A Dealer For the People

It's a tool to scrape data from dealerrater.com.

Features

It can extract and print reviews on the console, by default, from the Dealer McKaig Chevrolet Buick - A Dealer For The People.

How it works?

We'll fetch the "overly positive" reviews. Each review has the following data with weight and rules:

Data	Description	Priority	Rule
Recommended?	if a customer recommends the Dealer	1	When recommended is 1 point otherwise 0 points
Dealership Rating	A general rating from the customer	2	The value is used as a point, for example, 4.2
Features Rating	A rating given for each Dealer's feature	3	The sum of all values is used as a point, for example, 2.1 + 2.1
Content	Customer opinion	4	When has good words, sum 1 point but when has a bad word, decrease 1 point
Employee Rating	Employees involved on deal	5	The sum of all values is used as a point, for example, 2.1 + 2.1

Good and Bad words

The good and bad words are detected in the content of the customer's review. You can find the words in the .env file.

Docker Setup

To avoid install the Ruby on your machine, you can install and use Docker.

Now let's create an image containing the application:

docker build . -t dealer_rater

Run

Now we can run the code from this image:

docker run -t dealer_rater ruby scraper.rb --display=3 --pages=5 --slug=McKaig-Chevrolet-Buick-A-Dealer-For-The-People-dealer-reviews-23685 --source=dealer

Options

Attribute	Description
pages	Quantity of page to be scraped
display	Number of the best reviews to be displayed
slug	The slug of the dealer
source	The page type

Test

To make sure all test is passing, run:

docker run dealer_rater bundle exec rspec