Use requests.Session for all .get downloads

[mdavis-xyz]

This change is to make all requests to nemweb use the same requests.Session. The benefit of this is that the underlying TCP connection is re-used, instead of being re-established each time. (And also the TLS connection?) The result is that it's far faster to download lots of small files. (Another benefit is that things like user agent only need to be set up once.)

I tried downloading DUDETAIL with the script below. It took 93s using a session, and 285s without a session.

(I also changed the few http:// urls to https://. I think this also helps with re-using the TCP+TLS connections. Also https is available, so we should use it.)

from nemosis import dynamic_data_compiler, defaults, cache_compiler
from tempfile import TemporaryDirectory
from time import time

with TemporaryDirectory() as tmp_dir:
    start_t = time()

    cache_compiler(
        start_time='2020/01/01 00:00:00', 
        end_time='2024/11/01 00:05:00', 
        table_name='DUDETAIL', 
        raw_data_location=tmp_dir
    )
    end_t = time()
    print(f"Took {end_t - start_t}s")

[mdavis-xyz]

I've also added a r.raise_for_status() for each request, to partially resolve #60.