From 94a5214aa02f6ba684a3bc1ed39d1aff30adb4b3 Mon Sep 17 00:00:00 2001
From: Harlan Wilton <harlan@harlanzw.com>
Date: Mon, 13 Nov 2023 13:49:56 +1100
Subject: [PATCH] fix: filter i18n URLs based on non-prefixed path

---
 src/runtime/sitemap/builder/sitemap.ts  |  4 +-
 src/runtime/sitemap/urlset/filter.ts    | 28 ++++++++++----
 test/integration/i18n/filtering.test.ts | 51 +++++++++++++++++++++++++
 3 files changed, 74 insertions(+), 9 deletions(-)
 create mode 100644 test/integration/i18n/filtering.test.ts

diff --git a/src/runtime/sitemap/builder/sitemap.ts b/src/runtime/sitemap/builder/sitemap.ts
index 116564fd..f97d4ca4 100644
--- a/src/runtime/sitemap/builder/sitemap.ts
+++ b/src/runtime/sitemap/builder/sitemap.ts
@@ -93,7 +93,7 @@ export async function buildSitemap(sitemap: SitemapDefinition, resolvers: NitroU
       ).reverse()) as NitroRouteRules
 
       // apply top-level path without prefix, users can still target the localed path
-      if (autoI18n?.locales && autoI18n?.strategy === 'no_prefix') {
+      if (autoI18n?.locales && autoI18n?.strategy !== 'no_prefix') {
         // remove the locale path from the prefix, if it exists, need to use regex
         const match = path.match(new RegExp(`^/(${autoI18n.locales.map(l => l.code).join('|')})(.*)`))
         const pathWithoutPrefix = match?.[2]
@@ -117,7 +117,7 @@ export async function buildSitemap(sitemap: SitemapDefinition, resolvers: NitroU
     enhancedUrls = applyI18nEnhancements(enhancedUrls, { isI18nMapped, autoI18n, sitemapName: sitemap.sitemapName })
   // 3. filtered urls
   // TODO make sure include and exclude start with baseURL?
-  const filteredUrls = filterSitemapUrls(enhancedUrls, sitemap)
+  const filteredUrls = filterSitemapUrls(enhancedUrls, { autoI18n, ...sitemap })
   // 4. sort
   const sortedUrls = maybeSort(filteredUrls)
   // 5. maybe slice for chunked
diff --git a/src/runtime/sitemap/urlset/filter.ts b/src/runtime/sitemap/urlset/filter.ts
index b94b112b..87493898 100644
--- a/src/runtime/sitemap/urlset/filter.ts
+++ b/src/runtime/sitemap/urlset/filter.ts
@@ -1,6 +1,6 @@
 import { parseURL } from 'ufo'
 import { createRouter, toRouteMatcher } from 'radix3'
-import type { ResolvedSitemapUrl, SitemapDefinition } from '../../types'
+import type { ModuleRuntimeConfig, ResolvedSitemapUrl, SitemapDefinition } from '../../types'
 
 interface CreateFilterOptions {
   include?: (string | RegExp)[]
@@ -40,15 +40,29 @@ function createFilter(options: CreateFilterOptions = {}): (path: string) => bool
   }
 }
 
-export function filterSitemapUrls(_urls: ResolvedSitemapUrl[], filter: Pick<SitemapDefinition, 'sitemapName' | 'include' | 'exclude'>) {
+export function filterSitemapUrls(_urls: ResolvedSitemapUrl[], options: Pick<ModuleRuntimeConfig, 'autoI18n'> & Pick<SitemapDefinition, 'sitemapName' | 'include' | 'exclude'>) {
   // base may be wrong here
-  const urlFilter = createFilter(filter)
+  const urlFilter = createFilter(options)
   return _urls.filter((e) => {
-    if (e._sitemap && filter.sitemapName)
-      return e._sitemap === filter.sitemapName
+    if (e._sitemap && options.sitemapName)
+      return e._sitemap === options.sitemapName
     try {
-      const url = parseURL(e.loc)
-      return urlFilter(url.pathname)
+      const path = parseURL(e.loc).pathname
+      if (!urlFilter(path))
+        return false
+
+      const { autoI18n } = options
+      // if the non-prefixed locale is blocked then we block the prefixed versions
+      if (autoI18n?.locales && autoI18n?.strategy !== 'no_prefix') {
+        // remove the locale path from the prefix, if it exists, need to use regex
+        const match = path.match(new RegExp(`^/(${autoI18n.locales.map(l => l.code).join('|')})(.*)`))
+        const pathWithoutPrefix = match?.[2]
+        if (pathWithoutPrefix && pathWithoutPrefix !== path) {
+          if (!urlFilter(pathWithoutPrefix))
+            return false
+        }
+      }
+      return true
     }
     catch {
       // invalid URL
diff --git a/test/integration/i18n/filtering.test.ts b/test/integration/i18n/filtering.test.ts
new file mode 100644
index 00000000..797be2c0
--- /dev/null
+++ b/test/integration/i18n/filtering.test.ts
@@ -0,0 +1,51 @@
+import { describe, expect, it } from 'vitest'
+import { createResolver } from '@nuxt/kit'
+import { $fetch, setup } from '@nuxt/test-utils'
+
+const { resolve } = createResolver(import.meta.url)
+
+await setup({
+  rootDir: resolve('../../fixtures/i18n'),
+  nuxtConfig: {
+    sitemap: {
+      sitemaps: {
+        foo: {
+          urls: [
+            // custom blocked routes
+            '/admin',
+            '/es/admin',
+            '/fr/admin',
+            '/admin/foo',
+            '/es/admin/foo',
+            '/fr/admin/foo',
+            '/admin/foo/bar',
+            '/es/admin/foo/bar',
+            '/fr/admin/foo/bar',
+            // should be only route
+            '/valid',
+          ],
+          exclude: [
+            '/admin/**',
+          ],
+        },
+      },
+    },
+  },
+})
+describe('multi filtering', () => {
+  it('basic', async () => {
+    let sitemap = await $fetch('/foo-sitemap.xml')
+
+    // strip lastmod
+    sitemap = sitemap.replace(/<lastmod>.*<\/lastmod>/g, '')
+
+    expect(sitemap).toMatchInlineSnapshot(`
+      "<?xml version=\\"1.0\\" encoding=\\"UTF-8\\"?><?xml-stylesheet type=\\"text/xsl\\" href=\\"/__sitemap__/style.xsl\\"?>
+      <urlset xmlns:xsi=\\"http://www.w3.org/2001/XMLSchema-instance\\" xmlns:video=\\"http://www.google.com/schemas/sitemap-video/1.1\\" xmlns:xhtml=\\"http://www.w3.org/1999/xhtml\\" xmlns:image=\\"http://www.google.com/schemas/sitemap-image/1.1\\" xmlns:news=\\"http://www.google.com/schemas/sitemap-news/0.9\\" xsi:schemaLocation=\\"http://www.sitemaps.org/schemas/sitemap/0.9 http://www.sitemaps.org/schemas/sitemap/0.9/sitemap.xsd http://www.google.com/schemas/sitemap-image/1.1 http://www.google.com/schemas/sitemap-image/1.1/sitemap-image.xsd\\" xmlns=\\"http://www.sitemaps.org/schemas/sitemap/0.9\\">
+          <url>
+              <loc>https://nuxtseo.com/valid</loc>
+          </url>
+      </urlset>"
+    `)
+  }, 60000)
+})