sbt run
should suffice.
sbt unidoc
should generate some rudimentary documentation, which provides usage instructions, among other things. Or just check app/rootdoc.txt
.
The rest of this file is left unchanged.
Hello future colleague! In this assessment, we request that you implement an HTTP application that converts CSV data sets to JSON data sets. The application will provide the following endpoints:
Create a task with URI pointing to CSV dataset which will be converted to Json and reuturns taskId.
- 2 tasks can be run at the same time
- the task is executed immediately if
running-tasks < 2
Return informations about the task:
- lines processed
- avg lines processed (count/sec)
- state (
SCHEDULED/RUNNING/DONE/FAILED/CANCELED
) - result (uri where the JSON file can be downloaded)
Keep the connection open until the task isn't in a terminal state and send the updated response every 2 seconds.
Tasks in SCHEDULED
or RUNNING
state can be canceled.
Endpoint providing generated JSON files.
- Keep the state only in memory except generated json files.
- Take into account that input files don't have to fit into memory.
CSV datasets for testing purposes can be found at the following links:
Feel free to ask any questions through email or github.
Good Luck!