@adobe/helix-html-pipeline 5.6.2 → 5.7.0

package/CHANGELOG.md

@@ -1,3 +1,17 @@
+# [5.7.0](https://github.com/adobe/helix-html-pipeline/compare/v5.6.3...v5.7.0) (2024-02-05)
+
+
+### Features
+
+* render sitemap from `sitemap.json` ([#527](https://github.com/adobe/helix-html-pipeline/issues/527)) ([414e36e](https://github.com/adobe/helix-html-pipeline/commit/414e36ea59e37a0f7c018db43aa2fe126172688f))
+
+## [5.6.3](https://github.com/adobe/helix-html-pipeline/compare/v5.6.2...v5.6.3) (2024-01-30)
+
+
+### Bug Fixes
+
+* **deps:** update dependency @adobe/remark-gridtables to v3 ([#522](https://github.com/adobe/helix-html-pipeline/issues/522)) ([84c0dfc](https://github.com/adobe/helix-html-pipeline/commit/84c0dfc5a85cc73faa28ed7f09568c52de9dabfb))
+
 ## [5.6.2](https://github.com/adobe/helix-html-pipeline/compare/v5.6.1...v5.6.2) (2024-01-29)
 
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@adobe/helix-html-pipeline",
-  "version": "5.6.2",
+  "version": "5.7.0",
   "description": "Helix HTML Pipeline",
   "main": "src/index.js",
   "types": "src/index.d.ts",
@@ -45,7 +45,7 @@
     "@adobe/helix-markdown-support": "7.1.0",
     "@adobe/helix-shared-utils": "3.0.1",
     "@adobe/mdast-util-gridtables": "4.0.0",
-    "@adobe/remark-gridtables": "2.0.2",
+    "@adobe/remark-gridtables": "3.0.0",
     "cookie": "0.6.0",
     "github-slugger": "2.0.0",
     "hast-util-raw": "9.0.2",
@@ -53,7 +53,8 @@
     "hast-util-to-html": "9.0.0",
     "hast-util-to-string": "3.0.0",
     "hastscript": "9.0.0",
-    "jose": "5.2.0",
+    "jose": "5.2.1",
+    "lodash.escape": "4.0.1",
     "mdast-util-to-hast": "13.1.0",
     "mdast-util-to-string": "4.0.0",
     "mime": "4.0.1",
@@ -70,7 +71,7 @@
     "unist-util-visit-parents": "6.0.1"
   },
   "devDependencies": {
-    "@adobe/eslint-config-helix": "2.0.5",
+    "@adobe/eslint-config-helix": "2.0.6",
     "@markedjs/html-differ": "4.0.2",
     "@semantic-release/changelog": "6.0.3",
     "@semantic-release/git": "10.0.1",
@@ -81,11 +82,11 @@
     "eslint-plugin-header": "3.1.1",
     "eslint-plugin-import": "2.29.1",
     "esmock": "2.6.3",
-    "husky": "9.0.6",
+    "husky": "9.0.10",
     "js-yaml": "4.1.0",
     "jsdom": "24.0.0",
-    "junit-report-builder": "3.1.0",
-    "lint-staged": "15.2.0",
+    "junit-report-builder": "3.2.1",
+    "lint-staged": "15.2.1",
     "mocha": "10.2.0",
     "mocha-multi-reporters": "1.5.1",
     "mocha-suppress-logs": "0.4.1",

package/src/sitemap-pipe.js

@@ -9,6 +9,7 @@
  * OF ANY KIND, either express or implied. See the License for the specific language
  * governing permissions and limitations under the License.
  */
+import escape from 'lodash.escape';
 import { cleanupHeaderValue } from '@adobe/helix-shared-utils';
 import { authenticate, requireProject } from './steps/authenticate.js';
 import fetchConfig from './steps/fetch-config.js';
@@ -20,6 +21,49 @@ import setCustomResponseHeaders from './steps/set-custom-response-headers.js';
 import { PipelineStatusError } from './PipelineStatusError.js';
 import { PipelineResponse } from './PipelineResponse.js';
 
+async function generateSitemap(state, partition) {
+  const {
+    owner, repo, ref, contentBusId, s3Loader, log,
+    previewHost, liveHost, config: { host: prodCDN } = {},
+  } = state;
+  const ret = await s3Loader.getObject('helix-content-bus', `${contentBusId}/live/sitemap.json`);
+  if (ret.status !== 200) {
+    return ret;
+  }
+  let config;
+  try {
+    config = JSON.parse(ret.body);
+  } catch (e) {
+    log.info('failed to parse /sitemap.json', e);
+    throw new PipelineStatusError(404, `Failed to parse /sitemap.json: ${e.message}`);
+  }
+  const { data } = config;
+  if (!data || !Array.isArray(data)) {
+    throw new PipelineStatusError(404, 'Expected \'data\' array not found in /sitemap.json');
+  }
+  const host = partition === 'preview'
+    ? (previewHost || `${ref}--${repo}--${owner}.hlx.page`)
+    : (prodCDN || liveHost || `${ref}--${repo}--${owner}.hlx.live`);
+  const loc = ({ path, lastModified }) => `  <url>
+    <loc>
+      https://${host}${escape(path)}
+    </loc>
+    <lastmod>${new Date(lastModified * 1000).toISOString().substring(0, 10)}</lastmod>
+  </url>`;
+  const xml = [
+    '<?xml version="1.0" encoding="utf-8"?>',
+    '<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9" xmlns:xhtml="http://www.w3.org/1999/xhtml">',
+    ...data.map((record) => loc(record)),
+    '</urlset>',
+  ].join('\n');
+  return new PipelineResponse(xml, {
+    status: 200,
+    headers: {
+      'content-type': 'application/xml; charset=utf-8',
+    },
+  });
+}
+
 /**
  * Serves or renders the sitemap xml. The sitemap is always served from the preview content-bus
  * partition.
@@ -32,8 +76,9 @@ import { PipelineResponse } from './PipelineResponse.js';
  * @returns {PipelineResponse}
  */
 export async function sitemapPipe(state, req) {
-  const { log } = state;
+  const { partition, log } = state;
   state.type = 'sitemap';
+
   // force loading from preview
   state.partition = 'preview';
 
@@ -78,7 +123,14 @@ export async function sitemapPipe(state, req) {
     if (res.error !== 401) {
       await authenticate(state, req, res);
     }
-
+    if (res.status === 404) {
+      const ret = await generateSitemap(state, partition);
+      if (ret.status === 200) {
+        res.status = 200;
+        delete res.error;
+        state.content.data = ret.body;
+      }
+    }
     if (res.error) {
       // if content loading produced an error, we're done.
       throw new PipelineStatusError(res.status, res.error);