Make the test suite environment-independent, fix UTF-8 redirection, and a UI-call race

[tleonhardt]

Three fixes, all pre-existing on consolidated_toolbar and none specific to the
reserved-row toolbar work. Found while investigating a reviewer's report of failures in
unchanged tests.

1. The suite inherited the caller's colour environment

Rich and cmd2 both consult environment variables when deciding whether to emit styling, and
the suite inherited them. Exporting any one made large numbers of unrelated tests fail
depending on who ran it:

Variable	Tests failed before
NO_COLOR=1	15
FORCE_COLOR=1	53
TTY_COMPATIBLE=1	53

The failures look like product regressions, which makes them expensive to diagnose — this
cost a reviewer and me a full round trip. An autouse fixture now neutralizes them, and a
guard test fails if any reaches a test again. Tests that exercise these variables still set
them explicitly, because a test's own monkeypatching runs after the fixture.

2. Redirection and piping failed on non-UTF-8 systems (user-visible)

Command output is rendered by Rich and routinely contains non-ASCII, but redirection targets
and pipes were opened with the locale's encoding. On a system whose default is not UTF-8 —
a Windows console using a legacy code page — help > out.txt raised UnicodeEncodeError
and left the user an empty file plus advice to set PYTHONIOENCODING.

Both now use UTF-8 explicitly. Two tests that read redirected output back were relying on the
locale for decoding as well, so they name it too. CHANGELOG updated.

3. A race reported a timeout for a successful UI call (back-port)

Cherry-picked from the reserved-row branch, where it was found; it is not specific to that
work. CommandToolbar._call_in_ui() polls the pending future with a 0.1s timeout and
re-raises when the future is already done. That branch exists because
concurrent.futures.TimeoutError is TimeoutError on Python 3.11+, so a callback
raising a timeout of its own cannot be told apart by type from the poll expiring.

Re-raising the caught exception conflates the two: when the callback completes in the window
between the poll expiring and the future being inspected, the caller is told the call timed
out although it succeeded. It now asks the future for its outcome instead.

test_command_toolbar_ui_call_propagates_failures failed once in 40 runs on this branch
before the fix and not once in 60 after. The added regression test drives the interleaving
deterministically rather than relying on timing.

Verification

Full suite under every environment that previously broke it, three runs of the combined case:

baseline                      1937 passed, 2 skipped
NO_COLOR=1                    1937 passed, 2 skipped
FORCE_COLOR=1                 1937 passed, 2 skipped
TTY_COMPATIBLE=1              1937 passed, 2 skipped
TTY_INTERACTIVE=0             1937 passed, 2 skipped
non-UTF8 locale, no -Xutf8    1937 passed, 2 skipped
all hostile at once  x3       1937 passed, 2 skipped

make check, make test and make docs-test all pass.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.67%. Comparing base (dee4c9b) to head (7b70fbc).

Additional details and impacted files

@@                  Coverage Diff                  @@
##           consolidated_toolbar    #1752   +/-   ##
=====================================================
  Coverage                 99.67%   99.67%           
=====================================================
  Files                        25       25           
  Lines                      6464     6464           
=====================================================
  Hits                       6443     6443           
  Misses                       21       21           

Flag	Coverage Δ
unittests	99.67% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.