welcome at SebWalak.com

Operating on time ranges

categories: /software-development
tags: #go   #lib

For one of my recent projects I needed a tool to perform few basic operations on time ranges (called periods here).

I was doing a lot of time manipulation (however cool that sound I am not building a time machine) and was getting tired of polluting other code with this responsibility. I decided to encapsulate that in a separate library.

The library “github.com/outo/temporal/pops” (period operations) contains:

Contains constructs:

  • Period:

an immutable representation of time range with the lower boundary being inclusive and upper being exclusive

methods:

  • Check will identify the placement of specified time range with respect to the time range within this period
Periods relationship illustration
  • Get copy of startIncl and endIncl as multi return
  • GetStartIncl copy of startIncl
  • GetEndExcl copy of endIncl

Performance

each test performed on on an Intel® Core™ i7 4500U Processor with enough DDR3 SDRAM, consisted of 100 samples, each test based on checking the relationship between a target time range and a sliding and size-changing time range constructed to simulate each of the possible relationships took on average nearly 100 nanoseconds per PeriodCheck (with extremums at 80 and 160 nanoseconds)

  • Periods:

an immutable representation of practically (memory, language constraints) unlimited number of periods

methods:

  • Subtract will return a new Periods instance representing the time ranges within this object that are not within specified object
  • Union will return a new Periods instance representing the time ranges within this object and specified object
Periods subtract and union operation illustration

both methods’ result is:

  • defragmented - there aren’t going to be two ranges that are overlapping,
  • stitched - there aren’t going to be two ranges that are adjacent,
  • sorted - the ranges are appearing from the earliest to most recent

Note:

It is possible to create an instance of Periods with erratic time ranges by passing such to the constructing function CreatePeriods

Performance

each test performed on on an Intel® Core™ i7 4500U Processor with enough DDR3 SDRAM, consisted of 1000 samples of two Periods, each with about 30 distinct ranges (defragmented, stitched and sorted) took on average:

  • Subtract - 80 microseconds (producing nearly 30 ranges)
  • Union - 76 microseconds (producing about 20 ranges)
Thumbnail