Fixes issue with addOnce on ticker (#3938)

Fix destroyed textures throwing errors when destroyed again. (#3931)

Measure Text API, expanded (#3921) ...

* Added TextStyle.measure function (Applies to https://github.com/pixijs/pixi.js/issues/3903).

* Moved measure function from TextStyle to Text and made static.

* Updated JSDocs.

* Updated JSDocs again.

* Updates based on suggestions by @bigtimebuddy in PR.

* Created PIXI.MeasuredText to capture all the information for measuring text.

* Fixes for ESLint

* Fixes for ESLint

* Added PIXI prefix to returning MeasuredText.

* Added import for MeasuredText in Text.js.

* Refactored MeasuredText to TextMetrics and moved `measure` function into it.

* Forgot to move Text.wordWrap to TextMetrics.wordWrap

* Further cleanup of TextMetrics.

* Fixes suggested by @bigtimebuddy

* Removed clone and equals from TextMetrics as they are unnecessary.

* Destructured references to TextMetrics instance in Text and privatized TextMetrics.wordWrap.

* Deprecates calculateFontProperties & getFontStyle

* Moved getFontStyle to TextMetrics

* Made toFontString method on TextStyle

* Cache calls made to context.measureText for individual characters

* Use a smaller (10x10) canvas for TextMetrics

Improvements to how and what is stored in the texture caches. (#3834) ...

* Improvements to how and what is stored in the texture caches.
Built upon the good work from https://github.com/pixijs/pixi.js/pull/3789
+ For Text, the base textures were in the BaseTextureCache, but textures weren't in the TextureCache. This has been fixed
+ Now able to tell the BaseTexture the origin of where the canvas being passed through has come from. ie. was it from the text class? was it from the graphic class? was it from caching as a bitmap? More information now to track where these canvases (canvii??) have been generated from
+ BaseTexture has addBaseTextureToCache and removeBaseTextureFromCache functions, to match regular Texture. Any adding or removing to the BaseTextureCache is now through these functions

! Breaking change. When you call Texture.fromTextureFromCache - it no longer automatically removes it's base texture from the base texture cache. Not sure how people feel about this

? I don't like lines 233 to 236 in Texture.js, and line 591 in BaseTexture.js - I don't see why they are there or why they are needed, but I chickened out on removing them!

* Renaming function to more generic names

* Adding deprecation layer

* Fixed trying to add texture to BaseTexture cache causing unit test fail

* Fixed reading from the wrong cache

* Able to remove textures and base textures from their caches via object instances
Added debug warning if trying to add a texture to a cache id that already exists
Added unit tests

* Default string for texture origin

* Different object creation style

* textureCacheId is now an array of potential entries

* textureCacheId changed to textureCacheIds

* Removed unneeded string creation

Hot-fix for Ticker's destroy (#3904) ...

* Fixes an issue when destroying ticker on callback

* Handles more cases when destroying

Adds Ticker priority argument (#3859) ...

* Adds ticker priority

* Removes EventEmitter dependency from ticker

* Small optimization to avoid Function.call

* Changes ticker priority with UPDATE_PRIORITY const

* Fixes linting

* Use a linked list approach to ticker listeners

* Oops fixes test

* Minor fixes

* Removes unnecessary destroy check

* Minor optimizations

* Adds memory clean-up to Ticker

* Updated documentation

* Rename local `last` to `previous`

Added public hit test method. (#3878)

Merge branch 'master' into dev

Further fix issue described in #3851 (#3877) ...

Also undid some changes from #3857 because they were too complicated.

Merge branch 'master' into dev

Fix some interaction issues (#3857) ...

* Fix #3844

* Fix #3790

* Expand assertions in nested overlap tests.

Prepare plugin support for objects with multiple BaseTextures (#3830) ...

* Add ability for Prepare plugin to upload the all textures for AnimatedSprites

* Add a new function to loop through all the textures of an AnimatedSprite and add them to the queue
* Moved Duplicate `findBaseTextures` function from WebGL & Canvas Prepare to Base prepare as static
* Moved all the `findObjectType` methods to their appropiate class as static methods for consistency

* Updated Comments to reflect "static" not "private" methods

* Added the findTexturesForAnimatedSprites logic to the standard findBasicTextures

This cleans up the need to worry about ordering the registration of hooks and only calls upload once

* Split the prepare.register function into 2 registerFindHook & registerUploadHook

* Some simple cleanup before the PR

* comment formatting

* removed dependenvy on `extras` from prepare

* Fix small comment issues with pull request

Removed the [optional] reference in the comments for both
new register methods.
Added a small example to the docs

* Add deprecation method

* Added Check before calling hooks and reformatted JSDoc

Lint wouldn't pass without the params in the JSDoc

* added `instanceof` check and fixed up the return true only if a texture is found

Small Texture Cache enhancements (#3789) ...

* Textures track cache id and remove from cache in destroy.

* Added tests related to texture cache contents.

Merge branch 'master' into dev

Fix preparation of destroyed items (#3802) ...

* Added isDestroyed to DisplayObject, BaseTexture. Don't prepare destroyed items.

* isDestroyed -> _destroyed, marked as private

spritesheet url tests for #3715 (#3781)

Merge branch 'master' into dev

Fix another issue that was causing #3790 (#3800)

Merge branch 'master' into dev ...

# Conflicts:
#	src/core/textures/Texture.js

Fix #3790 and add unit test (#3796)

Fix #3794 (#3795) ...

Also add unit test for it and improve tests.

resolves #3783 (#3786)

Added core tests for Matrix, Ellipse, Circle and TilingSprite (#3699) ...

* Added matrix tests for prepend() and identity()

* Added contains if branch tests for Circle and Ellipse

* Removed trailing whitespace in circle.js

* Added TilingSprite tests for contains, width and height

* Removed .only tag in TilingSprite test

* Changed test values for prepand and identity tests, as per change request

Removes pending tests from Container (#3769)

[v4.4] Revert "Fix #3609 , about `filterClamp` & `filterArea` (#3626)" (#3760) ...

* Revert "Fix #3609 , about `filterClamp` & `filterArea`  (#3626)"

The real fix is to exclude "filterClamp" from uniformData
That fixes outlineFilter and glowFilter

* added a test, so finscn wont break it again ;)

[v4.4] fix SpriteRenderer for multiple webgl renderers, fix tests, add "withGL" (#3758) ...

* fix SpriteRenderer for multiple webgl renderers, fix tests, add "withGL"

* isWebGLSupported renamed to withGL

Uncomment & fix failing test from #3535. (#3638) ...

Resolves #3419 and #3628.

I just saw #3628 after I started working on this and I opened another PR (wrong) about the same bug #3628 looks good to me but adds an extra property, which I personally tend to dislike. This might have some side effects for cases that are untested, but I thought it's worth opening as an alternative to #3628.

centralise isWebGLSupported (#3740)

Fix cursors not working as expected in nested interactive objects (#3704) ...

* Changed pointerout to pointerleave.

* Add comment about pointerleave event.

* Fixed cursor changes being mishandled.

Cursor changes should not be attempted if it just just the default
value.

* Fix #3689

* Make change from #3703

* Changed pointerout to pointerleave in tests.

addChild/removeChild transform IDs were wrong (#3722) ...

* addChild/removeChild transform IDs were wrong

* better test

* uncomment expects

* also add removeChildren

* lint on tests