Throw qualified errors (#118)

* Throw qualified errors

* Tweaks

Author: Kikobeats

README.md

@@ -258,13 +258,6 @@ Call the instance for extracting content based on rules bundle provided at the c
 
 #### options
 
-##### html
-
-*Required*<br>
-Type: `String`
-
-The HTML markup for extracting the content.
-
 #### url
 
 *Required*<br>
@@ -276,6 +269,12 @@ It is used for resolve relative links that can be present in the HTML markup.
 
 it can be used as fallback field for different rules as well.
 
+##### html
+
+Type: `String`
+
+The HTML markup for extracting the content.
+
 #### rules
 
 Type: `Array`

packages/metascraper-clearbit-logo/index.js

@@ -12,7 +12,8 @@ const DEFAULTS = {
 const ENDPOINT = 'https://logo.clearbit.com'
 
 module.exports = opts => {
-  opts = Object.assign({}, DEFAULTS, opts)
+  opts = { ...DEFAULTS, ...opts }
+
   const { size, format, greyscale } = opts
 
   const clearbitLogo = async ({ url }) => {

packages/metascraper-clearbit-logo/test/index.js

@@ -2,23 +2,12 @@
 
 const should = require('should')
 
-const metascraper = require('metascraper')([
-  require('metascraper-author')(),
-  require('metascraper-date')(),
-  require('metascraper-description')(),
-  require('metascraper-image')(),
-  require('metascraper-logo')(),
-  require('metascraper-clearbit-logo')(),
-  require('metascraper-publisher')(),
-  require('metascraper-title')(),
-  require('metascraper-url')()
-])
+const metascraper = require('metascraper')([require('..')()])
 
 describe('metascraper-clearbit-logo', () => {
   it('returns when is possible resolve logo', async () => {
     const url = 'https://facebook.com'
-    const html = '<div></div>'
-    const meta = await metascraper({ html, url })
+    const meta = await metascraper({ url })
     should(meta.logo.indexOf('clearbit') !== -1).be.true()
   })
 

packages/metascraper-helpers/index.js

@@ -42,11 +42,9 @@ const REGEX_LOCATION = /^[A-Z\s]+\s+[-—–]\s+/
 
 const removeLocation = value => replace(value, REGEX_LOCATION, '')
 
-const urlTest = (url, { relative = true }) =>
+const isUrl = (url, { relative = false } = {}) =>
   relative ? isRelativeUrl(url) || urlRegex.test(url) : urlRegex.test(url)
 
-const isUrl = (url, opts = {}) => !isEmpty(url) && urlTest(url, opts)
-
 const absoluteUrl = (baseUrl, relativePath = '') =>
   resolveUrl(baseUrl, relativePath)
 
@@ -118,7 +116,11 @@ const publisher = value => isString(value) && condenseWhitespace(value)
 
 const author = value => isAuthor(value) && getAuthor(value)
 
-const url = (value, { url }) => isUrl(value) && normalizeUrl(url, value)
+const url = (value, { url }) => {
+  if (isEmpty(value)) return false
+  const absoluteUrl = normalizeUrl(url, value)
+  return isUrl(absoluteUrl) && absoluteUrl
+}
 
 const date = value => {
   if (!value) return false

packages/metascraper-logo-favicon/__snapshots__/index.js.snap-shot

@@ -1,11 +1,4 @@
 exports['create an absolute favicon url if the logo is not present 1'] = {
-  "lang": "en",
-  "author": "Mat Follas",
-  "title": "Vegetable Perfection: 100 delicious recipes for roots, bulbs, shoots and stems",
-  "publisher": "Amazon",
-  "image": "https://images-eu.ssl-images-amazon.com/images/I/61Q2c36lYeL._SX218_BO1,204,203,200_QL40_.jpg",
-  "description": "Buy Vegetable Perfection: 100 delicious recipes for roots, bulbs, shoots and stems by Mat Follas (ISBN: 9781849757096) from Amazon’s Book Store. Everyday low prices and free delivery on eligible orders.",
-  "logo": "https://www.amazon.co.uk/favicon.ico",
-  "url": "https://www.amazon.co.uk/Vegetable-Perfection-delicious-recipes-shoots/dp/1849757097"
+  "logo": "https://www.amazon.co.uk/favicon.ico"
 }
 

packages/metascraper-logo-favicon/test/index.js

@@ -6,18 +6,7 @@ const { resolve } = require('path')
 const { omit } = require('lodash')
 const fs = require('fs')
 
-const metascraper = require('metascraper')([
-  require('metascraper-amazon')(),
-  require('metascraper-author')(),
-  require('metascraper-date')(),
-  require('metascraper-description')(),
-  require('metascraper-image')(),
-  require('metascraper-logo')(),
-  require('..')(),
-  require('metascraper-publisher')(),
-  require('metascraper-title')(),
-  require('metascraper-url')()
-])
+const metascraper = require('metascraper')([require('..')()])
 
 const readFile = promisify(fs.readFile)
 

packages/metascraper/package.json

@@ -56,7 +56,8 @@
     "cheerio-advanced-selectors": "~2.0.1",
     "lodash": "~4.17.10",
     "p-reduce": "~1.0.0",
-    "sanitize-html": "~1.18.2"
+    "sanitize-html": "~1.18.2",
+    "whoops": "~4.0.1"
   },
   "devDependencies": {
     "clear-module": "latest",

packages/metascraper/src/index.js

@@ -1,20 +1,25 @@
 'use strict'
 
 const { isUrl } = require('@metascraper/helpers')
-const { isEmpty } = require('lodash')
+const whoops = require('whoops')
 
 const mergeRules = require('./merge-rules')
 const loadRules = require('./load-rules')
 const loadHTML = require('./load-html')
 const getData = require('./get-data')
 
+const MetascraperError = whoops('MetascraperError')
+
 module.exports = rules => {
   const loadedRules = loadRules(rules)
   return async ({ url, html, rules: inlineRules } = {}) => {
-    if (!isUrl(url)) throw new TypeError('You need to provide a valid url.')
-    if (isEmpty(html)) {
-      throw new TypeError('You need to provide a valid HTML markup.')
+    if (!isUrl(url)) {
+      throw new MetascraperError({
+        message: 'Need to provide a valid URL.',
+        code: 'INVALID_URL'
+      })
     }
+
     return getData({
       url,
       htmlDom: loadHTML(html),

packages/metascraper/test/unit/interface.js

@@ -2,32 +2,29 @@
 
 const should = require('should')
 
-const metascraper = require('../..')([
-  require('metascraper-author')(),
-  require('metascraper-date')(),
-  require('metascraper-description')(),
-  require('metascraper-video')(),
-  require('metascraper-image')(),
-  require('metascraper-lang')(),
-  require('metascraper-logo')(),
-  require('metascraper-publisher')(),
-  require('metascraper-title')(),
-  require('metascraper-url')()
-])
+const metascraper = require('../..')([])
 
 it('url is required', async () => {
   try {
     await metascraper()
   } catch (err) {
-    should(err).instanceof(TypeError)
+    should(err.name).be.equal('MetascraperError')
+    should(err.code).be.equal('INVALID_URL')
+    should(err.message).be.equal('INVALID_URL, Need to provide a valid URL.')
+  }
+  try {
+    await metascraper({ url: '' })
+  } catch (err) {
+    should(err.name).be.equal('MetascraperError')
+    should(err.code).be.equal('INVALID_URL')
+    should(err.message).be.equal('INVALID_URL, Need to provide a valid URL.')
   }
-})
-
-it('html is required', async () => {
   try {
-    await metascraper({ url: 'https://foo.com' })
+    await metascraper({ url: '/foo' })
   } catch (err) {
-    should(err).instanceof(TypeError)
+    should(err.name).be.equal('MetascraperError')
+    should(err.code).be.equal('INVALID_URL')
+    should(err.message).be.equal('INVALID_URL, Need to provide a valid URL.')
   }
 })