Implement a library and executable program in one file "rates.py"
that will convert currencies.

The program will use "rates.convert()" function and will accept these
parameters:

    -b --base: the base currency to convert from, required
    -c --currency: the currency to convert to, required,
                   can be specified multiple times
    -d --date: the date of the exchange rates in format YYYY-MM-DD,
               optional, if it's not specified use the current date

    N: number which will be converted, required, can be specified
       multiple times

and it will print the result to the stdout in this format.

Some examples:

$ ./rates.py -b EUR -c USD -d 2019-04-20 10
10.00 EUR is 11.25 USD

$ ./rates.py -b EUR -c USD -c CNY -d 2019-04-20 10 20
10.00 EUR is 11.25 USD
20.00 EUR is 22.50 USD
10.00 EUR is 75.45 CNY
20.00 EUR is 150.89 CNY

It should be possible to import the "rates.py" file as well as run
it directly.

In order to validate this exercise you should install python3-pytest
and run "./exercise.py".

As a data source for exchange rates you can use https://exchangeratesapi.io/
which has simple API.  To get rates for specific date you will simply get:

    https://api.exchangeratesapi.io/2010-01-12

There are optional parameters which you will need to use:

    base: the base currency
    symbols: the currencies to get exchange rates from base

The URL can look like this:

    https://api.exchangeratesapi.io/2010-01-12?base=USD&symbols=EUR

Hint: you can use requests module to do the HTTP communication.

As the webpage states the library should cache the downloaded rates,
it's good enough to cache it only in memory, as a bonus you can do file
caching.