Workaround for issue experienced with mktime on Micropython

[cefn]

This change fixes an issue I experienced when using the adafruit_gps module from micropython using a CircuitPython-compatibility wrapper. Where the time value provided is a floating point number, then the secs value is also a float after the remainder operation.

The implementation of struct_time in the compatibility layer fails when the struct_time is used to call micropython's mktime, which operates with a stricter interpretation of https://docs.python.org/3.1/library/time.html

The time value as returned by gmtime(), localtime(), and strptime(), and accepted by asctime(), mktime() and strftime(), is a sequence of 9 integers.

...hence requires the secs argument to be an integer when constructing the timestamp.

Wrapping the secs calculation in int() resolves the issue. The issue I experienced may be specific to the GPS module which returns the value in this form, and to the library downstream which tries to use it and dies, so feel free to ignore if there are reasons not to merge.

[tannewt]

Hours will have the same problem I believe. Mind converting time_utc to an int instead?

[cefn]

Ah yes, python floor division not integer division, so still a float. I'd pulled that int() out as I thought I'd been too conservative :)

[cefn]

It's quite a performance for me to get the full testing scenario for GPS set up again (reconfigure hardware, reconfigure software, go wandering outdoors and get a fix). Hopefully this new commit works but is untested for now.

[tannewt]

@jerryneedell Do you have a GPS set up and can test?

[jerryneedell]

I should be able to test it this evening.

[tannewt]

Thank you!

…

On Tue, Feb 27, 2018 at 10:17 AM jerryneedell ***@***.***> wrote: I should be able to test it this evening. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#3 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADNqYaqjVnmo7cY4cAqsUrLhm3sYYWsks5tZEazgaJpZM4ST33V> .

[jerryneedell]

Tested GPS acces with this PR on Metro_M4_Express_Revb - with CP 3.0 - seems to run as expected. No problems encountered.

[tannewt]

Thank you!