How to divide a time period into multiple periods

Question

I have a time period in two variables, from start_date to end_date. Is there an easy way to split it into a smaller periods, rounding the values up.

Here's an example:

I have period from 15.01.2016 to 10.06.2016. I want to split it into months, so I will have six periods:

01.01.2016 - 31.01.2016  
01.02.2016 - 31.02.2016   
01.03.2016 - 31.03.2016  
01.04.2016 - 31.04.2016  
01.05.2016 - 31.05.2016  
01.06.2016 - 31.06.2016 

I want to include the time between 01.01.2016 and 15.01.2016 regardless of the fact that it is not in the original period.

I've been looking for some ideas, but at this time it seems that the only way is using the start date and iterating using DateAndTime::Calculations to determine the borders of intervals, until hitting the end date.

Answer 1

from = Date.parse('15.01.2016')
to   = Date.parse('10.06.2016')
(from..to).group_by(&:month).map do |group|
  group.last.first.beginning_of_month..group.last.last.end_of_month
end
# => [Fri, 01 Jan 2016..Sun, 31 Jan 2016,
#  Mon, 01 Feb 2016..Mon, 29 Feb 2016,
#  Tue, 01 Mar 2016..Thu, 31 Mar 2016,
#  Fri, 01 Apr 2016..Sat, 30 Apr 2016,
#  Sun, 01 May 2016..Tue, 31 May 2016,
#  Wed, 01 Jun 2016..Thu, 30 Jun 2016]

Or, map dates to strings, if you need string representation:

(from..to).group_by(&:month).map do |group|
  "#{group.last.first.beginning_of_month} - #{group.last.last.end_of_month}"
end

# => ["2016-01-01 - 2016-01-31",
#     "2016-02-01 - 2016-02-29",
#     "2016-03-01 - 2016-03-31",
#     "2016-04-01 - 2016-04-30",
#     "2016-05-01 - 2016-05-31",
#     "2016-06-01 - 2016-06-30"]

To get precisely what you want, you can format the string representation of the dates:

(from..to).group_by(&:month).map do |group|
  "#{group.last.first.beginning_of_month.strftime('%d-%m-%Y')} - #{group.last.last.end_of_month.strftime('%d-%m-%Y')}"
end
#=> ["01-01-2016 - 31-01-2016",
#    "01-02-2016 - 29-02-2016",
#    "01-03-2016 - 31-03-2016",
#    "01-04-2016 - 30-04-2016",
#    "01-05-2016 - 31-05-2016",
#    "01-06-2016 - 30-06-2016"]

EDIT

To make sure ranges do not mix up because of different years, include it in grouping along with month:

from = Date.parse('15.01.2016')
to   = Date.parse('10.02.2017')
(from..to).group_by {|a| [a.year, a.month]}.map do |group|
  "#{group.last.first.beginning_of_month.strftime('%d-%m-%Y')} - #{group.last.last.end_of_month.strftime('%d-%m-%Y')}"
end
# => ["01-01-2016 - 31-01-2016",
#  "01-02-2016 - 29-02-2016",
#  "01-03-2016 - 31-03-2016",
#  "01-04-2016 - 30-04-2016",
#  "01-05-2016 - 31-05-2016",
#  "01-06-2016 - 30-06-2016",
#  "01-07-2016 - 31-07-2016",
#  "01-08-2016 - 31-08-2016",
#  "01-09-2016 - 30-09-2016",
#  "01-10-2016 - 31-10-2016",
#  "01-11-2016 - 30-11-2016",
#  "01-12-2016 - 31-12-2016",
#  "01-01-2017 - 31-01-2017",
#  "01-02-2017 - 28-02-2017"]

Answer 2

This approach avoids the need to examine every day in the date range and returns an array of strings in the desired format.

Date#>> shifts the date forward by the number of months given by the argument. Here I've shifted dates by one month (e.g., Date.parse("14-01-2016") >> 1 #=> #<Date: 2016-02-14...>). If the last day of the following month is greater than the last day of the given month, the last day of the following month is returned (e.g., Date.parse("30-01-2016") >> 1 #=> #<Date: 2016-02-29...>).

Code

require 'datetime'

def convert(sdate, edate)
  dfirst  = Date.strptime(sdate, "%d.%m.%Y")
  dlast   = Date.strptime(edate, "%d.%m.%Y")
  nmonths = 12*dlast.year + dlast.month - 12*dfirst.year - dfirst.month + 1
  dlast   = dfirst.end_of_month
  nmonths.times.map { |i|       
    "%s - %s" % [(dfirst >> i).strftime("01.%m.%Y"), (dlast >> i).strftime("%d.%m.%Y")]}
end

Example

convert "15.01.2016", "10.06.2016"
  #=> ["01.01.2016 - 31.01.2016", "01.02.2016 - 29.02.2016",
  #    "01.03.2016 - 31.03.2016", "01.04.2016 - 30.04.2016",
  #    "01.05.2016 - 31.05.2016", "01.06.2016 - 30.06.2016"]

Explanation

The steps are as follows.

sdate = "15.01.2016"
edate = "10.06.2016"

dfirst  = Date.strptime(sdate, "%d.%m.%Y")
  #=> #<Date: 2016-01-15 ((2457403j,0s,0n),+0s,2299161j)> 
dlast   = Date.strptime(edate, "%d.%m.%Y")
  #=> #<Date: 2016-06-10 ((2457550j,0s,0n),+0s,2299161j)> 
nmonths = 12*dlast.year + dlast.month - 12*dfirst.year - dfirst.month + 1
  #=>       12*24192    +     6       -   12*24192     -      1       + 1
  #=> 6 

Change dlast to last day of first month.

dlast   = d.first.end_of_month
  #=> #<Date: 2016-01-31 ((2457419j,0s,0n),+0s,2299161j)>

Map each month to the desired format.

nmonths.times.map { |i|       
  "%s - %s" % [(dfirst >> i).strftime("01.%m.%Y"), (dlast >> i).strftime("%d.%m.%Y")]}
  #=> ["01.01.2016 - 31.01.2016", "01.02.2016 - 29.02.2016",
  #    "01.03.2016 - 31.03.2016", "01.04.2016 - 30.04.2016",
  #    "01.05.2016 - 31.05.2016", "01.06.2016 - 30.06.2016"]

Answer 3

As mentioned in this thread: Is there a standard for inclusive/exclusive ends of time intervals? We should prefer inclusive start dates and exclusive end dates.

We can split a DateRange into days, weeks, months or years with the following tail recursive method:

  class DateRange
    def initialize(since_date, until_date)
      @since_date = since_date
      @until_date = until_date
    end

    def split_to(timespan)
      return [] if @since_date == @until_date
      end_date = [@since_date.send("next_#{timespan.to_s.singularize}"), @until_date].min
      [DateRange.new(@since_date, end_date)] + DateRange.new(end_date, @until_date).split_to(timespan)
    end
  end

This method takes :days, :weeks, :months or :years as a parameter. Notice that we rely on activesupport.