In convert_date(), round the seconds values for more accuracy

[keitherskine]

First off, many thanks for creating this package @willtrnr . It's greatly appreciated.

This PR should address an issue I found when converting datetime values from the floats in an .xlsb spreadsheet. The date in question was July 4 2020 at 10:00:00. In Excel, this is stored as the floating point number 44016.416666666664. However, when convert_date() is used to retrieve the datetime, something unexpected happened:

>>> pyxlsb.convert_date(44016.416666666664)
datetime.datetime(2020, 7, 4, 9, 59, 59)  # the time should be 10:00:00, not 09:59:59

In convert_seconds() the seconds values are cast to integers with int. This truncates the seconds, whereas round will I believe return a better, more correct value:

>>> (44016.416666666664 % 1) * 24 * 60 * 60
35999.99999979045
>>> int((44016.416666666664 % 1) * 24 * 60 * 60)  # existing code
35999
>>> round((44016.416666666664 % 1) * 24 * 60 * 60)  # new code
36000

Hence, this PR changes the int to round so the nearest seconds value is returned, not the lower one.

[sourcery-ai[bot]]

Sourcery Code Quality Report

❌  Merging this PR will decrease code quality in the affected files by 0.16%.

Quality metrics	Before	After	Change
Complexity	3.18 ⭐	3.18 ⭐	0.00
Method Length	48.00 ⭐	48.67 ⭐	0.67 👎
Working memory	6.91 🙂	6.91 🙂	0.00
Quality	76.23% ⭐	76.07% ⭐	-0.16% 👎

Other metrics	Before	After	Change
Lines	23	23	0

Changed files	Quality Before	Quality After	Quality Change
pyxlsb/init.py	76.23% ⭐	76.07% ⭐	-0.16% 👎

Here are some functions in these files that still need a tune-up:

File	Function	Complexity	Length	Working Memory	Quality	Recommendation

Legend and Explanation

The emojis denote the absolute quality of the code:

• ⭐ excellent
• 🙂 good
• 😞 poor
• ⛔ very poor

The 👍 and 👎 indicate whether the quality has improved or gotten worse with this pull request.

---

Please see our documentation here for details on how these metrics are calculated.

We are actively working on this report - lots more documentation and extra metrics to come!

Help us improve this quality report!

[willtrnr]

Good catch, round is indeed much better for the job. Thanks for the fix!