jrouaix / rusty-test

test in Rust

cargo run -- webserver
http://localhost:4242/filter?csvUri=https%3A%2F%2Fbeezupcdn.blob.core.windows.net%2Frecruitment%2Fbigfile.csv


target\debug\bztest.exe webserver
https://beezupcdn.blob.core.windows.net/recruitment/bigfile.csv

type bigfile.csv | target\debug\bztest.exe process -o json | grep 1234
type bigfile.csv | target\debug\bztest.exe process -o text | grep 1234

set before=%time%
type bigfile.csv | target\debug\bztest.exe process -o json > NUL
echo timed from %before% to %time%




cargo build --release

set before=%time%
type bigfile.csv | target\release\bztest.exe process -o text > NUL
echo timed from %before% to %time%

Generate documentation

cargo doc --open

Test

Introduction

• You are required to use the .NET Framework, using either .NET Core or .NET 4.5 or above.
• You are expected to use NuGet packages.
• You are required to create a local git repository.
• You are required to commit at least once at then end of each step (you may create as many intermediary as you wish, but must clearly indicate which commits are the final version for each step).

Evalutation methods and critera

• Each step will be reviewed separately. Do not skip any steps. Make your commits clear so we can easily find the final version of each step.
• You will be evaluated on your solutions:
  • code beauty (easy to read and understand),
  • efficiency (power, memory consumption, etc.),
  • sturdiness,
  • extensibility,
  • testing methods and coverage.

Before your start

• Some formulations in the specification are vague and generic on purpose. There is no 'perfect' or 'single' solution. We expect you to be able to explain and justify your choices if asked.
• We expect you to spend "few hours" on the test, if you cannot complete in a reasonable time (from your appreciation), you still can send what you achieved. Perhaps it's a good start.
• Feel free to cut out parts of the specs as long as you are confident enough in the way to implement it in your architecture and it feels redundant with another part. We'll ask you about missing parts.

Read the instructions again, take a break . . . and . . . Go !

Part 1 - Console

Step 1

Create a console application that reads a big CSV file (available here). For each row where the sum of integer value of ColumnC and integer value of ColumnD is greater than 100, print out the concatenation of ColumnA & ColumnB. Rows that can't be processed corretly shall by skipped.

Step 2

Replace the output from the previous step. Write a big JSON array of objects for the previous matched lines:

• Rows that can be processed correctly with C + D > 100: { "lineNumber": <FILE_LINE_NUMBER>, "type": "ok", "concatAB": "<PREVIOUS_AB_CONCAT>", "sumCD": <PREVIOUS_CD_SUM> }
• Rows that can't be processed correctly : { "lineNumber": <FILE_LINE_NUMBER>, "type": "error", "errorMessage": <ERROR_MESSAGE> }

Step 3

Output format (plain text, JSON, etc.) can now be configured by command line arguments. Integrate at least one new format.

Part 2 - API

Step 1

Create an Api project that can receive HTTP POST and GET requests on /filter?csvUri={<CSV_URI>}. The payload response should be a json object as described in Part1

Step 2

Now this api should be able to ouput Text, Json, CSV, XML ... Update the API to be able to return JSON or XML formats for the payload response.

Final (this step should be the easy one)

We are constantly trying to improve our test. Any constructive feedback (too long, too easy, too vague, missing parts, you wanted to show something else ...) on this test is welcome.