feat: Add initial benchmarks, integrate them into CI & add getters/settters for Scripts resource

[makspll]

Summary

• Moves tests to run centrally in the workspace tests directory
• Refactors test utilities to share a lot of the test discovery and running code between languages and benchmarks
• Adds criterion benchmarks
• Adds CI for snapshoting performance from main to bencher.dev

I've also noticed that Scripts didn't really have nice ways to be accessed by external crates, so I added a bunch of accessors

[Copilot]

Pull Request Overview

This PR centralizes test discovery and execution while introducing basic benchmarking support with Criterion and CI integration for performance snapshoting.

• Centralizes test running by moving tests to a common directory and refactoring test utilities.
• Introduces benchmark executors and grouping logic to run language-specific benchmarks.
• Adds CI workflow for continuous performance benchmarking with bencher.dev.

Reviewed Changes

Copilot reviewed 12 out of 18 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
tests/script_tests.rs	Introduces a centralized test runner with improved test discovery for Lua and Rhai integration tests.
.github/workflows/snapshot_benchmark_main.yml	Adds a new GitHub Actions workflow to run and snapshot performance benchmarks.
benches/benchmarks.rs	Implements benchmark grouping and execution for scripts using Criterion.
crates/testing_crates/test_utils/src/lib.rs	Updates test discovery logic with minor refactoring of directory traversal.
crates/testing_crates/script_integration_test_harness/src/lib.rs	Refactors integration test harness to support benchmarking and adds language-specific plugin initializers.
crates/bevy_mod_scripting_core/src/script.rs	Extends the Scripts API with additional collection methods.
Cargo.toml and language crate Cargo.toml files	Updates dependency and workspace configurations to support the changes and remove redundant tests from language crates.

Files not reviewed (6)

• assets/benchmarks/component/access.lua: Language not supported
• assets/benchmarks/component/access.rhai: Language not supported
• assets/benchmarks/component/get.lua: Language not supported
• assets/benchmarks/component/get.rhai: Language not supported
• crates/languages/bevy_mod_scripting_lua/tests/data/construct/construct_unit_struct.lua: Language not supported
• crates/languages/bevy_mod_scripting_rhai/tests/data/construct/unit_struct.rhai: Language not supported

Comments suppressed due to low confidence (2)

tests/script_tests.rs:35

• Using unwrap on the result of splitting the path assumes the delimiter 'tests{sep}data' is always present, which may lead to a panic if the expected structure is not met. Consider using a more robust method or adding a check to handle missing segments gracefully.

                .to_string_lossy().split(&format!("tests{}data", std::path::MAIN_SEPARATOR)).last().unwrap()

benches/benchmarks.rs:29

• Indexing directly into the split result at index 1 may lead to a panic if the resulting vector has fewer elements than expected. Consider verifying the vector length before indexing to ensure robust benchmark grouping.

let first_folder = path.split("/").collect::<Vec<&str>>()[1];