KPI Hub Server

Vapor server for KPIHub iOS app which is used for:

1. Parsing rozklad.kpi.ua or API requests to schedule.kpi.ua
2. Parsing campus.kpi.ua.

Hosing:

Server is hosted by digital ocean cloud.

Domain: kpihub.xyz

Used technologies

• Vapor
• vapor-routing (by poinfree)
• fluent-postgres (for storing groups)
• Docker, docker-compose
• Github Actions, Secrets

Requests

Rozklad

Receiving all groups:

GET /api/groups/all

Search group by name:

GET /api/groups/search?name=$(groupName)

Get group lessons:

GET /api/group/$(ID)/lessons

Campus

Student info:

GET /api/campus/userInfo?username=$(username)&password=$(password)

Student study sheet:

GET /api/campus/studySheet?username=$(username)&password=$(password)

How it works?

The main thing that the server does is parisng.

RozkladV2

Looks like rozklad.kpi.ua is migrating to new site schedule.kpi.ua. As rozklad.kpi.ua is not reacheable now, I created a wrapper RozkladV2 above new API.

RozkladV1

Every 28th on every mounth crone script updates group list. This is a rather labor-intensive task, and if you do it too often, the rozklad.kpi.ua will ban your IP.

When user request lessons for particular group, server goes to rozklad.kpi.ua and parse lessons live.

Campus

Every request for campus-related staff need to be provieded with campus credentials.

Server do not store any credential information. So that’s why u need to pass credentials for every request.

Server use this data for receiving Bearer token and cookies.

Routing

At the heart of the approach for routing was taken vapor-routing by pointfree. This library gives an opportunity to move routing-related code to separate library and use it in iOS application and server.

Parsing

This server use swift-parsing library for parsing HTML pages.

Parsers:

• GroupParser

  Receive groups IDs from HTML pages

• LessonsParser

  Extract lessons from two schedule tables (first and second weeks)

• StudySheetActivitiesParser

  Parse studysheet table for all years of study

• StudySheetLessonsParser

  Parse marks from a specific subject table