Covid19 data task for Cognizant qa engineer/junior developer position

A simple ephemeral(runt-output-terminate) program to fetch and analyse covid-19 data as per instructions below:

Use https://rapidapi.com/, "COVID-19 data" resource. Register to get API key, don't forget subscribe for the resource (free plan).

1.   Fetch Covid-19 data, 
   

https://covid-19-data.p.rapidapi.com/help/countries From https://rapidapi.com/Gramzivi/api/covid-19-data?endpoint=apiendpoint_80bfb292-72a4-485b-8e76-2900bd8246bd (Country code, Contry name, Longitute, Latitude)

2. Sort countries by Latitude Descending order. Select top 15 Countries;¬

3. Get last 10 days data from selected 15 countries, using endpoint. https://covid-19-data.p.rapidapi.com/report/country/code From https://rapidapi.com/Gramzivi/api/covid-19-data?endpoint=apiendpoint_a0a3b806-6a57-4726-8fe7-6883b6dbaa9e

4. Store daily data (Country code, date, confirmed, critical, deaths, recovered)

5. Select stored entries for every country, one day data, that has worst (biggest) death toll;

6. Convert it to XML. Email in html email body and xml-generated file attached.

Tech used: API: rapidapi.com covid-19-data okhttp - for fetching the data Jackson Object Mapper - for deserialization of json into pojos JAXB - for writting pojos to xml Java mail - for automated emails and attachments Xsl/Xslt - wrote a stylesheet and attempted transformation of xml to html.

Reflection: API constraints: rapidapi.com - at basic subscription, the COVID-19-data api, allows 1 request per second. This was not a problem for the fetch required in task 1, but it had serious implications for task 3. Since the endpoint giving daily report for all countries was locked for basic subscription, the only way was to perform multiple fetches (per each country, 10 fetches to get last 10 days data) as per method: getDailyDataDataForSelectedCountries(). This was not ideal too, as the api limits a basic user to 1 request per second.

Since the task was written as an ephemeral - run, output, terminate application, the simple workaround was to make the program sleep for 1.5s after each request Thread.sleep(1500) - see method getDailyDataDataForSelectedCountries(). This approach also required to fetch synchronously, because asynch fetches would make the program break. So to get 10 days data, for 15 countries results in 150 fetches - 1.5s each. Hence the app works, but is clunky - taking more than a minute to output results.

If the app was written to run on web-server, the fetching would have to be asynchronous, but given the 1request/second limitation of the basic subscription, data loading would probably take too long. So if this was a project in a commercial setting, most likely one would either have to upgrade to better subscription with rapidapi.com, or just find another api.

Pojos to XML to HTML: Managed to write xmls based on pojos. However, I am quite new to the format, and only used it for configuration purposes in the past, and so the transformation of it was challenging - I managed to write an xsl stylesheet to achieve the desired table output in html, this works in a browser, but I struggled to achieve the actual transformation to html document. (commented code 5.3 in TaskApplication)

Auto-emails: The commented code (5.4) in Task application also shows a workflow to compose an email and attach three generated xmls from /results directory. Feel free to comment out and try it with your details. Gmail likely to require a setting to allow less secure apps to connect.

For simplicity, I will provide a link to this repo in email, as well as three xml files that were generated using this app, as well as an xsl stylesheet. But feel free to run it, baring in mind it takes about 2-3 min to fetch daily data due to api limitations.