GH-<id>: [Python] Assert assume_timezone nonexistent/ambiguous results in test_compute

[anxkhn]

### Rationale for this change

In `python/pyarrow/tests/test_compute.py`, `test_assume_timezone` builds an
`expected` array and compares it against the kernel `result` with `.equals()`, but
four of those comparisons drop the `assert`, so the boolean result is discarded and
the comparison verifies nothing. Because of this, the `assume_timezone` output
branches for DST-nonexistent times (`nonexistent="shift_forward"` /
`"shift_backward"`) and DST-ambiguous times (`ambiguous` earliest / latest) are
executed but never actually checked. A regression that produced wrong timestamps on
those branches would still let the test pass. The earlier assertions in the same
test already use the correct `assert result.equals(pa.array(expected))` form, which
shows the four bare comparisons were meant to be assertions too.

### What changes are included in this PR?

Prefix the four bare `.equals()` comparisons in `test_assume_timezone` with
`assert`, matching the form already used elsewhere in the same test. This is a
test-only change; no production or kernel code is touched.

### Are these changes tested?

Yes. The change is itself a test fix: with the asserts in place,
`pytest python/pyarrow/tests/test_compute.py -k assume_timezone` still passes on the
current kernel (confirming the four expected/result pairings are correct) and would
now fail if the `assume_timezone` DST-nonexistent or DST-ambiguous branch regressed.
The test is gated on `@pytest.mark.pandas` and `@pytest.mark.timezone_data`, so it
runs in the CI jobs that have pandas and timezone data.

### Are there any user-facing changes?

No.

This change was made with the help of an AI coding assistant; I reviewed the diff,
confirmed each expected/result pairing against the C++ kernel semantics and pandas'
`tz_localize` behavior, and I own the change.

Notes on the PR body:

• The last paragraph is the AI-usage disclosure that Arrow's AI-generated-code
  guidance asks for. Keep it: the policy explicitly wants authors to "be upfront
  about AI usage." It leaks nothing about the local environment.
• Do NOT tag or ping any maintainer in the PR (Arrow policy: AI-assisted PRs must
  not ping maintainers).
• Arrow uses squash-merge and the merge bot adds Signed-off-by; do not commit
  with -s.

---

The diff (for reference)

diff --git a/python/pyarrow/tests/test_compute.py b/python/pyarrow/tests/test_compute.py
@@ def test_assume_timezone():
         timezone, nonexistent="shift_forward"))
     result = pc.assume_timezone(
         nonexistent_array, options=options_nonexistent_latest)
-    expected.equals(result)
+    assert expected.equals(result)

     expected = pa.array(nonexistent.tz_localize(
         timezone, nonexistent="shift_backward"))
     result = pc.assume_timezone(
         nonexistent_array, options=options_nonexistent_earliest)
-    expected.equals(result)
+    assert expected.equals(result)
@@
     expected = ambiguous.tz_localize(timezone, ambiguous=[True, True, True])
     result = pc.assume_timezone(
         ambiguous_array, options=options_ambiguous_earliest)
-    result.equals(pa.array(expected))
+    assert result.equals(pa.array(expected))

     expected = ambiguous.tz_localize(timezone, ambiguous=[False, False, False])
     result = pc.assume_timezone(
         ambiguous_array, options=options_ambiguous_latest)
-    result.equals(pa.array(expected))
+    assert result.equals(pa.array(expected))

Local commit message (already on the branch, Conventional Commits, leak-clean):

test(python): assert assume_timezone nonexistent/ambiguous results

Four comparisons in test_assume_timezone computed a result with .equals()
but dropped the assert, so the boolean was discarded and nothing was
verified. As a result the assume_timezone kernel's DST nonexistent
(shift_forward/shift_backward) and ambiguous (earliest/latest) output
branches were exercised but never checked, letting a regression pass.

Prefix the four bare comparisons with assert, matching the
assert result.equals(pa.array(expected)) form already used earlier in
the same test.

(The commit subject uses the test(python): Conventional-Commits form for local
history; the upstream PR title must be the GH-<id>: [Python] ... form above.
Arrow's squash-merge takes the PR title + body as the final commit message, so the
local subject is not what lands on main.)

[github-actions]

Thanks for opening a pull request!

If this is not a minor PR. Could you open an issue for this pull request on GitHub? https://github.com/apache/arrow/issues/new/choose

Opening GitHub issues ahead of time contributes to the Openness of the Apache Arrow project.

Then could you also rename the pull request title in the following format?

GH-${GITHUB_ISSUE_ID}: [${COMPONENT}] ${SUMMARY}

or

MINOR: [${COMPONENT}] ${SUMMARY}

See also:

• Other pull requests
• Contribution Guidelines - Contributing Overview

[Copilot]

Pull request overview

Fixes test_assume_timezone in python/pyarrow/tests/test_compute.py so that four previously non-asserted .equals() comparisons actually validate the assume_timezone results for DST-nonexistent and DST-ambiguous inputs.

Changes:

• Add missing assert keywords for two nonexistent-time (shift_forward / shift_backward) result comparisons.
• Add missing assert keywords for two ambiguous-time (earliest / latest) result comparisons.

[kou]

Please read https://arrow.apache.org/docs/dev/developers/overview.html carefully and try again.