support for millisecond-precise time inputs for jday

[davidcalhoun]

Based on #32, with some extra testing goodies this time. :)

[coveralls]

Coverage increased (+0.06%) to 83.869% when pulling afc0d75 on davidcalhoun:master into c6bc96c on shashwatak:develop.

[ezze]

@davidcalhoun, thanks a lot for your new pull request!

Are you sure that your new tests will pass when now.getUTCMilliseconds() returns 0?

I also want to ask you to look at new invjday function I ported from original Python library and check whether it's possible to add milliseconds support there.

[ezze]

Probably it's better to use predefined dates for testing instead of new Date().

[davidcalhoun]

Thanks for having a look. Will try to address your comments later this evening.

[ezze]

@davidcalhoun, have you any chance to look at this soon? Let me know if you still want to finish this task by yourself, or I'll try to continue it with my own efforts.

[davidcalhoun]

Ack, sorry I didn't make time for this :(. If you want to have a shot at it and have more time than me, go for it. :) . I won't mind if you close this.

[ezze]

@davidcalhoun, thanks a lot for your contribution again. I merged this with minor test fixes.