Enable StoreResolvedURIs when loading SDF

[iche033]

🎉 New feature

Summary

Replaces #2323 - see description in that PR.

This PR is needed so that bullet-featherstone plugin in gz-physics can load meshes with relative path URIs.

Checklist

• Signed all commits for DCO
• Added tests
• Added example and/or tutorial
• Updated documentation (as needed)
• Updated migration guide (as needed)
• Consider updating Python bindings (if the library has them)
• codecheck passed (See contributing)
• All tests passed (See test coverage)
• While waiting for a review on your PR, please help review another open pull request to support the maintainers

Note to maintainers: Remember to use Squash-Merge and edit the commit message to match the pull request summary while retaining Signed-off-by messages.

[iche033]

The ServerFixture.LoadSdfModelRelativeUri test in UNIT_Server_Rendering_TEST.cc was failing because it contains a mesh uri pointing to a non-existent dae file. With SetStoreResolvedURIs set to true, it returns an error and so the server stops loading the model. This breaks current behavior. To resolve this, I added logic in ddf7f46 to allow the server to proceed with presence of URI_LOOKUP errors.

[iche033]

The ServerFixture.LoadSdfModelRelativeUri test in UNIT_Server_Rendering_TEST.cc was failing because it contains a mesh uri pointing to a non-existent dae file. With SetStoreResolvedURIs set to true, it returns an error and so the server stops loading the model. This breaks current behavior. To resolve this, I added logic in ddf7f46 to allow the server to proceed with presence of URI_LOOKUP errors.

Looking for feedback on this. Any objections to allowing Server to continue loading with presence of URI errors?

[azeey]

I'm not sure this is a good idea. The URI_LOOKUP error is not just emitted for missing mesh files, but also when an included model is not found, which should be a blocker error IMO.

[iche033]

The other option is to not return the URI_LOOKUP error in sdformat when resolving URIs but that's also not ideal.

So with SetStoreResolvedURIs now set to true, we'll just be more strict and stop the Server from loading if an URI is not found. How does that sound?

[azeey]

Maybe we can define a parameter in ServerConfig that determines if a SDF errors prevent the Server from loading. Or it could be an environment variable. Either way, we can just set that in the test without affecting end users.

[iche033]

Maybe we can define a parameter in ServerConfig that determines if a SDF errors prevent the Server from loading.?

Added a new parameter in ServerConfig, blockOnSdfErrors to stop server from loading in presence of SDF errors. It can also be passed from the command line, gz sim -v 4 shapes.sdf --block-on-sdf-errors 1. fbddcb7

[azeey]

I don't know if we should make this available for end users. When I was suggesting the feature, I only had our unit tests in mind where we would use it only in a narrowly scoped use case (invalid mesh path). My concern is that users can now use this flag and would be confused when things don't work as expected. Is there an actual use case for enabling this?

[iche033]

More of a nice to have. I have sdf files with a few uri paths that don't exist on my filesystem and this lets me launch gazebo with only the models with valid URIs instead of exiting. I can also just comment them out.

I removed the changes here. We can revisit if we have a better or more common use case. afd6845

[azeey]

I think there's an unnecessary diff and DCO needs to be fixed. Otherwise, LGTM!

[azeey]

Remove this diff?

[iche033]

removed.

[iche033]

DCO needs to be fixed.

fixed and force pushed

[iche033]

fixed a couple of windows failures in 7488f7e.