gh-129205: Modernize test_eintr

[vstinner]

• Use f-string.
• Fix grammar: replace 'datas' with 'data' (and replace 'data' with 'item').
• Remove unused variables: 'pid' and 'old_mask'.

• Issue: Add os.readinto API for reading data into a caller provided buffer #129205

[vstinner]

cc @picnixz @cmaloney

[cmaloney]

Definitely more modern. I duplicated a lot of code between test_read and test_readinto in GH-129211 I'm planning to dedupe (cmaloney@ad161bc#diff-60657ac311c452c61e0a872c2d0c7b6e763196617d65cfb3a7f73d791c922271L126). Happy to do that separately or could be incorporated here.

Might be out of scope, but the:

proc = self.subprocess(*args, **kwargs)
with kill_on_error(proc)
    ...
    self.assertEqual(proc.wait(), 0)

across cases I think could be deduplicated with a context manager:

with self.checked_subprocess(code, *args, **kwargs):
    ...

[cmaloney]

nit: would it make sense to centralize the "make a string which sets sleep_time" (f'sleep_time = {self.sleep_time!r}') rather than the number of copies? (Possibly also data as well?)

Definitely nice having all the code for a test case in one place though, but feels fairly templated/repeated in structure at the moment.

[vstinner]

I'm not sure that it's worth it to have a helper just to format f'sleep_time = {self.sleep_time!r}' line.

[vstinner]

I duplicated a lot of code between test_read and test_readinto in #129211 I'm planning to dedupe (cmaloney@ad161bc#diff-60657ac311c452c61e0a872c2d0c7b6e763196617d65cfb3a7f73d791c922271L126).

I integrated your change.