feat: add le monde extractor (#415)

src/extractors/custom/index.js

@@ -126,3 +126,4 @@ export * from './techlog.iij.ad.jp';
 export * from './wired.jp';
 export * from './japan.zdnet.com';
 export * from './www.rbbtoday.com';
+export * from './www.lemonde.fr';

src/extractors/custom/www.lemonde.fr/index.js

@@ -0,0 +1,31 @@
+export const WwwLemondeFrExtractor = {
+  domain: 'www.lemonde.fr',
+
+  title: {
+    selectors: ['h1.article__title'],
+  },
+
+  author: {
+    selectors: ['.author__name'],
+  },
+
+  date_published: {
+    selectors: [['meta[name="og:article:published_time"]', 'value']],
+  },
+
+  dek: {
+    selectors: ['.article__desc'],
+  },
+
+  lead_image_url: {
+    selectors: [['meta[name="og:image"]', 'value']],
+  },
+
+  content: {
+    selectors: ['.article__content'],
+
+    transforms: {},
+
+    clean: [],
+  },
+};

src/extractors/custom/www.lemonde.fr/index.test.js

@@ -0,0 +1,86 @@
+import assert from 'assert';
+import URL from 'url';
+import cheerio from 'cheerio';
+
+import Mercury from 'mercury';
+import getExtractor from 'extractors/get-extractor';
+import { excerptContent } from 'utils/text';
+
+const fs = require('fs');
+
+describe('WwwLemondeFrExtractor', () => {
+  describe('initial test case', () => {
+    let result;
+    let url;
+    beforeAll(() => {
+      url =
+        'https://www.lemonde.fr/economie/article/2019/05/07/dans-ses-previsions-economiques-bruxelles-confirme-la-montee-des-perils_5459325_3234.html';
+      const html = fs.readFileSync(
+        './fixtures/www.lemonde.fr/1557235525251.html'
+      );
+      result = Mercury.parse(url, { html, fallback: false });
+    });
+
+    it('is selected properly', () => {
+      const extractor = getExtractor(url);
+      assert.equal(extractor.domain, URL.parse(url).hostname);
+    });
+
+    it('returns the title', async () => {
+      const { title } = await result;
+
+      assert.equal(
+        title,
+        `Les sombres perspectives économiques de la Commission européenne`
+      );
+    });
+
+    it('returns the author', async () => {
+      const { author } = await result;
+
+      assert.equal(author, `Cécile Ducourtieux`);
+    });
+
+    it('returns the date_published', async () => {
+      const { date_published } = await result;
+
+      assert.equal(date_published, `2019-05-07T11:59:43.000Z`);
+    });
+
+    it('returns the dek', async () => {
+      const { dek } = await result;
+
+      assert.equal(
+        dek,
+        'Elle abaisse ses prévisions pour 2019, avec un PIB à 1,4 % pour l’ensemble de l’UE, et à 1,2 % pour la zone euro.'
+      );
+    });
+
+    it('returns the lead_image_url', async () => {
+      const { lead_image_url } = await result;
+
+      assert.equal(
+        lead_image_url,
+        `https://img.lemde.fr/2019/05/07/316/0/3824/1912/1440/720/60/0/d105b14_dfjDE1I-caggQrT4gvHf2nZP.jpg`
+      );
+    });
+
+    it('returns the content', async () => {
+      const { content } = await result;
+
+      const $ = cheerio.load(content || '');
+
+      const first13 = excerptContent(
+        $('*')
+          .first()
+          .text(),
+        13
+      );
+
+      assert.equal(
+        first13,
+        'Les dirigeants européens qui doivent se réunir, jeudi 9 mai à Sibiu (Roumanie),'
+      );
+    });
+  });
+});