amp-apester: Support keyword lookup in AmpDocShadow (#27345)

* amp-apester: Support keyword lookup in AmpDocShadow

Query meta[name="keywords"] using method supported by
AmpDocShadow as well as other AmpDoc subtypes.

* extratctTitle --> extractTitle

* Add test note about clearing meta_

extensions/amp-apester-media/0.1/amp-apester-media.js

@@ -152,7 +152,7 @@ class AmpApesterMedia extends AMP.BaseElement {
         'data-apester-channel-id'
       ),
       renderer: true,
-      tags: extractTags(this.getAmpDoc().getRootNode(), this.element),
+      tags: extractTags(this.getAmpDoc(), this.element),
     };
   }
 

extensions/amp-apester-media/0.1/test/test-amp-apester-media-utils.js

@@ -14,18 +14,23 @@
  * limitations under the License.
  */
 
+import {Services} from '../../../../src/services';
 import {
   extractArticleTags,
   extractElementTags,
   extractTags,
-  extratctTitle,
+  extractTitle,
 } from '../utils';
 
 describes.realWin('amp-apester-media-utils', {}, (unused) => {
   beforeEach(() => {
     document.body.textContent = '';
     document.head.textContent = '';
   });
+  afterEach(() => {
+    // Clear cached AmpDoc meta since document is reused in each test
+    Services.ampdoc(document).meta_ = null;
+  });
 
   it('Extract element tags as empty array when there is no element', () => {
     const expected = [];
@@ -51,7 +56,7 @@ describes.realWin('amp-apester-media-utils', {}, (unused) => {
 
   it('Extract title works well with no dl json', () => {
     const expected = [];
-    const tags = extratctTitle(document);
+    const tags = extractTitle(document);
     expect(tags).to.deep.equal(expected);
   });
 
@@ -76,7 +81,7 @@ describes.realWin('amp-apester-media-utils', {}, (unused) => {
     }`;
     element.text = jsonLd;
     document.body.appendChild(element);
-    const tags = extratctTitle(document);
+    const tags = extractTitle(document);
     expect(tags).to.deep.equal(expected);
   });
 
@@ -101,7 +106,7 @@ describes.realWin('amp-apester-media-utils', {}, (unused) => {
     }`;
     element.text = jsonLd;
     document.body.appendChild(element);
-    const tags = extratctTitle(document);
+    const tags = extractTitle(document);
     expect(tags).to.deep.equal(expected);
   });
 
@@ -129,7 +134,7 @@ describes.realWin('amp-apester-media-utils', {}, (unused) => {
       document.body.appendChild(element);
     }
 
-    const tags = extratctTitle(document);
+    const tags = extractTitle(document);
     expect(tags).to.deep.equal(expected);
   });
 
@@ -139,7 +144,8 @@ describes.realWin('amp-apester-media-utils', {}, (unused) => {
     meta.setAttribute('name', 'keywords');
     meta.setAttribute('content', '');
     document.head.appendChild(meta);
-    const tags = extractArticleTags(document);
+    const ampdoc = Services.ampdoc(document);
+    const tags = extractArticleTags(ampdoc);
     expect(tags).to.deep.equal(expected);
   });
 
@@ -149,7 +155,8 @@ describes.realWin('amp-apester-media-utils', {}, (unused) => {
     meta.setAttribute('name', 'keywords');
     meta.setAttribute('content', '');
     document.head.appendChild(meta);
-    const tags = extractArticleTags(document);
+    const ampdoc = Services.ampdoc(document);
+    const tags = extractArticleTags(ampdoc);
     expect(tags).to.deep.equal(expected);
   });
 
@@ -159,7 +166,8 @@ describes.realWin('amp-apester-media-utils', {}, (unused) => {
     meta.setAttribute('name', 'keywords');
     meta.setAttribute('content', 'this is, the, tag');
     document.head.appendChild(meta);
-    const tags = extractArticleTags(document);
+    const ampdoc = Services.ampdoc(document);
+    const tags = extractArticleTags(ampdoc);
     expect(tags).to.deep.equal(expected);
   });
 
@@ -201,7 +209,8 @@ describes.realWin('amp-apester-media-utils', {}, (unused) => {
     scriptElement.text = jsonLd;
     document.body.appendChild(scriptElement);
 
-    const tags = extractTags(document, element);
+    const ampdoc = Services.ampdoc(document);
+    const tags = extractTags(ampdoc, element);
     expect(tags).to.deep.equal(expected);
   });
 
@@ -233,7 +242,8 @@ describes.realWin('amp-apester-media-utils', {}, (unused) => {
     scriptElement.text = jsonLd;
     document.body.appendChild(scriptElement);
 
-    const tags = extractTags(document, element);
+    const ampdoc = Services.ampdoc(document);
+    const tags = extractTags(ampdoc, element);
     expect(tags).to.deep.equal(expected);
   });
 });

extensions/amp-apester-media/0.1/utils.js

@@ -88,10 +88,10 @@ export function extractElementTags(element) {
 
 /**
  * Extract article's first 5 words from the title and use them for tags.
- * @param {!Node} root
+ * @param {!Document|!ShadowRoot} root
  * @return {!Array<string>}
  */
-export function extratctTitle(root) {
+export function extractTitle(root) {
   const scriptTags = toArray(
     root.querySelectorAll('script[type="application/ld+json"]')
   );
@@ -117,31 +117,29 @@ export function extratctTitle(root) {
 }
 /**
  * Extracts article meta keywords
- * @param {*} root
- * @return {*} TODO(#23582): Specify return type
+ * @param {!../../../src/service/ampdoc-impl.AmpDoc} ampdoc
+ * @return {Array<string>}
  */
-export function extractArticleTags(root) {
-  const metaKeywords = root.querySelector('meta[name="keywords"]') || {
-    content: '',
-  };
-  return metaKeywords.content
-    .trim()
+export function extractArticleTags(ampdoc) {
+  return (ampdoc.getMetaByName('keywords') || '')
     .split(',')
-    .filter((e) => e)
-    .map((e) => e.trim());
+    .map((e) => e.trim())
+    .filter((e) => e);
 }
 
 /**
  * Extracts tags from a given element and document.
- * @param {!Node} root
+ * @param {!../../../src/service/ampdoc-impl.AmpDoc} ampdoc
  * @param {!Node} element
  * @return {Array<string>}
  */
-export function extractTags(root, element) {
+export function extractTags(ampdoc, element) {
   const extractedTags = extractElementTags(element) || [];
-  const articleMetaTags = extractArticleTags(root);
+  const articleMetaTags = extractArticleTags(ampdoc);
   const concatedTags = extractedTags.concat(
-    articleMetaTags.length ? articleMetaTags : extratctTitle(root) || []
+    articleMetaTags.length
+      ? articleMetaTags
+      : extractTitle(ampdoc.getRootNode()) || []
   );
   const loweredCase = concatedTags.map((tag) => tag.toLowerCase().trim());
   const noDuplication = loweredCase.filter(