A platform for browsing the top free and paid android games and engage in real-time game-specific discussion with players around the world.
Firebase, AngularJS, AngularFire, Java
###Introduction There are a wide variety of games available on the Google Play Store. But there is no convenient way for gamers to interact with other players of the games they're interested in. The review section is currently the main way by which the gamers interact.
Through this project, I provide a platform for the gamers to interact in real time. This platform has been expanded to work with all the top games from Google Play Store.
The users would login to the application using Google authentication and browse through the list of top mobile games. The games are listed according to the ranking in the Play Store. The users can select a game and view more details about it like the developer info, user rating, description, price, logo, etc. They can also interact with other players in real time in the messages section.
###Challenges ####Unavailability of API Official Google Play API does not support searching and retrieving of application information. I tried using unofficial python n java APIs like egirault API and Android Market API. These libraries use depreciated functions and hence, do not work anymore. 42matters is the only solution that seems to offer the required features, but is a paid service.
So, I wrote a crawler that scrapes the Play Store for top free and paid games. It extracts all the required information about the games and stores it in the backend. This crawler was developed in Java using jSoup library.
####Realtime Updates Finding the right backend technology that allows real-time updates to facilitate live discussion was a challenge.
For backend architecture, I considered various options. Parse was a good option but it does not support real time updates. socket.io was perfect for chat message transfer, but it does not store data or state. Hence, I finalized Firebase as the backend because it supports realtime data transfer and has a nice AngularJS API with three-way data binding.
###Approach
####Data Collection The Java crawler accesses Google Play Store game section and extracts all the available information about the top games like display image, title, rating, rank, description, developer information and price. This information is then stored in Firebase using its Java API. Since the Java program would terminate before the data is completely stored in the firebase instance, I keep the connection alive by setting the running thread to sleep mode.
####Back-end A firebase data store instance is used to store the information about the games as well as the chat messages. To reduce impact to the end user and to minimize the page-loading time, data is denormalized and stored in separate sections called messages, details and gamesList. This way, only the required data is downloaded to the client machine.
####Front-end The web application was developed using AngularJS, AngularFire and Bootstrap. AngularFire along with AngularJS provides three-way binding of data and with this, realtime synchronization of chat messages became possible. The application was then hosted to Firebase application server using Firebase CLI.\
The user interface was developed using Twitter Bootstrap and the design is responsive in nature and adapts to the browser window size. This ensures that regardless of the monitor size, users can have a good user experience with the application.