ipushpull development test project

This project is a partially implemented web API that returns historical stock market data (open, close, high and low prices) for stocks in the Indian Nifty50 stock index.

The project is implemented using python 3.9 and the starlette ASGI web framework.

Getting started

• Clone or fork this repository
• Install requirements using pip install -r requirements.txt
• Run the server using python -m nifty
• Access the endpoint at localhost:8888/nifty/stocks/{symbol}

Summary of requirements

1) Return historical price data

Implement the price_data function in __main__.py to return open, close, high and low prices for the requested symbol as JSON records.

For example:

GET /nifty/stocks/tatamotors/

should return

[
    {
        "date": "26/12/2003",
        "open": 435.8,
        "high": 440.5,
        "low": 431.65,
        "close": 438.6
    },
    {
        ...
    }
]

• The data is stored in the file data/nifty50_all.csv
• The endpoint should return one record for each row of data in the file
• Returned data should be sorted by date, most recent data first
• If an invalid symbol is requested, the endpoint should return 400 with an appropriate error message
• The solution should allow the dataset to be updated (e.g. new data added) without restarting the app

2) Allow the price data to be filtered by year

Add a query parameter to the endpoint so that calling apps can request data for a single year.

For example:

GET /nifty/stocks/tatamotors/?year=2017

• This should only return rows for the specified year (and symbol)
• If there is no data for the specified year, an empty list should be returned
• If the year is invalid, the endpoint should return 400 and an appropriate error message

3) Extend the endpoint to allow new data to be added

• The endpoint should only accept JSON and allow prices for one or more days to be added to the dataset
• It should only allow new data to be added, it should not allow an existing value to be updated
• Any subset of OPEN, CLOSE, HIGH, LOW should be accepted - no other price-types are acceptable
• Updates should be validated as follows:
  • Dates must be in the format DD/MM/YYYY
  • Prices must be within 1 standard deviation of the prior 50 values for that combination of symbol and price-type
• New data should be persisted and immediately accessible via GET

Additional information

• You should use python 3.9 or above
• You may use any appropriate open source libs as part of your solution
• Please upload your project to github.com or similar and provide a link
• If you have questions please email support@ipushpull.com

Thanks for taking our test!