Add option to prefix sitemap (#9846)

* Add option to prefix sitemap * Fix call resolve twice * let to const * Apply suggestions from code review Co-authored-by: Emanuele Stoppa <my.burning@gmail.com> * change changeset patch to minor * use node:test * Update changeset * Add regex validation for prefix * Update .changeset/eighty-falcons-tease.md Co-authored-by: Sarah Rainsberger <sarah@rainsberger.ca> * Update prefix regex in SitemapOptionsSchema --------- Co-authored-by: Emanuele Stoppa <my.burning@gmail.com> Co-authored-by: Sarah Rainsberger <sarah@rainsberger.ca>
withastro · Feb 20, 2024 · 9b78c99 · 9b78c99
1 parent 3c73441
commit 9b78c99
Show file tree

Hide file tree

Showing 4 changed files with 125 additions and 9 deletions.
diff --git a/.changeset/eighty-falcons-tease.md b/.changeset/eighty-falcons-tease.md
@@ -0,0 +1,24 @@
+---
+"@astrojs/sitemap": minor
+---
+
+Adds a new configuration option `prefix` that allows you to change the default `sitemap-*.xml` file name.
+
+By default, running `astro build` creates both `sitemap-index.xml` and `sitemap-0.xml` in your output directory. 
+
+To change the names of these files (e.g. to `astrosite-index.xml` and `astrosite-0.xml`), set the `prefix` option in your `sitemap` integration configuration:
+
+```
+import { defineConfig } from 'astro/config';
+import sitemap from '@astrojs/sitemap';
+export default defineConfig({
+  site: 'https://example.com',
+  integrations: [
+    sitemap({
+      prefix: 'astrosite-',
+    }),
+  ],
+});
+```
+
+This option is useful when Google Search Console is unable to fetch your default sitemap files, but can read renamed files.
diff --git a/packages/integrations/sitemap/src/index.ts b/packages/integrations/sitemap/src/index.ts
@@ -1,12 +1,14 @@
 import type { AstroConfig, AstroIntegration } from 'astro';
-import path from 'node:path';
+import path, { resolve } from 'node:path';
 import { fileURLToPath } from 'node:url';
 import type { EnumChangefreq, LinkItem as LinkItemBase, SitemapItemLoose } from 'sitemap';
-import { simpleSitemapAndIndex } from 'sitemap';
+import { SitemapAndIndexStream, SitemapStream, streamToPromise } from 'sitemap';
 import { ZodError } from 'zod';
 
 import { generateSitemap } from './generate-sitemap.js';
 import { validateOptions } from './validate-options.js';
+import { createWriteStream } from 'node:fs';
+import { Readable } from 'node:stream';
 
 export { EnumChangefreq as ChangeFreqEnum } from 'sitemap';
 export type ChangeFreq = `${EnumChangefreq}`;
@@ -33,6 +35,8 @@ export type SitemapOptions =
 			lastmod?: Date;
 			priority?: number;
 
+			prefix?: string;
+
 			// called for each sitemap item just before to save them on disk, sync or async
 			serialize?(item: SitemapItem): SitemapItem | Promise<SitemapItem | undefined> | undefined;
 	  }
@@ -44,7 +48,6 @@ function formatConfigErrorMessage(err: ZodError) {
 }
 
 const PKG_NAME = '@astrojs/sitemap';
-const OUTFILE = 'sitemap-index.xml';
 const STATUS_CODE_PAGES = new Set(['404', '500']);
 
 function isStatusCodePage(pathname: string): boolean {
@@ -77,7 +80,8 @@ const createPlugin = (options?: SitemapOptions): AstroIntegration => {
 
 					const opts = validateOptions(config.site, options);
 
-					const { filter, customPages, serialize, entryLimit } = opts;
+					const { filter, customPages, serialize, entryLimit, prefix = 'sitemap-' } = opts;
+					const OUTFILE = `${prefix}index.xml`;
 
 					let finalSiteUrl: URL;
 					if (config.site) {
@@ -166,13 +170,22 @@ const createPlugin = (options?: SitemapOptions): AstroIntegration => {
 						}
 					}
 					const destDir = fileURLToPath(dir);
-					await simpleSitemapAndIndex({
-						hostname: finalSiteUrl.href,
-						destinationDir: destDir,
-						sourceData: urlData,
+
+					const sms = new SitemapAndIndexStream({
 						limit: entryLimit,
-						gzip: false,
+						getSitemapStream: (i) => {
+							const sitemapStream = new SitemapStream({ hostname: finalSiteUrl.href });
+							const fileName = `${prefix}${i}.xml`;
+
+							const ws = sitemapStream.pipe(createWriteStream(resolve(destDir + fileName)));
+
+							return [new URL(fileName, finalSiteUrl.href).toString(), sitemapStream, ws];
+						},
 					});
+
+					sms.pipe(createWriteStream(resolve(destDir + OUTFILE)));
+					await streamToPromise(Readable.from(urlData).pipe(sms));
+					sms.end();
 					logger.info(`\`${OUTFILE}\` created at \`${path.relative(process.cwd(), destDir)}\``);
 				} catch (err) {
 					if (err instanceof ZodError) {

diff --git a/packages/integrations/sitemap/src/schema.ts b/packages/integrations/sitemap/src/schema.ts
@@ -34,6 +34,13 @@ export const SitemapOptionsSchema = z
 		changefreq: z.nativeEnum(ChangeFreq).optional(),
 		lastmod: z.date().optional(),
 		priority: z.number().min(0).max(1).optional(),
+
+		prefix: z
+			.string()
+			.regex(/^[a-zA-Z\-_]+$/gm, {
+				message: 'Only English alphabet symbols, hyphen and underscore allowed',
+			})
+			.optional(),
 	})
 	.strict()
 	.default(SITEMAP_CONFIG_DEFAULTS);
diff --git a/packages/integrations/sitemap/test/prefix.test.js b/packages/integrations/sitemap/test/prefix.test.js
@@ -0,0 +1,72 @@
+import { loadFixture, readXML } from './test-utils.js';
+import { sitemap } from './fixtures/static/deps.mjs';
+import assert from 'node:assert/strict';
+import { before, describe, it } from 'node:test';
+
+describe('Prefix support', () => {
+	/** @type {import('./test-utils.js').Fixture} */
+	let fixture;
+	const prefix = 'test-';
+
+	describe('Static', () => {
+		before(async () => {
+			fixture = await loadFixture({
+				root: './fixtures/static/',
+				integrations: [
+					sitemap(),
+					sitemap({
+						prefix,
+					}),
+				],
+			});
+			await fixture.build();
+		});
+
+		it('Content is same', async () => {
+			const data = await readXML(fixture.readFile('/sitemap-0.xml'));
+			const prefixData = await readXML(fixture.readFile(`/${prefix}0.xml`));
+			assert.deepEqual(prefixData, data);
+		});
+
+		it('Index file load correct sitemap', async () => {
+			const data = await readXML(fixture.readFile('/sitemap-index.xml'));
+			const sitemapUrl = data.sitemapindex.sitemap[0].loc[0];
+			assert.strictEqual(sitemapUrl, 'http://example.com/sitemap-0.xml');
+
+			const prefixData = await readXML(fixture.readFile(`/${prefix}index.xml`));
+			const prefixSitemapUrl = prefixData.sitemapindex.sitemap[0].loc[0];
+			assert.strictEqual(prefixSitemapUrl, `http://example.com/${prefix}0.xml`);
+		});
+	});
+
+	describe('SSR', () => {
+		before(async () => {
+			fixture = await loadFixture({
+				root: './fixtures/ssr/',
+				integrations: [
+					sitemap(),
+					sitemap({
+						prefix,
+					}),
+				],
+			});
+			await fixture.build();
+		});
+
+		it('Content is same', async () => {
+			const data = await readXML(fixture.readFile('/client/sitemap-0.xml'));
+			const prefixData = await readXML(fixture.readFile(`/client/${prefix}0.xml`));
+			assert.deepEqual(prefixData, data);
+		});
+
+		it('Index file load correct sitemap', async () => {
+			const data = await readXML(fixture.readFile('/client/sitemap-index.xml'));
+			const sitemapUrl = data.sitemapindex.sitemap[0].loc[0];
+			assert.strictEqual(sitemapUrl, 'http://example.com/sitemap-0.xml');
+
+			const prefixData = await readXML(fixture.readFile(`/client/${prefix}index.xml`));
+			const prefixSitemapUrl = prefixData.sitemapindex.sitemap[0].loc[0];
+			assert.strictEqual(prefixSitemapUrl, `http://example.com/${prefix}0.xml`);
+		});
+	});
+});