Application is designed to process employees data and generate corresponding reports.

File examples are represented in the following folder:

hubrick-backend-challenge/data

There are several assumptions were made in this challenge:

• Assume that files can contains wrong data (also in wrong format), duplicate data
• If Employee was not properly parsed (inc. cases as wrong department, no/bad age), this record are taken out of report generation and information about it put into errors.log file
• Some handy 3rd party libraries were not used by the challenge conditions and were replaced with simplified analogs(self-made ones) as ex. com.fasterxml.jackson.dataformat:jackson-dataformat-csv is nice to work with csv
• Console is not used to show all information about application work. Instead of that, errors are written in separate file /out/errors.log
• Every developer know what percentile means and how to work with it. But what's about the proper algorthytm, how to calculate it. Honestly I "googled" for it and found out at least several variations which provide close but still a bit different results. I've choosen on of them. Implementation is under ReportService.percentile

Comments to code:

• Somewhere in code you can find static imports, which somebody like, somebody not really. Sometimes I find its handy to make code cleaner.
• I used to use javax.annotation.Nonnull/Nullable to provide method/api contracts. but there are not a part of jdk.
• Also I was missed about lombok and guava libraries, which could make code more nicer.
• In some places I've used BigDecimal for more precise results and comparing double values. I also notices that not everybody is cool with it.
• I've put some comments for better readability

Additional Features:

• errors logged in separate file
• you can use ~ in path parameter to provide relative to user home folder path.
• ReportService designed in the way to be able to generate reports with different groupings and aggregating different values.
• ReportService also allow calculte different percentiles.

To build application run:

mvn clean install

To execute application run:

java -jar target/hubrick-challenge.jar <PATH-TO-FOLDER-WITH-DATA>