Merge pull request #280 from ndaidong/v7.x.x

Change method to deal with `source` and `description`

dist/cjs/package.json

@@ -1,5 +1,5 @@
 {
   "name": "article-parser-cjs",
-  "version": "7.0.0rc3",
+  "version": "7.0.0rc4",
   "main": "./article-parser.js"
 }

package.json

@@ -1,5 +1,5 @@
 {
-  "version": "7.0.0rc3",
+  "version": "7.0.0rc4",
   "name": "article-parser",
   "description": "To extract main article from given URL",
   "homepage": "https://ndaidong.github.io/article-parser-demo/",
@@ -37,7 +37,8 @@
     "linkedom": "^0.14.12",
     "sanitize-html": "^2.7.0",
     "string-comparison": "^1.1.0",
-    "urlpattern-polyfill": "^5.0.3"
+    "tldts": "^5.7.84",
+    "urlpattern-polyfill": "^5.0.5"
   },
   "standard": {
     "ignore": [
@@ -48,7 +49,7 @@
     "@types/sanitize-html": "^2.6.2",
     "cross-env": "^7.0.3",
     "esbuild": "^0.14.49",
-    "jest": "^28.1.2",
+    "jest": "^28.1.3",
     "nock": "^13.2.8",
     "standard": "^17.0.0"
   },

src/config.js

@@ -42,8 +42,8 @@ const htmlCrushOptions = {
 const parserOptions = {
   wordsPerMinute: 300, // to estimate "time to read"
   urlsCompareAlgorithm: 'levenshtein', // to find the best url from list
-  descriptionLengthThreshold: 40, // min num of chars required for description
-  descriptionTruncateLen: 156, // max num of chars generated for description
+  descriptionLengthThreshold: 210, // min num of chars required for description
+  descriptionTruncateLen: 210, // max num of chars generated for description
   contentLengthThreshold: 200 // content must have at least 200 chars
 }
 

src/main.test.js

@@ -74,6 +74,7 @@ describe('test extract(bad url)', () => {
 })
 
 describe('test extract(regular article url)', () => {
+  const expDesc = "Navigation here Few can name a rational peach that isn't a conscientious goldfish! One cannot separate snakes from plucky pomegranates? Draped neatly on a hanger, the melons could be said to resemble knowledgeable pigs."
   const cases = [
     {
       input: {
@@ -101,7 +102,7 @@ describe('test extract(regular article url)', () => {
       validate: (result, expect) => {
         expect(result).toBeTruthy()
         expect(result.title).toEqual('Article title here')
-        expect(result.description).toEqual('Few words to summarize this article content')
+        expect(result.description).toEqual(expDesc)
       }
     }
   ]
@@ -124,6 +125,6 @@ describe('test extract(regular article url)', () => {
     const result = await extract(html)
     expect(result).toBeTruthy()
     expect(result.title).toEqual('Article title here')
-    expect(result.description).toEqual('Few words to summarize this article content')
+    expect(result.description).toEqual(expDesc)
   })
 })

src/utils/linker.js

@@ -18,16 +18,6 @@ export const isValid = (url = '') => {
   }
 }
 
-export const getHostname = (url) => {
-  try {
-    const { hostname } = new URL(url)
-    return hostname.replace('www.', '')
-  } catch (err) {
-    logger.error(err)
-    return ''
-  }
-}
-
 export const absolutify = (fullUrl = '', relativeUrl = '') => {
   try {
     const result = new URL(relativeUrl, fullUrl)

src/utils/linker.test.js

@@ -6,7 +6,6 @@ import { readFileSync } from 'fs'
 import { isString } from 'bellajs'
 
 import {
-  getHostname,
   chooseBestUrl,
   isValid as isValidUrl,
   purify as purifyUrl,
@@ -180,41 +179,6 @@ describe('test absolutifyUrl()', () => {
   })
 })
 
-describe('test getHostname()', () => {
-  const entries = [
-    {
-      url: '',
-      expected: ''
-    },
-    {
-      url: {},
-      expected: ''
-    },
-    {
-      url: 'https://www.some.where/article/abc-xyz',
-      expected: 'some.where'
-    },
-    {
-      url: 'https://www.alpha.some.where/blog/authors/article/abc-xyz',
-      expected: 'alpha.some.where'
-    },
-    {
-      url: 'https://10.1.1.5:1888/article/abc-xyz',
-      expected: '10.1.1.5'
-    }
-  ]
-  entries.forEach((entry) => {
-    const {
-      url,
-      expected
-    } = entry
-    test(`absolutifyUrl("${url}") must become "${expected}"`, () => {
-      const result = getHostname(url)
-      expect(result).toEqual(expected)
-    })
-  })
-})
-
 describe('test chooseBestUrl()', () => {
   test('test chooseBestUrl an actual case', () => {
     const title = 'Google đã ra giá mua Fitbit'

src/utils/parseFromHtml.js

@@ -1,6 +1,7 @@
 // utils -> parseFromHtml
 
 import { stripTags, truncate, unique, pipe } from 'bellajs'
+import { getDomain } from 'tldts'
 
 import { cleanify, cleanAndMinify as cleanAndMinifyHtml } from './html.js'
 
@@ -9,7 +10,6 @@ import {
   purify as purifyUrl,
   absolutify as absolutifyUrl,
   normalize as normalizeUrls,
-  getHostname,
   chooseBestUrl
 } from './linker.js'
 
@@ -28,9 +28,20 @@ import logger from './logger.js'
 import { getParserOptions } from '../config.js'
 
 const summarize = (desc, txt, threshold, maxlen) => {
-  return desc.length < threshold
-    ? truncate(txt, maxlen).replace(/\n/g, ' ')
-    : desc
+  const removeFirstParts = (str) => {
+    const arr = str.split(' - ')
+    if (arr.length > 1) {
+      arr.shift()
+      return arr.join(' ')
+    }
+    return str
+  }
+  const metadesc = removeFirstParts(desc)
+  if (metadesc.length > threshold) {
+    return metadesc
+  }
+  const extradesc = truncate(txt, maxlen).replace(/\n/g, ' ')
+  return removeFirstParts(extradesc)
 }
 
 export default async (inputHtml, inputUrl = '') => {
@@ -46,7 +57,6 @@ export default async (inputHtml, inputUrl = '') => {
     description: metaDesc,
     image: metaImg,
     author,
-    source,
     published
   } = meta
 
@@ -129,7 +139,7 @@ export default async (inputHtml, inputUrl = '') => {
     image,
     content,
     author,
-    source: source || getHostname(bestUrl),
+    source: getDomain(bestUrl),
     published,
     ttr: getTimeToRead(textContent)
   }

src/utils/parseFromHtml.test.js

@@ -9,6 +9,7 @@ import parseFromHtml from './parseFromHtml.js'
 import { addTransformations } from './transformation.js'
 
 describe('test parseFromHtml()', () => {
+  const expDesc = "Navigation here Few can name a rational peach that isn't a conscientious goldfish! One cannot separate snakes from plucky pomegranates? Draped neatly on a hanger, the melons could be said to resemble knowledgeable pigs."
   const cases = [
     {
       input: {
@@ -66,7 +67,7 @@ describe('test parseFromHtml()', () => {
       },
       expectation: (result, expect) => {
         expect(result.title).toEqual('Article title here')
-        expect(result.description).toEqual('Few words to summarize this article content')
+        expect(result.description).toEqual(expDesc)
         expect(result.content).toEqual(expect.stringContaining('<a target="_blank" href="https://otherwhere.com/descriptions/rational-peach">'))
         expect(result.content).toEqual(expect.stringContaining('<a target="_blank" href="https://somewhere.com/dict/watermelon">'))
       }

test-data/regular-article.html

@@ -15,7 +15,7 @@
     <meta property="og:title" content="Article title here">
     <meta property="og:type" content="article">
     <meta property="og:url" content="https://somewhere.com/path/to/article-title-here">
-    <meta property="og:description" content="Few words to summarize this article content">
+    <meta property="og:description" content="Navigation here Few can name a rational peach that isn't a conscientious goldfish! One cannot separate snakes from plucky pomegranates? Draped neatly on a hanger, the melons could be said to resemble knowledgeable pigs.">
     <meta property="og:image" content="https://somewhere.com/path/to/image.jpg">
     <meta property="article:published_time" content="2021-12-15T10:00:00.000+07:00">
     <meta property="article:modified_time" content="2021-12-16T09:00:00.000+07:00">

test-data/vnn-article.html

@@ -15,7 +15,7 @@
     <meta property="og:title" content="Article title here">
     <meta property="og:type" content="article">
     <meta property="og:url" content="https://vnn.vn/path/to/article-title-here">
-    <meta property="og:description" content="Few words to summarize this article content">
+    <meta property="og:description" content="VNN News - Few words to summarize this article content">
     <meta property="og:image" content="https://somewhere.com/path/to/image.jpg">
     <meta property="article:published_time" content="2021-12-15T10:00:00.000+07:00">
     <meta property="article:modified_time" content="2021-12-16T09:00:00.000+07:00">