Music should automatically clear headers and footers

[Nick-Mazuk]

Resolves: #9043
Resolves: #9340
Resolves: #9159

This PR removes collisions between headers/footers and music. When header or footer text becomes too tall, the music will automatically move out of the way.

• By default, the footer offset is now 0mm (because it's unnecessary)
• The header/footer text is now rendered correctly
• There is a configurable allowed gap between the music and header/footer

Here's a diagram of what's happening:

header-footer.collision.avoidance.mov

[Jojo-Schmitz]

Some test failures, crashes actually

[Nick-Mazuk]

Fixed the failing tests.

Edit: for some reason, the tests are passing on my machine but are failing on CI. Working on figuring out why that's happening now.

[Jojo-Schmitz]

For me it crashes whan opening an existing score, and with a failed assertion:

Fatal: ASSERT: "!strcmp(MStyle::valueType(idx),"Ms::Spatium")" in file .../libmscore/score.h, line 896

So it got to be in layoutpage.cpp, where you're using styleP() (Edit: were using)

[Jojo-Schmitz]

Still fails engraving_tests.
BTW: consistent with the backport to 3.x, which fails the tst_layout_elements, both without any good hint at what the problem might be. Maybe I also did something wrong, as it works for header but not not (fully) for footer

[DmitryArefiev]

Hi @Nick-Mazuk , should it work like this?

[Nick-Mazuk]

Hi @Nick-Mazuk , should it work like this?

No, that's a bug. Thanks for finding it, I'll fix it!

[DmitryArefiev]

Hi @Nick-Mazuk , should it work like this?

No, that's a bug. Thanks for finding it, I'll fix it!

Cool! Then just assign to me when it's ready

[Nick-Mazuk]

As I understand, the page break should lock the amount of systems on one particular page. So the footer should not move system with a page break to a next page.
But I guess my case above with a such tall footer text is a very rare workflow so let's promote it now

Oh, I see what you mean now. From the video, that is the behavior I'd expect, though I can see how someone would not expect that behavior. I'll double-check with Simon when I speak with him tomorrow to confirm the desired result.

[Jojo-Schmitz]

The Merge commit doesn't belong into this (or any) PR

[oktophonie]

The behaviour with regard to page breaks seems as I would expect: a page break simply means that the next system will be on the following page, but it can't force more systems to fit on the preceding page than there is space for.

There's scope for some sort of "keep systems together on page" functionality in the future (analogous to "keep bars together on system") but that would be an entirely separate issue from this.

[DmitryArefiev]

Got it. Thanks for review Simon!

Lets merge it into master

[DmitryArefiev]

Please merge into master (@RomanPudashkin )