added nock for better mocking

jacktuck · Sep 25, 2018 · 97563b8 · 97563b8
1 parent 947cd08
commit 97563b8
Show file tree

Hide file tree

Showing 13 changed files with 322 additions and 383 deletions.
diff --git a/package-lock.json b/package-lock.json
diff --git a/package.json b/package.json
@@ -8,6 +8,7 @@
   "main": "dist/index.js",
   "dependencies": {
     "debug": "^3.1.0",
+    "he": "^1.2.0",
     "htmlparser2": "^3.9.2",
     "iconv-lite": "^0.4.24",
     "node-fetch": "^2.2.0",
@@ -25,6 +26,7 @@
     "eslint-plugin-standard": "^4.0.0",
     "eslint-plugin-typescript": "^0.12.0",
     "jest": "^23.6.0",
+    "nock": "^10.0.0",
     "ts-jest": "^23.1.4",
     "tsc": "^1.20150623.0",
     "tslint": "^5.11.0",

diff --git a/src/index.ts b/src/index.ts
@@ -26,6 +26,7 @@ import {
 } from './schema'
 
 import { Metadata, Opts } from './types'
+import he from 'he'
 
 function unfurl (url: string, opts?: Opts): Promise<Metadata> {
   if (opts === undefined) {
@@ -280,8 +281,6 @@ function getMetadata (ctx, opts: Opts) {
             parser.reset()
           }
         }
-      }, {
-        decodeEntities: true
       })
 
       parser.write(text)
@@ -311,9 +310,10 @@ function parse (ctx) {
         continue
       }
 
-      if (item.type === 'string') {
-        metaValue = metaValue.toString()
-      } else if (item.type === 'number') {
+      // decoding html entities
+      metaValue = he.decode(he.decode(metaValue.toString()))
+
+      if (item.type === 'number') {
         metaValue = parseInt(metaValue, 10)
       } else if (item.type === 'url') {
         metaValue = resolveUrl(ctx.url, metaValue)

diff --git a/test/basic/double escaped-edge-case.html b/test/basic/double escaped-edge-case.html
@@ -0,0 +1,11 @@
+
+<!DOCTYPE html>
+<html>
+<head>
+  <meta charset="utf-8">
+  <meta name="description" content="&amp;quot;" />
+</head>
+<body></body>
+</html>
+
+
diff --git a/test/basic/test.ts b/test/basic/test.ts
@@ -1,17 +1,35 @@
 const unfurl = require('../../src/')
-import TestServer from '../server'
+import nock from 'nock'
+test('should detect title, description and keywords', async () => {
+  nock('http://localhost')
+    .get('/html/basic')
+    .replyWithFile(200, __dirname + '/basic.html', {
+      'Content-Type': 'text/html'
+    })
+
+  const result = await unfurl('http://localhost/html/basic')
 
-const port = process.env.port
-const baseUrl = `http://localhost:${port}`
+  const expected = {
+    favicon: 'http://localhost/favicon.ico',
+    description: 'aaa',
+    keywords: ['a', 'b', 'c'],
+    title: 'ccc'
+  }
 
-beforeAll(then => TestServer.listen(port, then))
-afterAll(then => TestServer.close(then))
+  expect(result).toEqual(expected)
+})
 
 test('should detect title, description and keywords', async () => {
-  const result = await unfurl(baseUrl + '/html/basic')
+  nock('http://localhost')
+    .get('/html/basic')
+    .replyWithFile(200, __dirname + '/basic.html', {
+      'Content-Type': 'text/html'
+    })
+
+  const result = await unfurl('http://localhost/html/basic')
 
   const expected = {
-    favicon: 'http://localhost:9000/favicon.ico',
+    favicon: 'http://localhost/favicon.ico',
     description: 'aaa',
     keywords: ['a', 'b', 'c'],
     title: 'ccc'
@@ -21,10 +39,16 @@ test('should detect title, description and keywords', async () => {
 })
 
 test('should detect last dupe of title, description and keywords', async () => {
-  const result = await unfurl(baseUrl + '/html/basic-duplicates')
+  nock('http://localhost')
+    .get('/html/basic-duplicates')
+    .replyWithFile(200, __dirname + '/basic-duplicates.html', {
+      'Content-Type': 'text/html'
+    })
+
+  const result = await unfurl('http://localhost/html/basic-duplicates')
 
   const expected = {
-    favicon: 'http://localhost:9000/favicon.ico',
+    favicon: 'http://localhost/favicon.ico',
     description: 'aaa',
     keywords: ['a', 'b', 'c'],
     title: 'ccc'
@@ -34,10 +58,16 @@ test('should detect last dupe of title, description and keywords', async () => {
 })
 
 test('should detect last dupe of title, description and keywords', async () => {
-  const result = await unfurl(baseUrl + '/html/keyword-edge-cases')
+  nock('http://localhost')
+  .get('/html/keyword-edge-cases')
+  .replyWithFile(200, __dirname + '/keyword-edge-cases.html', {
+    'Content-Type': 'text/html'
+  })
+
+  const result = await unfurl('http://localhost/html/keyword-edge-cases')
 
   const expected = {
-    favicon: 'http://localhost:9000/favicon.ico',
+    favicon: 'http://localhost/favicon.ico',
     keywords: ['foo', 'bar', 'baz quix', 'foo', 'foo']
   }
 

diff --git a/test/encoding/test.ts b/test/encoding/test.ts
@@ -1,14 +1,16 @@
 const unfurl = require('../../src/')
-import TestServer from '../server'
 
-const port = process.env.port
-const baseUrl = `http://localhost:${port}`
-
-beforeAll(then => TestServer.listen(port, then))
-afterAll(then => TestServer.close(then))
+import fs from 'fs'
+import nock from 'nock'
 
 test('should detect GB2312 charset (HTML 4) and convert to UTF-8', async () => {
-  const result = await unfurl(baseUrl + '/encoding/html/4')
+  nock('http://localhost')
+    .get('/html4/gb2312')
+    .replyWithFile(200, __dirname + '/html_4.html', { 'Content-Type': 'text/html' })
+
+    // require('http').get('http://localhost/html4/gb2312', console.log)
+
+  const result = await unfurl('http://localhost/html4/gb2312')
 
   const expected = {
     description: '腾讯网从2003年创立至今，已经成为集新闻信息，区域垂直生活服务、社会化媒体资讯和产品为一体的互联网媒体平台。腾讯网下设新闻、科技、财经、娱乐、体育、汽车、时尚等多个频道，充分满足用户对不同类型资讯的需求。同时专注不同领域内容，打造精品栏目，并顺应技术发展趋势，推出网络直播等创新形式，改变了用户获取资讯的方式和习惯。',
@@ -20,7 +22,11 @@ test('should detect GB2312 charset (HTML 4) and convert to UTF-8', async () => {
 })
 
 test('should detect GB2312 charset (HTML 5) and convert to UTF-8', async () => {
-  const result = await unfurl(baseUrl + '/encoding/html/5')
+  nock('http://localhost')
+    .get('/html5/gb2312')
+    .replyWithFile(200, __dirname + '/html_5.html', { 'Content-Type': 'text/html' })
+
+  const result = await unfurl('http://localhost/html5/gb2312')
 
   const expected = {
     description: '腾讯网从2003年创立至今，已经成为集新闻信息，区域垂直生活服务、社会化媒体资讯和产品为一体的互联网媒体平台。腾讯网下设新闻、科技、财经、娱乐、体育、汽车、时尚等多个频道，充分满足用户对不同类型资讯的需求。同时专注不同领域内容，打造精品栏目，并顺应技术发展趋势，推出网络直播等创新形式，改变了用户获取资讯的方式和习惯。',
@@ -32,11 +38,15 @@ test('should detect GB2312 charset (HTML 5) and convert to UTF-8', async () => {
 })
 
 test('should detect EUC-JP charset (HTML 5) and convert to UTF-8', async () => {
-  const result = await unfurl(baseUrl + '/encoding/html/euc-jp')
+  nock('http://localhost')
+    .get('/html5/euc-jp')
+    .replyWithFile(200, __dirname + '/euc-jp.html', { 'Content-Type': 'text/html' })
+
+  const result = await unfurl('http://localhost/html5/euc-jp')
 
   const expected = {
     description: '楽天市場はインターネット通販が楽しめる総合ショッピングモール。楽天スーパーポイントがどんどん貯まる！使える！毎日お得なクーポンも。あす楽利用で翌日にお届け。食品から家電、ファッション、ベビー用品、コスメまで、充実の品揃え。',
-    favicon: 'http://localhost:9000/favicon.ico',
+    favicon: 'http://localhost/favicon.ico',
     keywords: ['通販', 'インターネット通販', 'オンラインショッピング', '楽天', '楽天市場', 'rakuten', 'ラクテン', 'らくてん'],
     open_graph: {
       description: '楽天市場はインターネット通販が楽しめる総合ショッピングモール。楽天スーパーポイントがどんどん貯まる！使える！毎日お得なクーポンも。あす楽利用で翌日にお届け。食品から家電、ファッション、ベビー用品、コスメまで、充実の品揃え。',

diff --git a/test/general/content-type.test.ts b/test/general/content-type.test.ts
@@ -1,20 +1,21 @@
 import fs from 'fs'
 import http from 'http'
 import iconv from 'iconv-lite'
-import TestServer from '../server'
 
 const unfurl = require('../../src/')
 import UnexpectedError from '../../src/unexpectedError'
 
-const port = process.env.port
-const baseUrl = `http://localhost:${port}`
-
-beforeAll(then => TestServer.listen(port, then))
-afterAll(then => TestServer.close(then))
+import nock from 'nock'
 
 test('should throw bad content type error', async () => {
   try {
-    await unfurl(baseUrl + '/image')
+    nock('http://localhost')
+      .get('/image')
+      .reply(200, '', {
+        'Content-Type': 'image/png'
+      })
+
+    await unfurl('http://localhost/image')
   } catch (err) {
     expect(err.name).toEqual(UnexpectedError.EXPECTED_HTML.name)
     expect(err.message).toEqual(UnexpectedError.EXPECTED_HTML.message)