Giter Site home page Giter Site logo

historical-bank-ruby's Introduction

historical-bank-ruby

Gem Version Build Status

Description

This gem provides a bank that can serve historical rates, contrary to most bank implementations that provide only current market rates. Fixer is used as the provider of the rates. Fixer is based on EU Central Bank As the HTTP requests can add latency to your calls, a RatesStore (cache) based on Redis was added, making it super-fast.

You can use it as your default bank and keep calling the standard money gem methods (Money#exchange_to, Bank#exchange_with). On top of that, we've added a few more methods that allow accessing historical rates (Money#exchange_to_historical, Bank#exchange_with_historical).

Base currency

An exchange rate has a base currency and a quote (or counter) currency. More specifically, it is the price of 1 unit of base currency in the quote currency. For example, if base currency is EUR, quote currency is USD, and the rate is 1.25, this means that 1 EUR is equal to 1.25 USD.

All the rates fetched and cached by this Bank are relative to a single base currency which is defined in the configuration block. This helps to optimize fetching and caching rates.

Default base currency is EUR, but it can be changed in the config.

Dates, times, timezones

The timezone used throughout this gem is the UTC timezone.

All processed rates (fetched from Fixer, added manually, cached in Redis) are considered to be the closing (end of day) rates for their associated dates in UTC. For example, when we have a cached rate of EUR->USD on January 10th 2017 with value 1.25, this means that 1 EUR was equivalent to 1.25 USD on January 10th at 23:59:59. This is the historical rate that the bank will use for exchanging EUR with USD on that date. Consequently, a rate for a certain date fetched from Fixer becomes available at 00:00 UTC on date+1.

For convenience, methods that accept Dates as arguments can accept Times as well. When a Time is used, it is first converted into the UTC-equivalent Date, and method is executed as if that Date was passed instead. For example, when the Time 2017-01-10 02:50:00 +04:00 is passed as argument to #exchange_with_historical,

from_money = Money.new(100_00, 'EUR')
to_currency = 'USD'

# 2017-01-10 02:50:00 +0400
bank.exchange_with_historical(from_money, to_currency, Time.new(2017, 1, 10, 2, 50, 0, '+04:00'))
# => #<Money fractional:10585 currency:USD>

it is equivalent to passing the Date 2017-01-09

# 2017-01-09
bank.exchange_with_historical(from_money, to_currency, Date.new(2017, 1, 9))
# => #<Money fractional:10585 currency:USD>

Caching

We've implemented 2 layers of caching in order to obliterate latency! First layer is memory (instance variable in the bank object), and second is Redis. If desired rate is not found in memory, the bank tries to look it up in Redis. If that fails too, a request to Fixer is made.

When we fetch rates from Fixer, they are cached in Redis and memory too. Similarly, when the rate is found in Redis, it is again cached in memory.

Caching diagram

Pretty simple and fast!

Singleton

The bank follows the Singleton pattern, as it inherits from money gem's Money::Bank::Base. This also helps preserve the memory cache across calls. Don't worry, the bank is thread-safe!

Requirements

  • OpenExchangeRates Enterprise or Unlimited plan - It is needed for calling the /time-series.json endpoint which serves historical rates.
  • Redis >= 2.8 (versions of Redis earlier than 2.8 may also work fine, but this gem has only been tested with 2.8 and above.)
  • Ruby >= 2.0

Installation

gem install historical-bank

Alternatively, if you're using bundler, you can add

gem 'historical-bank'

to your Gemfile and run bundle install

Usage

Example scripts demonstrating all functionality can be found in examples/.

Configuration

Money::Bank::Historical.configure do |config|
  # (optional) currency relative to which all the rates are stored (default: EUR)
  config.base_currency = Money::Currency.new('USD')

  # (optional) the URL of the Redis server (default: 'redis://localhost:6379')
  config.redis_url = 'redis://localhost:6379'

  # (optional) Redis namespace to prefix all keys (default: 'currency')
  config.redis_namespace = 'currency_historical_gem'

  # (optional) set a timeout for the Fixer calls (default: 15 seconds)
  config.timeout = 20
end

Rails

In Rails, config should be set inside an initializer (config/initializers). If you have the money-rails gem installed, you can add this on top of the config/initializers/money.rb file.

# config/initializers/money.rb

require 'money/bank/historical'

Money::Bank::Historical.configure do |config|
  # ....
end


MoneyRails.configure do |config|
  # ...

  # if you want to set it as the default bank
  config.default_bank = Money::Bank::Historical.instance

  # ...
end

Sinatra

In a Sinatra app (or any other kind of app), simply set the config before you call Money::Bank::Historical.instance.

# app.rb

require 'money/bank/historical'

Money::Bank::Historical.configure do |config|
  # ....
end

class App < Sinatra::Base

  configure do
    # if you want to set it as the default bank
    Money.default_bank = Money::Bank::Historical.instance
  end

  # ...
end

Dates

The minimum date for which we can fetch rates is January 1st 1999. This limitation is set by the OpenExchangeRates API.

However, if you want to overcome these limitations manually, you can add past (and even future!) rates using Historical#add_rate and #add_rates.

Exchange

You can use the bank object for the exchange

from_money = Money.new(100_00, 'EUR')
to_currency = 'GBP'

bank = Money::Bank::Historical.instance

# exchange money with rates from December 10th 2016
bank.exchange_with_historical(from_money, to_currency, Date.new(2016, 12, 10))
# => #<Money fractional:8399 currency:GBP>

# can also pass a Time/DateTime object, it's converted into the respective UTC Date
bank.exchange_with_historical(from_money, to_currency, Time.utc(2016, 10, 2, 11, 0, 0))
# => #<Money fractional:8691 currency:GBP>

Or perform it directly on the Money object

from_money.exchange_to_historical(to_currency, Date.new(2016, 12, 10))
# => #<Money fractional:8399 currency:GBP>

Bank#exchange_with and Money#exchange_to can still be used. In this case, recent rates are needed, so yesterday's closing rates are used for the calculation

bank.exchange_with(from_money, to_currency)

# set the default bank and create a new Money object that will use it
Money.default_bank = Money::Bank::Historical.instance
from_money = Money.new(100_00, 'EUR')
from_money.exchange_to(to_currency)

Adding and retrieving rates

Adding rates will not be needed in most cases as the rates are fetched from Fixer. However, #add_rate and #get_rate were implemented in order to conform to the Bank API. An extra #add_rates method was implemented for setting rates in bulk.

#add_rate and add_rates can prove quite when testing, as you can't afford HTTP requests there. Only thing you need to do is initialize the bank, and add some default rates before your tests run.

Get single rate

#get_rate accepts both ISO strings and Money::Currency objects

bank.get_rate(Money::Currency.new('GBP'), 'CAD', Date.new(2016, 10, 1))
# => #<BigDecimal:7fd39fd2cb78,'0.1703941289 451827243E1',27(45)>

Getting without a datetime will return yesterday's closing rate, e.g. bank.get_rate('CAD', 'CNY').

Add single rate

#add_rate adds a single rate to the Redis cache. It accepts both ISO strings and Money::Currency objects. Added rates should be relative to the base currency.

date = Date.new(2016, 5, 18)

bank.add_rate('EUR', 'USD', 1.2, date)
bank.add_rate(Money::Currency.new('USD'), Money::Currency.new('GBP'), 0.8, date)

# 100 EUR = 100 * 1.2 USD = 100 * 1.2 * 0.8 GBP = 96 GBP
bank.exchange_with_historical(from_money, to_currency, date)
# => #<Money fractional:9600 currency:GBP>

Adding without a datetime will set the rate to yesterday's closing rate

bank.add_rate('EUR', 'USD', 1.4)
bank.add_rate(Money::Currency.new('USD'), Money::Currency.new('GBP'), 0.6)

# 100 EUR = 100 * 1.4 USD = 100 * 1.4 * 0.6 GBP = 84 GBP
bank.exchange_with(from_money, to_currency)
# => #<Money fractional:8400 currency:GBP>

Trying to add a rate that is not relative to the base currency will fail. This is because all cached rates are relative to the base currency.

bank.add_rate('EUR', 'GBP', 0.96, date)
# ArgumentError: `from_currency` (EUR) or `to_currency` (GBP) should match the base currency USD

Add rates in bulk

#add_rates can be used to add multiple historical rates (relative to the base currency) to the Redis cache.

rates = {
  'EUR' => {
    '2015-09-10' => 0.11, # 1 USD = 0.11 EUR
    '2015-09-11' => 0.22
  },
  'GBP' => {
    '2015-09-10' => 0.44, # 1 USD = 0.44 GBP
    '2015-09-11' => 0.55
  }
}
bank.add_rates(rates)

# 100 EUR = 100 / 0.11 USD = 100 / 0.11 * 0.44 GBP = 400 GBP
bank.exchange_with_historical(from_money, to_currency, Date.new(2015, 9, 10))
# => #<Money fractional:40000 currency:GBP>

historical-bank-ruby's People

Contributors

jeopard avatar gugaiz avatar

Watchers

James Cloos avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.