fix: incorrect parsing on medium.com (#477)

* fix: medium extractor now pulls content

* fix: remove youtube caption if no preview available

* fix: remove youtube node if no image

* fix: removing dek from medium.com extractor

src/extractors/custom/medium.com/index.js

@@ -1,22 +1,16 @@
 export const MediumExtractor = {
   domain: 'medium.com',
 
-  supportedDomains: ['trackchanges.postlight.com'],
-
   title: {
-    selectors: ['h1'],
+    selectors: ['h1', ['meta[name="og:title"]', 'value']],
   },
 
   author: {
     selectors: [['meta[name="author"]', 'value']],
   },
 
   content: {
-    selectors: [
-      ['.section-content'],
-      '.section-content',
-      'article > div > section',
-    ],
+    selectors: ['article'],
 
     // Is there anything in the content you selected that needs transformed
     // before it's consumable content? E.g., unusual lazy loaded images
@@ -25,14 +19,18 @@ export const MediumExtractor = {
       iframe: $node => {
         const ytRe = /https:\/\/i.embed.ly\/.+url=https:\/\/i\.ytimg\.com\/vi\/(\w+)\//;
         const thumb = decodeURIComponent($node.attr('data-thumbnail'));
+        const $parent = $node.parents('figure');
 
         if (ytRe.test(thumb)) {
           const [_, youtubeId] = thumb.match(ytRe); // eslint-disable-line
           $node.attr('src', `https://www.youtube.com/embed/${youtubeId}`);
-          const $parent = $node.parents('figure');
           const $caption = $parent.find('figcaption');
           $parent.empty().append([$node, $caption]);
+          return;
         }
+
+        // If we can't draw the YouTube preview, remove the figure.
+        $parent.remove();
       },
 
       // rewrite figures to pull out image and caption, remove rest
@@ -42,29 +40,33 @@ export const MediumExtractor = {
 
         const $img = $node.find('img').slice(-1)[0];
         const $caption = $node.find('figcaption');
+
         $node.empty().append([$img, $caption]);
       },
+
+      // Remove any smaller images that did not get caught by the generic image
+      // cleaner (author photo 48px, leading sentence images 79px, etc.).
+      img: $node => {
+        const width = parseInt($node.attr('width'), 10);
+        if (width < 100) $node.remove();
+      },
     },
 
     // Is there anything that is in the result that shouldn't be?
     // The clean selectors will remove anything that matches from
     // the result
-    clean: [],
+    clean: ['span', 'svg'],
   },
 
   date_published: {
-    selectors: [['time[datetime]', 'datetime']],
+    selectors: [['meta[name="article:published_time"]', 'value']],
   },
 
   lead_image_url: {
     selectors: [['meta[name="og:image"]', 'value']],
   },
 
-  dek: {
-    selectors: [
-      // enter selectors
-    ],
-  },
+  dek: null,
 
   next_page_url: {
     selectors: [

src/extractors/custom/medium.com/index.test.js

@@ -44,7 +44,7 @@ describe('MediumExtractor', () => {
     it('returns the date_published', async () => {
       const { date_published } = await result;
 
-      assert.equal(date_published, '2016-10-19T14:24:20.323Z');
+      assert.equal(date_published, '2016-10-19T14:30:56.529Z');
     });
 
     it('returns the dek', async () => {
@@ -62,7 +62,7 @@ describe('MediumExtractor', () => {
       // the article.
       assert.equal(
         lead_image_url,
-        'https://cdn-images-1.medium.com/max/1200/1*3Gzaug9mRc8vvx1cuQWkog.png'
+        'https://miro.medium.com/max/540/1*3Gzaug9mRc8vvx1cuQWkog.png'
       );
     });
 
@@ -78,12 +78,9 @@ describe('MediumExtractor', () => {
         13
       );
 
-      // testing that youtube video transform is working
-      assert.equal(/IAoy3ia2ivI/.test(content), true);
-
       assert.equal(
         first13,
-        'Video of WTF? My talk at the White House Frontiers ConferenceLast Thursday, I'
+        'Last Thursday, I had the honor to be one of the warmup acts'
       );
     });
   });