SERP Parser
serp-parser is small lib writen in typescript used to extract search engine rank position from the html.
Instalation
npm i serp-parser
yarn add serp-parser
Usage - Google SERP extraction
GoogleSERP accepts both html that is extracted with any headless browser lib (puppeteer, phantomjs...) that have enabled javascript as well as html page structure from no-js-enabled requests from for example request lib. For full enabled js html we use GoogleSERP class, and for nojs pages GoogleNojsSERP class.
With html from headless browser we use full GoogleSERP parser
import { GoogleSERP } from 'serp-parser'
const parser = new GoogleSERP(html);
console.dir(parser.serp);
Or on es5 with request lib, we get nojs Google results, so we use GoogleNojsSERP parser that is separate class in the lib
var request = require("request")
var sp = require("serp-parser")
request('https://www.google.com/search?q=google', function (error, response, html) {
if (!error && response.statusCode == 200) {
parser = new sp.GoogleNojsSERP(html);
console.dir(parser.serp);
}
});
It will return serp object with array of results with domain, position, title, url, cached url, similar url, link type, sitelinks and snippet
{
"keyword: "google",
"totalResults": 15860000000,
"timeTaken": 0.61,
"currentPage": 1,
"pagination": [
{ page: 1,
path: "" },
{ page: 2,
path: "/search?q=google&safe=off&gl=US&pws=0&nfpr=1&ei=N1QvXKbhOLCC5wLlvLa4Dg&start=10&sa=N&ved=0ahUKEwjm2Mn2ktTfAhUwwVkKHWWeDecQ8tMDCOwB" },
...
],
"videos": [
{ title: "The Matrix YouTube Movies Science Fiction - 1999 $ From $3.99",
sitelink: "https://www.youtube.com/watch?v=3DfOTKGvtOM",
date: 2018-10-28T23:00:00.000Z,
source: "YouTube",
channel: "Warner Movies On Demand",
videoDuration: "2:23" },
...
],
"thumbnailGroups": [
{ "heading": "Organization software",
"thumbnails:": [ {
"sitelink": "/search?safe=off&gl=US&pws=0&nfpr=1&q=Microsoft&stick=H4sIAAAAAAAAAONgFuLUz9U3MDFNNk9S4gAzi8tMtGSyk630k0qLM_NSi4v1M4uLS1OLrIozU1LLEyuLVzGKp1n5F6Un5mVWJZZk5ucpFOenlZQnFqUCAMQud6xPAAAA&sa=X&ved=2ahUKEwjm2Mn2ktTfAhUwwVkKHWWeDecQxA0wHXoECAQQBQ",
"title": "Microsoft Corporation"
},
...
]
},
...
],
"organic": [
{
"domain": "www.google.com",
"position": 1,
"title": "Google",
"url": "https://www.google.com/",
"cachedUrl": "https://webcache.googleusercontent.com/search?q=cache:y14FcUQOGl4J:https://www.google.com/+&cd=1&hl=en&ct=clnk&gl=us",
"similarUrl": "/search?safe=off&gl=US&pws=0&nfpr=1&q=related:https://www.google.com/+google&tbo=1&sa=X&ved=2ahUKEwjm2Mn2ktTfAhUwwVkKHWWeDecQHzAAegQIARAG",
"linkType": "HOME",
"sitelinks": [
{ "title": "Google Docs",
"snippet": "Google Docs brings your documents to life with smart ...",
"type": "card" },
{ "title": "Google News",
"snippet": "Comprehensive up-to-date news coverage, aggregated from ...",
"type": "card" },
...
],
"snippet": "Settings Your data in Search Help Send feedback. AllImages. Account · Assistant · Search · Maps · YouTube · Play · News · Gmail · Contacts · Drive · Calendar."
},
{
"domain": "www.google.org",
"position": 2,
"title": "Google.org: Home",
"url": "https://www.google.org/",
"cachedUrl": "https://webcache.googleusercontent.com/search?q=cache:Nm9ycLj-SKoJ:https://www.google.org/+&cd=24&hl=en&ct=clnk&gl=us",
"similarUrl": "/search?safe=off&gl=US&pws=0&nfpr=1&q=related:https://www.google.org/+google&tbo=1&sa=X&ved=2ahUKEwjm2Mn2ktTfAhUwwVkKHWWeDecQHzAXegQIDBAF",
"linkType": "HOME",
"snippet": "Data-driven, human-focused philanthropy powered by Google. We bring the best of Google to innovative nonprofits that are committed to creating a world that..."
},
...
],
"relatedKeywords": [
{ keyword: google search,
path: "/search?safe=off&gl=US&pws=0&nfpr=1&q=google+search&sa=X&ved=2ahUKEwjm2Mn2ktTfAhUwwVkKHWWeDecQ1QIoAHoECA0QAQ" },
{ keyword: google account,
path: "/search?safe=off&gl=US&pws=0&nfpr=1&q=google+account&sa=X&ved=2ahUKEwjm2Mn2ktTfAhUwwVkKHWWeDecQ1QIoAXoECA0QAg" },
...
]
}
Usage - Bing SERP extraction
Note: Only BingNojsSerp is implemented so far.
BingSERP works the same as GoogleSerp. It accepts both html that is extracted with any headless browser lib (puppeteer, phantomjs...) that have enabled javascript as well as html page structure from no-js-enabled requests from for example request lib. For full enabled js html we use BingSERP class, and for nojs pages BingNojsSERP class.
With html from headless browser we use full BingSERP parser
import { BingNojsSERP } from 'serp-parser'
const parser = new BingNojsSERP(html);
console.dir(parser.serp);
Or on es5 with request lib, we get nojs Bing results, so we use BingNojsSERP parser that is separate class in the lib
var request = require("request")
var sp = require("serp-parser")
request('https://www.bing.com/search?q=bing', function (error, response, html) {
if (!error && response.statusCode == 200) {
parser = new sp.BingNojsSERP(html);
console.dir(parser.serp);
}
});
Parse options - specify what features to parse
On each module you can specify the parse options as a second parameter to only parse SERP features that you want, for GoogleSERP here are all the options:
// Options for GoogleSERP
#DEF_OPTIONS = {
organic: true,
related: true,
pagination: true,
ads: true,
hotels: true,
videos: true,
thumbnails: true,
shop: true,
stories: true,
locals: true,
};
// Options for GoogleNojsSERP and BingNojsSERP
#DEF_OPTIONS = {
organic: true,
related: true,
ads: true,
hotels: true,
};
Usage: In this example only organic feature will be parsed
import { GoogleSERP } from 'serp-parser'
const parser = new GoogleSERP(html, {organic: true});
console.dir(parser.serp);
Roadmap
We are working on enriching parsed data to grab all existing and new SERP features from Google search page results, as well as knowlege graph, ads, and any related info. Also we will add more Search engines along the way.
Anyone willing to help - please submit issues, feature requests, fork away and send PR's