];
+
+  it.each(privateIps)("блокирует %s", (ip) => expect(isPrivateAddress(ip)).toBe(true));
+  it.each(publicIps)("пропускает %s", (ip) => expect(isPrivateAddress(ip)).toBe(false));
+});
+
+describe("assertPublicUrl", () => {
+  const publicResolve = async () => [{ address: "93.184.216.34", family: 4 }];
+  const privateResolve = async () => [{ address: "172.18.0.2", family: 4 }];
+  const mixedResolve = async () => [
+    { address: "93.184.216.34", family: 4 },
+    { address: "10.0.0.5", family: 4 },
+  ];
+
+  it("пропускает публичный https-URL", async () => {
+    const url = await assertPublicUrl("https://example.com/article", publicResolve);
+    expect(url.hostname).toBe("example.com");
+  });
+
+  it.each([
+    "file:///etc/passwd",
+    "ftp://example.com/x",
+    "chrome://settings",
+    "not a url",
+  ])("блокирует %s", async (raw) => {
+    await expect(assertPublicUrl(raw, publicResolve)).rejects.toThrow(BlockedUrlError);
+  });
+
+  it("блокирует localhost и .local без резолва", async () => {
+    await expect(assertPublicUrl("http://localhost:3000/x", publicResolve)).rejects.toThrow(BlockedUrlError);
+    await expect(assertPublicUrl("http://printer.local/x", publicResolve)).rejects.toThrow(BlockedUrlError);
+  });
+
+  it("блокирует литеральный приватный IP", async () => {
+    await expect(assertPublicUrl("http://192.168.1.1/admin", publicResolve)).rejects.toThrow(BlockedUrlError);
+    await expect(assertPublicUrl("http://[::1]:8080/", publicResolve)).rejects.toThrow(BlockedUrlError);
+  });
+
+  it("блокирует хост, резолвящийся в приватный IP (включая частично)", async () => {
+    await expect(assertPublicUrl("https://evil.example/x", privateResolve)).rejects.toThrow(BlockedUrlError);
+    await expect(assertPublicUrl("https://evil.example/x", mixedResolve)).rejects.toThrow(BlockedUrlError);
+  });
+
+  it("блокирует хост, который не резолвится", async () => {
+    const failResolve = async () => { throw new Error("ENOTFOUND"); };
+    await expect(assertPublicUrl("https://nope.example/x", failResolve)).rejects.toThrow(BlockedUrlError);
+  });
+});

src/lib/clean-pdf/ssrf.ts

@@ -0,0 +1,82 @@
+import { lookup } from "node:dns/promises";
+import { isIP } from "node:net";
+
+export class BlockedUrlError extends Error {}
+
+export type LookupFn = (
+  hostname: string,
+  opts: { all: true },
+) => Promise<{ address: string; family: number }[]>;
+
+function v4ToInt(ip: string): number {
+  return ip.split(".").reduce((acc, o) => acc * 256 + Number(o), 0);
+}
+
+// [начало включительно, конец включительно] в виде 32-битных чисел
+const V4_PRIVATE: Array<[number, number]> = [
+  ["0.0.0.0", "0.255.255.255"],       // "this network"
+  ["10.0.0.0", "10.255.255.255"],     // RFC1918
+  ["100.64.0.0", "100.127.255.255"],  // CGNAT
+  ["127.0.0.0", "127.255.255.255"],   // loopback
+  ["169.254.0.0", "169.254.255.255"], // link-local / cloud metadata
+  ["172.16.0.0", "172.31.255.255"],   // RFC1918 (docker-сети попадают сюда)
+  ["192.0.0.0", "192.0.0.255"],       // IETF protocol assignments
+  ["192.168.0.0", "192.168.255.255"], // RFC1918
+  ["198.18.0.0", "198.19.255.255"],   // benchmarking
+  ["224.0.0.0", "255.255.255.255"],   // multicast + reserved + broadcast
+].map(([a, b]) => [v4ToInt(a), v4ToInt(b)] as [number, number]);
+
+function isPrivateV4(ip: string): boolean {
+  const n = v4ToInt(ip);
+  return V4_PRIVATE.some(([lo, hi]) => n >= lo && n <= hi);
+}
+
+export function isPrivateAddress(ip: string): boolean {
+  if (isIP(ip) === 4) return isPrivateV4(ip);
+  if (isIP(ip) !== 6) return true; // не IP — не пропускаем
+
+  const lower = ip.toLowerCase();
+  // v4-mapped: ::ffff:10.0.0.1
+  const mapped = lower.match(/^::ffff:(\d+\.\d+\.\d+\.\d+)$/);
+  if (mapped) return isPrivateV4(mapped[1]);
+
+  if (lower === "::" || lower === "::1") return true;
+  // fc00::/7 (ULA), fe80::/10 (link-local)
+  const firstGroup = parseInt(lower.split(":")[0] || "0", 16);
+  if (firstGroup >= 0xfc00 && firstGroup <= 0xfdff) return true;
+  if (firstGroup >= 0xfe80 && firstGroup <= 0xfebf) return true;
+  return false;
+}
+
+export async function assertPublicUrl(raw: string, resolve: LookupFn = lookup): Promise<URL> {
+  let url: URL;
+  try {
+    url = new URL(raw);
+  } catch {
+    throw new BlockedUrlError("Некорректный URL");
+  }
+  if (url.protocol !== "http:" && url.protocol !== "https:") {
+    throw new BlockedUrlError("Поддерживаются только http и https");
+  }
+
+  const hostname = url.hostname.replace(/^\[|\]$/g, ""); // [::1] → ::1
+  if (hostname === "localhost" || hostname.endsWith(".local") || hostname.endsWith(".internal")) {
+    throw new BlockedUrlError("Адрес недоступен");
+  }
+
+  if (isIP(hostname)) {
+    if (isPrivateAddress(hostname)) throw new BlockedUrlError("Адрес недоступен");
+    return url;
+  }
+
+  let addresses: { address: string; family: number }[];
+  try {
+    addresses = await resolve(hostname, { all: true });
+  } catch {
+    throw new BlockedUrlError("Не удалось определить адрес сайта");
+  }
+  if (addresses.length === 0 || addresses.some((a) => isPrivateAddress(a.address))) {
+    throw new BlockedUrlError("Адрес недоступен");
+  }
+  return url;
+}