Loopback fixes

__tests__/__snapshots__/index.spec.ts.snap

@@ -1,7 +1,15 @@
-// Jest Snapshot v1, https://goo.gl/fbAQLP
+// Jest Snapshot v1, https://jestjs.io/docs/snapshot-testing
 
 exports[`#getLinkPreview() no link in text should fail gracefully 1`] = `"link-preview-js did not receive a valid a url or text"`;
 
+exports[`#getLinkPreview() should block .internal hostnames 1`] = `"link-preview-js did not receive a valid a url or text"`;
+
+exports[`#getLinkPreview() should block .local hostnames 1`] = `"link-preview-js did not receive a valid a url or text"`;
+
+exports[`#getLinkPreview() should block nip.io wildcard hostnames 1`] = `"link-preview-js did not receive a valid a url or text"`;
+
+exports[`#getLinkPreview() should block sslip.io wildcard hostnames 1`] = `"link-preview-js did not receive a valid a url or text"`;
+
 exports[`#getLinkPreview() should handle empty strings gracefully 1`] = `"link-preview-js did not receive a valid url or text"`;
 
 exports[`#getLinkPreview() should handle malformed urls gracefully 1`] = `"link-preview-js did not receive a valid a url or text"`;

__tests__/index.spec.ts

@@ -1,6 +1,22 @@
 import { getLinkPreview, getPreviewFromContent } from "../index";
+import { CONSTANTS } from "../constants";
 import prefetchedResponse from "./sampleResponse.json";
 
+describe(`#REGEX_LOOPBACK`, () => {
+  it(`matches IPv6 loopback and local ranges`, () => {
+    expect(CONSTANTS.REGEX_LOOPBACK.test(`::1`)).toBe(true);
+    expect(CONSTANTS.REGEX_LOOPBACK.test(`::ffff:127.0.0.1`)).toBe(true);
+    expect(CONSTANTS.REGEX_LOOPBACK.test(`fc00::1`)).toBe(true);
+    expect(CONSTANTS.REGEX_LOOPBACK.test(`fd12:3456:789a::1`)).toBe(true);
+    expect(CONSTANTS.REGEX_LOOPBACK.test(`fe80::abcd`)).toBe(true);
+    expect(CONSTANTS.REGEX_LOOPBACK.test(`febf::abcd`)).toBe(true);
+  });
+
+  it(`does not match non-local IPv6 addresses`, () => {
+    expect(CONSTANTS.REGEX_LOOPBACK.test(`2001:4860:4860::8888`)).toBe(false);
+  });
+});
+
 describe(`#getLinkPreview()`, () => {
   it(`should extract link info from just URL`, async () => {
     const linkInfo: any = await getLinkPreview(`https://www.youtube.com/watch?v=wuClZjOdT30`, {
@@ -150,6 +166,32 @@ describe(`#getLinkPreview()`, () => {
     ).rejects.toThrowErrorMatchingSnapshot();
   });
 
+  it(`should block .internal hostnames`, async () => {
+    await expect(
+      getLinkPreview(
+        `http://metadata.google.internal/computeMetadata/v1/instance/service-accounts/default/token`,
+      ),
+    ).rejects.toThrowErrorMatchingSnapshot();
+  });
+
+  it(`should block .local hostnames`, async () => {
+    await expect(
+      getLinkPreview(`http://printer.local/status`),
+    ).rejects.toThrowErrorMatchingSnapshot();
+  });
+
+  it(`should block nip.io wildcard hostnames`, async () => {
+    await expect(
+      getLinkPreview(`http://169.254.169.254.nip.io/latest/meta-data/iam/security-credentials/`),
+    ).rejects.toThrowErrorMatchingSnapshot();
+  });
+
+  it(`should block sslip.io wildcard hostnames`, async () => {
+    await expect(
+      getLinkPreview(`http://127.0.0.1.sslip.io/`),
+    ).rejects.toThrowErrorMatchingSnapshot();
+  });
+
   it(`should handle empty strings gracefully`, async () => {
     await expect(getLinkPreview(``)).rejects.toThrowErrorMatchingSnapshot();
   });

constants.ts

@@ -5,6 +5,9 @@ export const CONSTANTS = {
       "(?:(?:https?|ftp)://)" +
       // user:pass authentication
       "(?:\\S+(?::\\S*)?@)?" +
+      // block internal-only hostnames and wildcard DNS rebinding services
+      "(?![^/?#]+\\.(?:internal|local)(?::\\d{2,5})?(?:[/?#]|$))" +
+      "(?![^/?#]+\\.(?:nip|sslip)\\.io(?::\\d{2,5})?(?:[/?#]|$))" +
       "(?:" +
       // IP address exclusion
       // private & local networks
@@ -34,7 +37,7 @@ export const CONSTANTS = {
       // resource path
       "(?:[/?#]\\S*)?" +
       "$",
-    "i"
+    "i",
   ),
 
   REGEX_LOOPBACK: new RegExp(
@@ -63,8 +66,20 @@ export const CONSTANTS = {
       "|" +
       // Carrier-Grade NAT (CGNAT): 100.64.0.0 - 100.127.255.255
       "(?:100\\.(?:6[4-9]|[7-9]\\d|1[0-1]\\d)(?:\\.\\d{1,3}){2})" +
+      "|" +
+      // IPv6 loopback
+      "(?:::1)" +
+      "|" +
+      // IPv4-mapped IPv6 loopback: ::ffff:127.0.0.0/104
+      "(?:::ffff:127(?:\\.\\d{1,3}){3})" +
+      "|" +
+      // IPv6 Unique Local Address (ULA): fc00::/7
+      "(?:f[c-d][0-9a-f]{2}:[0-9a-f:]+)" +
+      "|" +
+      // IPv6 link-local unicast: fe80::/10
+      "(?:fe[89ab][0-9a-f]:[0-9a-f:]+)" +
       "$",
-    "i"
+    "i",
   ),
 
   REGEX_CONTENT_TYPE_IMAGE: new RegExp("image/.*", "i"),

index.ts

@@ -413,6 +413,10 @@ export async function getLinkPreview(text: string, options?: ILinkPreviewOptions
     const resolvedUrl = await options.resolveDNSHost(detectedUrl);
 
     throwOnLoopback(resolvedUrl);
+  } else {
+    console.error(
+      "[link-preview-js] You are not resolving DNS addresses (resolveDNSHost option) before fetching a link. This can cause loopback attacks. Always try to resolve DNS addresses",
+    );
   }
 
   const timeout = options?.timeout ?? 3000; // 3 second timeout default