fix(core): prevent 404 when accessing /page.html (#7184)

jest.config.mjs

@@ -12,6 +12,7 @@ process.env.TZ = 'UTC';
 const ignorePatterns = [
   '/node_modules/',
   '__fixtures__',
+  '__mocks__',
   '/testUtils.ts',
   '/packages/docusaurus/lib',
   '/packages/docusaurus-logger/lib',

packages/docusaurus/src/client/__tests__/__mocks__/@generated/routes.ts

@@ -0,0 +1,30 @@
+/**
+ * Copyright (c) Facebook, Inc. and its affiliates.
+ *
+ * This source code is licensed under the MIT license found in the
+ * LICENSE file in the root directory of this source tree.
+ */
+
+export default [
+  {
+    path: '/page.html',
+    exact: true,
+    component: '',
+  },
+  {
+    path: '/docs',
+    exact: false,
+    component: '',
+    routes: [
+      {
+        path: '/docs/installation',
+        exact: true,
+        component: '',
+      },
+    ],
+  },
+  {
+    path: '*',
+    component: '',
+  },
+];

packages/docusaurus/src/client/__tests__/normalizeLocation.test.ts

@@ -43,6 +43,33 @@ describe('normalizeLocation', () => {
     });
   });
 
+  it('removes html extension', () => {
+    expect(
+      normalizeLocation({
+        pathname: '/docs/installation.html',
+      }),
+    ).toEqual({
+      pathname: '/docs/installation',
+    });
+    expect(
+      normalizeLocation({
+        pathname: '/docs/introduction/foo.html',
+        search: '',
+        hash: '#bar',
+      }),
+    ).toEqual({
+      pathname: '/docs/introduction/foo',
+      search: '',
+      hash: '#bar',
+    });
+  });
+
+  it('does not strip extension if the route location has one', () => {
+    expect(normalizeLocation({pathname: '/page.html'})).toEqual({
+      pathname: '/page.html',
+    });
+  });
+
   it('leaves pathnames untouched', () => {
     const replaceMock = jest.spyOn(String.prototype, 'replace');
 
@@ -72,18 +99,6 @@ describe('normalizeLocation', () => {
     });
     expect(replaceMock).toBeCalledTimes(1);
 
-    expect(
-      normalizeLocation({
-        pathname: '/docs/introduction/foo.html',
-        search: '',
-        hash: '#bar',
-      }),
-    ).toEqual({
-      pathname: '/docs/introduction/foo.html',
-      search: '',
-      hash: '#bar',
-    });
-
     expect(
       normalizeLocation({
         pathname: '/',

packages/docusaurus/src/client/normalizeLocation.ts

@@ -5,23 +5,33 @@
  * LICENSE file in the root directory of this source tree.
  */
 
+import {matchRoutes} from 'react-router-config';
+import routes from '@generated/routes';
 import type {Location} from 'history';
 
 // Memoize previously normalized pathnames.
-const pathnames: {[rawPathname: string]: string} = {};
+const pathnames = new Map<string, string>();
 
 export default function normalizeLocation<T extends Location>(location: T): T {
-  if (pathnames[location.pathname]) {
+  if (pathnames.has(location.pathname)) {
     return {
       ...location,
-      pathname: pathnames[location.pathname],
+      pathname: pathnames.get(location.pathname),
     };
   }
 
+  // If the location was registered with an `.html` extension, we don't strip it
+  // away, or it will render to a 404 page.
+  const matchedRoutes = matchRoutes(routes, location.pathname);
+  if (matchedRoutes.some(({route}) => route.exact === true)) {
+    pathnames.set(location.pathname, location.pathname);
+    return location;
+  }
+
   const pathname =
-    location.pathname.trim().replace(/\/index\.html$/, '') || '/';
+    location.pathname.trim().replace(/(?:\/index)?\.html$/, '') || '/';
 
-  pathnames[location.pathname] = pathname;
+  pathnames.set(location.pathname, pathname);
 
   return {
     ...location,

website/_dogfooding/_docs tests/dummy.md

@@ -1 +1,5 @@
+---
+slug: dummy.html
+---
+
 # Just a dummy page