Add Python 3.10 and 3.11 support

[jerematt]

This also updates the manylinux base for all python3 versions to a newer (supported) release.

Tested in Python 3.11:

root@ed0fd0434619:/pypredict# python3
Python 3.11.4 (main, Jul 28 2023, 05:02:22) [GCC 12.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import predict
>>> tle='''0 LEMUR 1
... 1 40044U 14033AL  23221.74491137  .00002349  00000-0  33527-3 0  9990
... 2 40044  97.6724  33.3927 0055349 242.1156 117.4442 14.76595710491566
... '''
>>> qth=(0,0,0)
>>> import time
>>> predict.quick_predict(tle=tle, qth=qth, ts=time.time())[0]
{'norad_id': 40044, 'name': '0 LEMUR 1', 'epoch': 1691726231.9165578, 'latitude': -17.13144263315735, 'longitude': 18.44200581459458, 'azimuth': 134.08233213751, 'elevation': 0.024514276043866044, 'orbital_velocity': 27081.368214171, 'footprint': 5556.775550153253, 'altitude': 656.9883901810581, 'slant_range': 2961.4649877898605, 'eclipse_depth': -8.078600804508218, 'orbital_phase': 74.52818693865964, 'orbital_model': 'SGP4', 'visibility': b'V', 'sunlit': 1, 'orbit': 49177, 'geostationary': 0, 'has_aos': 1, 'decayed': 0, 'doppler': 1832.236825634777, 'eci_x': 5362.917004655394, 'eci_y': 4057.277134434491, 'eci_z': -2060.265299041267, 'eci_vx': 2.3998370520358243, 'eci_vy': 0.4991610080967417, 'eci_vz': 7.112043693659583, 'eci_sun_x': -113149993.411927, 'eci_sun_y': 92613872.07420123, 'eci_sun_z': 40151430.18263137, 'eci_obs_x': 6042.611378274802, 'eci_obs_y': 2041.4405016832857, 'eci_obs_z': 0.0, 'beta_angle': -72.59264563087656}

Tested in python 3.7 (due to new manylinux):

root@7d623a9bd15b:/pypredict/wheelhouse# python3
Python 3.7.0 (default, Oct 16 2018, 07:10:55) 
[GCC 6.3.0 20170516] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import predict
>>> tle='''0 LEMUR 1
... 1 40044U 14033AL  23221.74491137  .00002349  00000-0  33527-3 0  9990
... 2 40044  97.6724  33.3927 0055349 242.1156 117.4442 14.76595710491566'''
>>> qth=(0,0,0)
>>> import time
>>> ts=time.time()
>>> predict.quick_predict(tle=tle, qth=qth, ts=ts)[0]
{'norad_id': 40044, 'name': '0 LEMUR 1', 'epoch': 1691726232.0131924, 'latitude': -17.125548721242176, 'longitude': 18.440737662512376, 'azimuth': 134.07376763522498, 'elevation': 0.029358561973790488, 'orbital_velocity': 27081.3574243417, 'footprint': 5556.783744853028, 'altitude': 656.990493629095, 'slant_range': 2960.9341801699593, 'eclipse_depth': -8.078994064946237, 'orbital_phase': 74.53241505128307, 'orbital_model': 'SGP4', 'visibility': b'V', 'sunlit': 1, 'orbit': 49177, 'geostationary': 0, 'has_aos': 1, 'decayed': 0, 'doppler': 1832.0535648101843, 'eci_x': 5363.148895534224, 'eci_y': 4057.3253515128736, 'eci_z': -2059.577980621891, 'eci_vx': 2.3992428089761497, 'eci_vy': 0.49871144502385606, 'eci_vz': 7.112272550593836, 'eci_sun_x': -113149995.28166385, 'eci_sun_y': 92613870.11336055, 'eci_sun_z': 40151429.332538374, 'eci_obs_x': 6042.596991889211, 'eci_obs_y': 2041.4830845197966, 'eci_obs_z': 0.0, 'beta_angle': -72.59264472677555}

[rachaelacollins]

Hey @jerematt, do you think anything else is needed here to support 3.11? Does it just need to be tested?

[jerematt]

Nope, should be good to go, tested functionality on 3.11 and seems good, then also tested 3.7.0 to be sure that the manylinux base image change didnt negatively impact compatibility with that version. 2.7 hasnt changed as its still based off the previous manylinux image.

If you or Matt could approve, I'd be happy to merge and deploy a new release to Pypi.