Skip to content

Latest commit

 

History

History
101 lines (74 loc) · 3.22 KB

README.md

File metadata and controls

101 lines (74 loc) · 3.22 KB
Raw

Mímr

Mímr (from the norse god Mímir) is a NodeJS script to track and mail changes of specific elements on a web-site using CSS selectors.

Installation

This script requires node>= 6, npm and yarn(secure npm alternative). You have to clone this repository, because the package is not yet available in the npm registry.

$ git clone https://github.com/eppfel/mimr.git

Inside the folder you install all dependecies with:

$ yarn install

Usage

Run with Node

You can execute this script directly with node by calling index.js :

$ node index.js

All results will be stored in store.json, so, if you call the script again, it will only show you new findings.

But first you need to configure a config file:

Configuration in config.json

The script tracks page elements specified in the config.json in the pages array:

{
  "pages": [{
    "name": "Github Trending Repositories",
    "url": "https://github.com/trending",
    "selector": ".explore-content h3 a"
  }]
}

Each page needs the following parameters:

Parameter Value
name a name for each page (completely free)
url an url to an html page to be tracked
selector a CSS selector to retrieve a set of elements from the page

Additionally you can specifiy mail credentials to send the results via mail. Just leave the mail setting, if you do not want to use it.

{
  "mail": {
    "host": "smtp.example.org",
    "username": "[email protected]",
    "password": "***",
    "from": "[email protected]",
    "to": "[email protected]"
	}
}

Mimir uses nodemailer. See their documentation for more details on the parameters.

You'll find an example configuration in config.example.json.

Use as module

Include it in your projects folder or in your node_modules folder to import it via require().

const mimr = require('./mimr')

const pages = [{"name": "Github Trending Repositories", "url": "https://github.com/trending", "selector": ".explore-content h3 a"}]
const oldFindings = []

mimr(pages, oldFindings).then(([err, newFindings, allFindings]) => {
  for (let page of newFindings) {
    console.log(page.finds)
  }
}).catch(console.error)

mimr takes an array with pages to track (see configuration) and optionally findings from a previous run to compare against.

It returns a promise that resolves an array with three elements:

  • err: If some errors occur during extraction this will not directly break execution
  • allFindings: These are all findings for each page provided, with an added attribute finds, which as an array of Strings or Objects with of found elements.
  • newFindings: These are the findings for each page provided as allFindings minus the elements provided by oldFindings.

Roadmap

  • Impliment this as a service with internal scheduling to support features like a results web server or RSS Feed
  • More comprehensive results with
    • extracting any HTML attributes
    • define custom templates for your findings
  • Tests
  • Browserify – to use from directly in the browser or a browser extension