Quick start

Overview

Short overview about how to start the new search application from A-Z quickly. You need heroku account and some free time.

Deploy to Heroku

Open installator on Heroku and run it in about one minute. Heroku will install dependencies like Elasticsearch or MongoDB automatically.

Installing on heroku

Prepare JSON file

Prepare your JSON with list of something (movies, songs, restaurants, etc). You can use example one with movies https://raw.githubusercontent.com/itemsapi/itemsapi-example-data/master/items/movies-processed.json.

It is good if you have image field because application will look nicer. If you provide unique permalink field for each items your app will have better URL's so then better SEO. It's also worth noting name field is responsible for autocomplete.

JSON file with list of movies

Upload JSON data

Put your JSON url or JSON content into website configuration formular and submit.

Installation form.

See your application

Once the uploading JSON is successful your application is created. You have very fast application with search, filters, items list, item page and also showing similar items. Application is ready for further configuration and customization.

App overview. List of the items.

Login to admin panel

Open http://localhost:4000/admin/login and log-in to administration panel with admin and itemsapi credentials.

Login to administration panel

Administration panel

Once you login successfully you will have access to admin panel. This is a place for all management and app configuration. You can manage there items, configure filters, improve search behaviour or even set google analytics.

Administration panel

Configure filters

Application generator is not perfect so it can generate unnecessary filters. Go to http://localhost:4000/admin/collections/edit. In Aggregations box you can delete them. You can also configure them in Edit collection box i.e. changing title, size or position on website.

Configure filters (aggregations)

Configure autocomplete

Autocomplete should work from start. The requirement is the name field in your dataset.

Fast autocomplete

Configure similar items

Showing similar items might be great for discovering new things. To show items recommendations (i.e. similar movies by tags or by actors) go to admin settings and set recommendation field. Right now similar items visually looks like this:

Similar items

Customize Template

To customize templates just go into /views/basic folder and open one of the file. These files are responsible for how the pages look like. The template engine is written in nunjucks (javascript counterpart of jinja2 and twig) and it is flexible and easy to learn.

The javascript and css files are located in /assets/ directory.

Nunjucks template. Twig and Jinja2 counterpart
Was this page helpful ?