Improve GLTF support #2069

riccardobl · 2023-09-04T13:07:37Z

This is the initial work to improve and extend the gltf support in jme.
This pr:

Fix an issue with cache poisoning when manipulating the first instance of a cached spatial
Adds the UserData loader from Simsilica and makes it the default (as expected from a jme loader)
Moves all the extensions related code to packages with their fully qualified name to make it easier to figure out which code belongs to which official or unofficial extension
Adds a new JME_speaker extension to load speaker nodes directly from blender scenes into jme (this will require an extension also on the blender side)
Makes default extension registrable statically, this allows specific modules (think about (j)bullet, minie etc) to statically register their own extension in the gltf loader.

Currently testing:

JME_physics extension automatically registered by jbullet to load rigidbodies from blender scenes (needs blender addon)

… the first instance of a spatial from poisoning the cache.

…nto packages named after their fully qualified name

Ali-RS · 2023-09-05T17:55:39Z

...lugins/src/gltf/java/com/jme3/scene/plugins/gltf/ext/JME_speaker/SpeakerExtensionLoader.java

+ final float volume = jsonObject.getAsJsonPrimitive("volume").getAsFloat();
+ final float pitch = jsonObject.getAsJsonPrimitive("pitch").getAsFloat();
+ final float attenuation = jsonObject.getAsJsonPrimitive("attenuation").getAsFloat(); // unused
+ final float distanceMax = jsonObject.getAsJsonPrimitive("distance_max").getAsFloat();
+ final float distanceReference = jsonObject.getAsJsonPrimitive("distance_reference").getAsFloat();
+ final float volume_min = jsonObject.getAsJsonPrimitive("volume_min").getAsFloat(); // unused
+ final float volume_max = jsonObject.getAsJsonPrimitive("volume_max").getAsFloat(); // unused
+ final float angleOuterCone = jsonObject.getAsJsonPrimitive("angle_outer_cone").getAsFloat();
+ final float angleInnerCone = jsonObject.getAsJsonPrimitive("angle_inner_cone").getAsFloat();
+ final float outerConeVolume = jsonObject.getAsJsonPrimitive("outer_cone_volume").getAsFloat(); // unused
+ final String soundPath = jsonObject.getAsJsonPrimitive("sound_path").getAsString();


Properties are named in camel case in the rest of the Gltf extensions so I would suggest keeping the same naming convention here as well, please. (i.e. distance_max -> distanceMax,...)

Ali-RS · 2023-09-05T17:59:42Z

jme3-plugins/src/gltf/java/com/jme3/scene/plugins/gltf/MaterialAdapter.java

@@ -63,7 +61,7 @@ public abstract class MaterialAdapter {

 protected abstract MatParam adaptMatParam(MatParam param);

- protected void init(AssetManager assetManager) {
+ public void init(AssetManager assetManager) {


Why method access modifier is changed to "public" here?

Ali-RS · 2023-09-05T18:08:53Z

jme3-plugins/src/gltf/java/com/jme3/scene/plugins/gltf/UserDataLoader.java

+/*
+ * $Id$
+ *
+ * Copyright (c) 2019, Simsilica, LLC
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in
+ * the documentation and/or other materials provided with the
+ * distribution.
+ *
+ * 3. Neither the name of the copyright holder nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */


Please move the copyright notice to the top.

Ali-RS · 2023-09-05T18:15:32Z

jme3-plugins/src/gltf/java/com/jme3/scene/plugins/gltf/UserDataLoader.java

+ public Object handleExtras(
+ GltfLoader loader,
+ String parentName,
+ JsonElement parent,
+ JsonElement extras,
+ Object input
+ ) {
+ log.fine(
+ "handleExtras(" +
+ loader +
+ ", " +
+ parentName +
+ ", " +
+ parent +
+ ", " +
+ extras +
+ ", " +
+ input +
+ ")"
+ );


Regarding the code-formatting here and some other places, I like to place them in one line until the line length exceeds the max length, only then I would move them to the new line but this is just my own preference, not sure about others though!

Edit:
FYI, this is how the original code is formatted, which sounds nicer to me.
(4 lines vs. 20 lines)

https://github.com/Simsilica/JmeConvert/blob/ad52c7729cb9a2345eaeb88cc037fb64348d8779/src/main/java/com/simsilica/jmec/gltf/GltfExtrasLoader.java#L73-L76

Ali-RS · 2023-09-05T18:24:24Z

@riccardobl thanks for your contribution

Please update the copyright date to 2023 on the modified classes.

Ali-RS · 2023-09-06T05:21:26Z

jme3-plugins/src/gltf/java/com/jme3/scene/plugins/gltf/UserDataLoader.java

+ } else if (jp.isString()) {
+ return jp.getAsString();
+ }
+ log.fine("Unhandled primitive:" + jp);


Seems this should be a "warning"?

Ali-RS · 2023-09-06T05:22:18Z

jme3-plugins/src/gltf/java/com/jme3/scene/plugins/gltf/UserDataLoader.java

+ } else if (el instanceof JsonPrimitive) {
+ return toAttribute(el.getAsJsonPrimitive(), nested);
+ }
+ log.fine("Unhandled extras element:" + el);


Seems this should be a "warning"?

stephengold · 2023-09-06T20:36:15Z

In general, a PR should do one thing. The scope of a PR should be narrow, addressing a single refactor, issue, or feature. See Creating Good Pull Requests.

According to the description, this PR includes 3 new features, a bug fix, and major refactoring.

I request that the changes be split up into multiple PRs. This will make the changes easier to understand and review.

riccardobl · 2023-09-06T21:29:58Z

In general, a PR should do one thing. The scope of a PR should be narrow, addressing a single refactor, issue, or feature. See Creating Good Pull Requests.

According to the description, this PR includes 3 new features, a bug fix, and major refactoring.

I request that the changes be split up into multiple PRs. This will make the changes easier to understand and review.

You are right, but since i am submitting these PRs in between projects i am doing with the engine, i have limited time and working on multiple PRs that rely on each other is very time consuming.

stephengold · 2023-09-06T21:31:03Z

Another thing: package names use only lowercase letters and digits (no underscores). See the Google Java Style Guide section 5.2.1.

stephengold · 2023-09-06T22:00:09Z

i am submitting these PRs in between projects i am doing with the engine, i have limited time

Are you suggesting your PRs deserve special treatment because you work on other projects and have limited time?

riccardobl · 2023-09-06T23:12:47Z

Another thing: package names use only lowercase letters and digits (no underscores). See the Google Java Style Guide section 5.2.1.

I would make an exception in this case to match their registered extension name in gltf (https://github.com/KhronosGroup/glTF/blob/main/extensions/README.md) but i don't have a strong opinion in this regard

i am submitting these PRs in between projects i am doing with the engine, i have limited time

Are you suggesting your PRs deserve special treatment because you work on other projects and have limited time?

I am suggesting that if you want the perfect PRs we can leave them open indefinitely until i have time to follow all the rules, or we can accept my bad PRs to get features and bug fixes merged quickly as long as the code works.

I am always working on a fork, so i don't have any urge to get my code merged, and besides, last time i checked i had push rights to the repo.

Ali-RS · 2023-09-07T07:59:43Z

Regarding package names, even though I do like your approach but yeah that seems to be not a standard Java package naming(?). Also as most of those extensions turn out to be a single class file anyway, I would recommend having a gltf.ext.khr package for all Khronos extensions, (but if you like you can also divide them like gltf.ext.khr.light, gltf.ext.khr.material, gltf.ext.khr.texture,....)

Next, we can have a gltf.ext.jme package for JME extensions. Similarly, if you think a JME extension is going to have multiple related classes (which sounds unlikely to me) you can make a subpackage for that. (e.g. gltf.ext.jme.audio,...)

Ali-RS · 2023-09-07T08:14:14Z

Regarding PR complexity, looks like UserDataLoader is imported from the JmeConvert library(?), if so, I am already using that in my project and it works just fine, so I think we can skip reviewing the functionality of that class.

Ali-RS · 2023-09-07T08:26:29Z

...lugins/src/gltf/java/com/jme3/scene/plugins/gltf/ext/JME_speaker/SpeakerExtensionLoader.java

+ * An extension for the GLTF loader that loads speaker nodes.
+ * (This extension requires the jme-extras addon for blender)
+ */
+public class SpeakerExtensionLoader implements ExtensionLoader {


I am fine with SpeakerExtensionLoader naming, however, I think something like AudioNodeExtensionLoader or AudioExtensionLoader would be nicer, as in JME, I believe we do not use the term "Speaker".

Ali-RS · 2023-09-07T13:03:27Z

Next, we can have a gltf.ext.jme package for JME extensions.

I think we can even omit jme from above, so we have just gltf.ext for custom JME extensions and gltf.ext.khr for Khronos extensions. What do you think?

Just my two cents!

stephengold · 2023-09-07T15:18:55Z

we can leave them open indefinitely until i have time to follow all the rules, or we can accept my bad PRs to get features and bug fixes merged quickly

The goal is to do what's best for the project in the long term. Code added to the project requires maintenance, so we want PRs to be in as good shape as possible before they're integrated, while the burden of maintenance falls on the submittor, not the community.

If this PR is urgent for some reason, please document that reason.

besides, last time i checked i had push rights to the repo.

As do many other people. Why even bring that up? This is a discussion of proposed changes to the codebase, not a discussion of Riccardo. Let's keep the discussion as impersonal as possible.

riccardobl · 2023-09-07T15:39:07Z

we can leave them open indefinitely until i have time to follow all the rules, or we can accept my bad PRs to get features and bug fixes merged quickly

The goal is to do what's best for the project in the long term. Code added to the project requires maintenance, so we want PRs to be in as good shape as possible before they're integrated, while the burden of maintenance falls on the submittor, not the community.

If this PR is urgent for some reason, please document that reason.

besides, last time i checked i had push rights to the repo.

As do many other people. Why even bring that up? This is a discussion of proposed changes to the codebase, not a discussion of Riccardo. Let's keep the discussion as impersonal as possible.

Does the size of the PR reflects on the quality of the code?

As do many other people. Why even bring that up? This is a discussion of proposed changes to the codebase, not a discussion of Riccardo. Let's keep the discussion as impersonal as possible.

Pretend the PR wasn't submitted and consider the single commits as if they were pushed to the master, you will see that every change has its own commit and the fact they are grouped on a single PR by the github interface doesn't make it less true.

stephengold · 2023-09-07T16:44:13Z

Does the size of the PR reflects on the quality of the code?

No.

riccardobl and others added 5 commits September 4, 2023 14:46

Never return the cached spatial directly: prevent the manipulation of…

90a28b1

… the first instance of a spatial from poisoning the cache.

Add extension to load speaker nodes

da8186d

Add userdata loader and make it the default

d8e6b27

Make default extensions registrable statically, move all extensions i…

394973d

…nto packages named after their fully qualified name

auto-format

4390373

Ali-RS reviewed Sep 5, 2023

View reviewed changes

Ali-RS reviewed Sep 6, 2023

View reviewed changes

Ali-RS reviewed Sep 7, 2023

View reviewed changes

riccardobl marked this pull request as draft October 14, 2023 12:43

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Improve GLTF support #2069

Improve GLTF support #2069

riccardobl commented Sep 4, 2023 •

edited

Ali-RS Sep 5, 2023 •

edited

Ali-RS Sep 5, 2023 •

edited

Ali-RS Sep 5, 2023 •

edited

Ali-RS Sep 5, 2023 •

edited

Ali-RS commented Sep 5, 2023 •

edited

Ali-RS Sep 6, 2023

Ali-RS Sep 6, 2023

stephengold commented Sep 6, 2023

riccardobl commented Sep 6, 2023

stephengold commented Sep 6, 2023

stephengold commented Sep 6, 2023

riccardobl commented Sep 6, 2023 •

edited

Ali-RS commented Sep 7, 2023 •

edited

Ali-RS commented Sep 7, 2023 •

edited

Ali-RS Sep 7, 2023 •

edited

Ali-RS commented Sep 7, 2023 •

edited

stephengold commented Sep 7, 2023 •

edited

riccardobl commented Sep 7, 2023

stephengold commented Sep 7, 2023

Improve GLTF support #2069

Are you sure you want to change the base?

Improve GLTF support #2069

Conversation

riccardobl commented Sep 4, 2023 • edited

Ali-RS Sep 5, 2023 • edited

Choose a reason for hiding this comment

Ali-RS Sep 5, 2023 • edited

Choose a reason for hiding this comment

Ali-RS Sep 5, 2023 • edited

Choose a reason for hiding this comment

Ali-RS Sep 5, 2023 • edited

Choose a reason for hiding this comment

Ali-RS commented Sep 5, 2023 • edited

Ali-RS Sep 6, 2023

Choose a reason for hiding this comment

Ali-RS Sep 6, 2023

Choose a reason for hiding this comment

stephengold commented Sep 6, 2023

riccardobl commented Sep 6, 2023

stephengold commented Sep 6, 2023

stephengold commented Sep 6, 2023

riccardobl commented Sep 6, 2023 • edited

Ali-RS commented Sep 7, 2023 • edited

Ali-RS commented Sep 7, 2023 • edited

Ali-RS Sep 7, 2023 • edited

Choose a reason for hiding this comment

Ali-RS commented Sep 7, 2023 • edited

stephengold commented Sep 7, 2023 • edited

riccardobl commented Sep 7, 2023

stephengold commented Sep 7, 2023

riccardobl commented Sep 4, 2023 •

edited

Ali-RS Sep 5, 2023 •

edited

Ali-RS Sep 5, 2023 •

edited

Ali-RS Sep 5, 2023 •

edited

Ali-RS Sep 5, 2023 •

edited

Ali-RS commented Sep 5, 2023 •

edited

riccardobl commented Sep 6, 2023 •

edited

Ali-RS commented Sep 7, 2023 •

edited

Ali-RS commented Sep 7, 2023 •

edited

Ali-RS Sep 7, 2023 •

edited

Ali-RS commented Sep 7, 2023 •

edited

stephengold commented Sep 7, 2023 •

edited