PERF: Avoid fragmentation of DataFrame in read_sas

[phofl]

• [x] closes #48595 (Replace xxxx with the Github issue number)
• [ ] Tests added and passed if fixing a bug or adding a new feature
• [x] All code checks passed.
• [x] Added type annotations to new arguments/methods/functions.
• [x] Added an entry in the latest doc/source/whatsnew/vX.X.X.rst file if fixing a bug or adding a new feature.

Not sure how to test this, the op file is 13mb which is way to large for our repository.

[ParfaitG]

I know we want to limit network calls, for testing, how about reading OP's URL that points to a CDC .xpt file. Currently, read_sas can parse online files:

pd.read_sas("https://wwwn.cdc.gov/Nchs/Nhanes/2017-2018/DR1TOT_J.XPT")

There is also a pytest decorator check in case link changes to use before test. Given the date and government site, link looks reliable over the years:

@pytest.mark.network
@pytest.mark.slow
@tm.network(
    url="https://wwwn.cdc.gov/Nchs/Nhanes/2017-2018/DR1TOT_J.XPT",
    check_before_test=True,
)

[phofl]

This runs forever (25secs), since we don't have slow builds anymore, I don't think this is worth it

[mroeschke]

Thanks @phofl