Fix slider sub/add-page rendering

[brisvag]

• closes QSS theming inconsistencies with sliders #338
• closes Styling QRangeSlider #292

This was a surprisingly small fix, that I could have never found out by myself, but codex found super quickly :P

Both problems from the issue are solved!

Before:

After:

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 84.92%. Comparing base (a32a3a4) to head (708c7a3).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #345      +/-   ##
==========================================
- Coverage   84.96%   84.92%   -0.05%     
==========================================
  Files          49       49              
  Lines        3912     3914       +2     
==========================================
  Hits         3324     3324              
- Misses        588      590       +2     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[brisvag]

Actually, I somehow didnt notice that extra handle... 🤔

[brisvag]

Ok I think it works now:

[Czaki]

Looking on implementation, it looks correct.
What code did you use for testing?

[tlambert03]

thanks @brisvag

i'm happy with this too. I had claude make a big contact sheet of all kinds of before/after cases, and in almost every case it looks equal/better:

the only vaguely interesting case is the plain, unstyled QDoubleSlider, on macos:

it looks like the add-page (unfilled area to the right of the slider) is not getting the native style anymore. It it subtle, but it is noticeable:

so, i think we're probably paying a small cost here (losing nativeness) for the case where people are using stylesheets.

claude suggestion for a "cleaner fix"

Upstream QSlider::paintEvent (qtbase qslider.cpp ≈ lines 245–256) draws the slider in one pass with SC_SliderGroove | SC_SliderHandle (+ ticks). The two-pass split in _generic_slider.py:299-330 is the actual deviation. Collapsing to one combined call would:

• still fix the QSS ::sub-page/::add-page case (same code path as the PR),
• avoid the shared-NSSlider state pollution and likely restore the native blue accent,
• match what Qt itself does.

The reason superqt splits into two passes today is the interleaved manual tick-drawing for the Monterey QSS hack (paintEvent lines 309–328). That manual tick pass can run after a single combined drawComplexControl — it draws on top either way.

but i haven't checked to see A) how much work that is and B) whether it retains the native look, while still getting the fixes for the styled case

[brisvag]

What code did you use for testing?

Just my eyes 🫠 copilot threw some tests together but as usual they are kinda useless and hard to parse, so I just tried myself. Luckily @tlambert03 came in with the big comparison :P

The reason superqt splits into two passes today is the interleaved manual tick-drawing for the Monterey QSS hack (paintEvent lines 309–328). That manual tick pass can run after a single combined drawComplexControl — it draws on top either way.

With my limited understanding, it seems a bit tricky, because RangeSlider inherits from the generic slider code and AFAICT it's necessary to have the split in there? I played around with that part of thhe code but I'm lost :P