* Clean up docs in CharArray.
Dangling comment here, typo there. The only substantive changes here are avoiding a duplicate check for equal size in equals() and avoiding an explicit call to Long.toString() that I don't think is required (it's already being appended to a String).
* Dubious? Avoid heap pollution in appendAll().
This requires making `appendAll(T...)` final, because the JVM can apparently get "heap pollution" when generic-typed varargs are used by a method that isn't at least one of private, static, and/or final. I haven't found any documentation on what "heap pollution" entails, so this might not be needed. I chose not to apply the SafeVarargs annotation because it isn't used anywhere else in libGDX, even though it is a valid compiler warning to omit it. The check for non-zero length before wasn't really... helpful, I think, since the loop just wouldn't run if array was empty. But, the parameter could be null in some situations, and iterating over that would crash with an NPE, so now it checks that the parameter is not null.
* The method is writer(), not asWriter().
* CharArray shouldn't behave differently from...
... every other Array type here; none of the others throw Exceptions when the Array is empty.
* Added initial prototype for a Universal Tiled Map Loader.
This will seamlessly handle any map type passed into it. Passing it along to the appropriate loader based on file type and properties.
The way properties will be handled not decided yet. There are 2 potential methods. But it's ugly.
Might refactor and consolidate atlas param instead.
Added
* Finalized version of the new universal tiled map loader.
Cleaned up a mismatch in the TmjMapLoader, so it uses its own parameters the same way as all the other maploaders do.
* Added a test which loads an assorted pick of maps from other tests. All using the TiledMapLoader.
* Apply formatter
* Refactored the tiledmap parameters. Removed the parameter subclasses from the loaders. All Parameters should live in BaseTiledMapLoader.Parameters (which they already were anyway).
Changed any test which was using the AtlasTmxMapLoader.AtlasTiledMapLoaderParameters to use
BaseTiledMapLoader.Parameters instead.
* Apply formatter
---------
Co-authored-by: GitHub Action <action@github.com>
Co-authored-by: Tomski <tomwojciechowski@asidik.com>
* Updated a the getTileSetDependencyFileHandle method inside the TmxMapLoader to reflect PR #7136's code duplication fix.
Did the same for the TmjMapLoader as well.
* Apply formatter
---------
Co-authored-by: GitHub Action <action@github.com>
Originally JsonValue.get() ignored case. So it worked accidentally. But recently it was changed. Checking against "tileSets" was a mistake to begin with so it needs to be fixed. The TMX map loader IS case-sensitive, so it should match. Which is why I am not suggesting to use the getIgnoreCase() method that was added.
* For issue #7024
You can now create HexagonalTiledMapRenderer objects when passing in new empty TiledMap objects. This will no longer result in a crash and allows similar functionality to all the other TiledMapRenderers.
Also added getters and setters for the necessary fields which would be needed for a user to set it up manually.
* Apply formatter
---------
Co-authored-by: GitHub Action <action@github.com>
* #7651: add default value loading for tiles with a class
* #7651: add test for tile with a nested class
* #7651: add test for object with a class
* make renderMapLayer public
* Apply formatter
* add a proper warning message instead of throwing a NullpointerException when projectFilePath is not passed for a map with class properties/classes
* Apply formatter
---------
Co-authored-by: GitHub Action <action@github.com>
Added 3 new uncompressed base64 tiledmaps which we can use in GWT. As those maps could not be loaded in GWT so the original test was useless.
Created a specific test FOR Testing the AtlasTmxMapLoader stuff in GWT to replicate the normal non-gwt test.
* Added TemplateObject loading feature into BaseTmxMapLoader class.
* Added a new test for Template Objects. Test is an exact replica of the TiledMapObjectLoadingTest except using template objects. We should compare results to that test, should look exactly the same.
Since template objects don't support embedded tilesets, I created a new tileset to use based on the old one for this new map.
* Updates to Tmj template loading.
* Added new map files and modified the Test for Template map objects. This test can swap between both TMJ and TMX map types for comparison to show proper parsing of object properties between formats.
* Modified Test to better show which loader you were currently using.
Added missing import for gwt test.
Small fix in tmj loader
* Updated TestMaps, fixed typo in property name.
* Rebased and Updated the Tiled Template functionality to support PointMapObjects and TextMapObjects.
Refactored some duplicate functionality after coming across some unique issue's with TextMapObjects.
Updated Tests to include the 2 new map objects and templates.
* Apply formatter
* Rebased against latest Master.
Updated some comments because of broken formatting.
Refactored a section of code. And updated a text string to look for proper lowercase "tilesets" because get is no longer case-insensitive.
* Apply formatter
---------
Co-authored-by: GitHub Action <action@github.com>
* parseRoot -> parseValue, makes it useful in more situations.
* Fixed multiple patterns doing capture all at the same time.
* Fixed capture all not setting the root value name.
* Fixed patterns matching a single value not using the first when other patterns prevent parsing from stopping.
* JsonWriter/String improvements.
* Avoid boxing.
* Use stack only for depths < current.
* Avoid comma check after name for set().
* JsonMatcher, added parseRoot(), start(), patterns constructor, no patterns captures whole document, javadoc.
parseRoot is ugly, but it's otherwise inconvenient to obtain the root JsonValue.
* Multisample FBO to support OpenGL ES 3.0
* Exposed members of GLFrameBuffer's internal classes
* Moved unit tests from GL31 to GL30, updated CHANGES
* Added frame buffer tests to GWT
* Adjust test, make sure shader is gles 3.0 compliant, webgl complained.
Webgl also doesn't like non sequential glDrawBuffer arguments.
Changed the MRT test slightly so its a bit easier to debug whats going on.
* Apply formatter
* Fix the masking of transfer.
Clear re-used fbo in MRT to ensures transfer is corrrect
* restore transfer
* Apply formatter
---------
Co-authored-by: Tom Wojciechowski <tomwojciechowski@asidik.com>
Co-authored-by: GitHub Action <action@github.com>
* #7651: add default value loading for tiles with a class
* #7651: add test for tile with a nested class
* #7651: add test for object with a class
* make renderMapLayer public
* Apply formatter
---------
Co-authored-by: GitHub Action <action@github.com>
