- How to build a package is at http://python-packaging.readthedocs.org/en/latest/index.html
- Codacy
- Code climate
- Lots of detail about calendar conversion is at http://norbyhus.dk/calendar.php
- A mistake in Wolfram Alpha (and a good test case) http://mathematica.stackexchange.com/questions/97818/why-does-mathematicas-date-for-the-battle-agincourt-differ-from-the-us-navys
- This page has lots of information about conversion algorithms: https://en.wikipedia.org/wiki/Julian_day
- Conversion sample code: http://aa.usno.navy.mil/faq/docs/JD_Formula.php
- Swedish historical calendar
- Alaskan historical calendar
- French revolutionary calendar
- Julian day numbers
- Roman historical calendar - not Julian until 4BC, due to leap year corrections.
- Revised Julian Calendar - https://en.wikipedia.org/wiki/Revised_Julian_calendar
- Modified Julian Date - http://tycho.usno.navy.mil/mjd.html
- A special case of calendars - all dates are the same, but the years have different numbers.
- Eg. Juche calendar https://en.wikipedia.org/wiki/North_Korean_calendar
- Given a calendar, we should be able map a given year number to the start and end days of the year.
- Eg. the British tax year, the Jewish year.
- Easter
- Thanksgiving
- Rosh Hashanah
- Hannukah
- Leap days
- Soviet five-day work week
- Phases of the moon
For example, the set of public holidays in the UK is a day collection. It isn't a single recurrent event, but we can check if any day is in the collection, or count the number of such days between any two dates.
A function from the set of pairs of days, to a set of measurements. For example, age, measured in weeks for young babies, months for infants, years and quarter years for children, years for older people.
- Days which are in a different year in Julian and Gregorian.
- Epochs: Microsoft, Java, Unix, Julian.
- End days, Millenium, Mayan, Unix 32-bit rollover.
- Dates which are in different years depending on when the year starts.
- Dates which are leap days of the Julian calendar.
- Always do
import ..helpersetc. instead ofimport calexicon.helpers, except in tests where we can equally expect to run the tests from a the root folder and from a virtualenv which contains an installed version of the package.
We have used numbers from navy.mil to set up some of the conversion. However it looks like they have a mistake. This page is supposed to be Gregorian dates to Julian day numbers, but the results are: http://aa.usno.navy.mil/cgi-bin/aa_jdconv.pl?form=1&year=100&month=3&day=1&era=1&hr=0&min=0&sec=0.0 The Julian date for CE 100 March 1 00:00:00.0 UT is JD 1757642.500000
http://aa.usno.navy.mil/cgi-bin/aa_jdconv.pl?form=1&year=100&month=2&day=28&era=1&hr=0&min=0&sec=0.0 The Julian date for CE 100 February 28 00:00:00.0 UT is JD 1757640.500000
This suggests that there is a day in between these two. But by the Gregorian calendar they were adjacent - there was no leap year in 100.