Merge pull request #291 from ndaidong/7.0.3

v7.0.3

dist/cjs/package.json

@@ -1,5 +1,5 @@
 {
   "name": "article-parser-cjs",
-  "version": "7.0.2",
+  "version": "7.0.3",
   "main": "./article-parser.js"
 }

package.json

@@ -1,5 +1,5 @@
 {
-  "version": "7.0.2",
+  "version": "7.0.3",
   "name": "article-parser",
   "description": "To extract main article from given URL",
   "homepage": "https://ndaidong.github.io/article-parser-demo/",
@@ -9,6 +9,10 @@
   },
   "author": "@ndaidong",
   "main": "./src/main.js",
+  "exports": {
+    "import": "./src/main.js",
+    "require": "./dist/cjs/article-parser.js"
+  },
   "browser": {
     "linkedom": "./src/browser/linkedom.js",
     "./main.js": "./dist/article-parser.browser.js"
@@ -30,13 +34,12 @@
   "dependencies": {
     "@mozilla/readability": "^0.4.2",
     "axios": "^0.27.2",
-    "bellajs": "^11.0.5",
-    "html-crush": "^5.1.3",
-    "linkedom": "^0.14.12",
-    "sanitize-html": "^2.7.1",
+    "bellajs": "^11.0.7",
+    "html-crush": "^5.1.6",
+    "linkedom": "^0.14.14",
+    "sanitize-html": "^2.7.2",
     "string-comparison": "^1.1.0",
-    "tldts": "^5.7.90",
-    "urlpattern-polyfill": "^5.0.6"
+    "urlpattern-polyfill": "^5.0.9"
   },
   "standard": {
     "ignore": [
@@ -46,8 +49,8 @@
   "devDependencies": {
     "@types/sanitize-html": "^2.6.2",
     "cross-env": "^7.0.3",
-    "esbuild": "^0.15.6",
-    "jest": "^29.0.1",
+    "esbuild": "^0.15.7",
+    "jest": "^29.0.3",
     "nock": "^13.2.9",
     "standard": "^17.0.0"
   },

src/config.js

@@ -4,7 +4,7 @@ import { clone, copies } from 'bellajs'
 
 const requestOptions = {
   headers: {
-    'user-agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Firefox/102.0',
+    'user-agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:104.0) Gecko/20100101 Firefox/104.0',
     accept: 'text/html; charset=utf-8',
     'accept-encoding': 'deflate,zlib,gzip'
   },
@@ -15,20 +15,49 @@ const requestOptions = {
 }
 
 const sanitizeHtmlOptions = {
-  allowedTags: ['h1', 'h2', 'h3', 'h4', 'h5', 'u', 'b', 'i', 'em', 'strong', 'small', 'sup', 'sub', 'div', 'span', 'p', 'article', 'blockquote', 'section', 'details', 'summary', 'pre', 'code', 'ul', 'ol', 'li', 'dd', 'dl', 'table', 'th', 'tr', 'td', 'thead', 'tbody', 'tfood', 'fieldset', 'legend', 'figure', 'figcaption', 'img', 'picture', 'video', 'audio', 'source', 'iframe', 'progress', 'br', 'p', 'hr', 'label', 'abbr', 'a', 'svg'],
+  allowedTags: [
+    'h1', 'h2', 'h3', 'h4', 'h5', 'h6',
+    'u', 'b', 'i', 'em', 'strong', 'small', 'sup', 'sub',
+    'div', 'span', 'p', 'article', 'blockquote', 'section',
+    'details', 'summary',
+    'pre', 'code',
+    'ul', 'ol', 'li', 'dd', 'dl',
+    'table', 'th', 'tr', 'td', 'thead', 'tbody', 'tfood',
+    'fieldset', 'legend',
+    'figure', 'figcaption', 'img', 'picture',
+    'video', 'audio', 'source',
+    'iframe',
+    'progress',
+    'br', 'p', 'hr',
+    'label',
+    'abbr',
+    'a',
+    'svg'
+  ],
   allowedAttributes: {
+    h1: ['id'],
+    h2: ['id'],
+    h3: ['id'],
+    h4: ['id'],
+    h5: ['id'],
+    h6: ['id'],
     a: ['href', 'target', 'title'],
     abbr: ['title'],
     progress: ['value', 'max'],
-    img: ['src', 'srcset', 'alt', 'width', 'height', 'style', 'title'],
+    img: ['src', 'srcset', 'alt', 'title'],
     picture: ['media', 'srcset'],
     video: ['controls', 'width', 'height', 'autoplay', 'muted', 'loop', 'src'],
-    audio: ['controls'],
+    audio: ['controls', 'width', 'height', 'autoplay', 'muted', 'loop', 'src'],
     source: ['src', 'srcset', 'data-srcset', 'type', 'media', 'sizes'],
     iframe: ['src', 'frameborder', 'height', 'width', 'scrolling', 'allow'],
     svg: ['width', 'height'] // sanitize-html does not support svg fully yet
   },
-  allowedIframeDomains: ['youtube.com', 'twitter.com', 'facebook.com', 'vimeo.com']
+  allowedIframeDomains: [
+    'youtube.com', 'vimeo.com', 'odysee.com',
+    'soundcloud.com', 'audius.co',
+    'github.com', 'codepen.com',
+    'twitter.com', 'facebook.com', 'instagram.com'
+  ]
 }
 
 /**

src/config.test.js

@@ -24,7 +24,7 @@ test('Testing setRequestOptions/getRequestOptions methods', () => {
   const actual = getRequestOptions()
   const expectedHeader = {
     authorization: 'bearer <token>',
-    'user-agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Firefox/102.0',
+    'user-agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:104.0) Gecko/20100101 Firefox/104.0',
     accept: 'text/html; charset=utf-8',
     'accept-encoding': 'deflate,zlib,gzip'
   }

src/utils/linker.js

@@ -143,3 +143,8 @@ export const chooseBestUrl = (candidates = [], title = '') => {
     value: shortestUrl
   }).value
 }
+
+export const getDomain = (url) => {
+  const host = (new URL(url)).host
+  return host.replace('www.', '')
+}

src/utils/parseFromHtml.js

@@ -1,7 +1,6 @@
 // utils -> parseFromHtml
 
 import { stripTags, truncate, unique, pipe } from 'bellajs'
-import { getDomain } from 'tldts'
 
 import { cleanify, cleanAndMinify as cleanAndMinifyHtml } from './html.js'
 
@@ -10,7 +9,8 @@ import {
   purify as purifyUrl,
   absolutify as absolutifyUrl,
   normalize as normalizeUrls,
-  chooseBestUrl
+  chooseBestUrl,
+  getDomain
 } from './linker.js'
 
 import extractMetaData from './extractMetaData.js'