BasicShaderTestSuite: Rework so each shader reports its own result

[YYDan]

Description

BasicShaderTestSuite was never covered by the summer 2024 refactor campaign, and it does not fit the standard split pattern — it needs investigation before any work is committed to.

The suite is 1,194 lines reporting 21 results, but the assert count is misleading: most facts contain a single assert_ call inside a loop over a shared shaders[] array. So shader_get/set_uniform_f (83 lines), shader_set_uniform_f_array (75 lines), shader_set_uniform_f_buffer (88 lines), shader_set_uniform_i (79 lines), shader_set_uniform_i_array (74 lines) and shader_get_sampler_index (69 lines) each report one pass/fail for what is actually one check per shader in the array. Two further facts, normals_test (76 lines) and gl_front_facing/sv_is_front_face (79 lines), contain no direct asserts at all and verify through helpers.

The right shape here is almost certainly addTheory() over the shader list rather than a flat fact split — that would give one reported result per shader per uniform function, which is what we actually want, without duplicating the loop body.

The two helper-based facts need a separate decision about how their verification should surface as a result.

This issue should start with an investigation and a written proposal, not with edits.

Script Name

BasicShaderTestSuite

Steps To The Task

• Audit the suite and document how many distinct checks are actually performed behind each of the 21 reported results
• Evaluate addTheory() over the shaders[] array as the primary approach and post the proposal on this issue for review before starting
• Decide how normals_test and gl_front_facing/sv_is_front_face should report, given they verify through helpers rather than direct asserts
• Implement the agreed approach
• Add section comment headers per uniform family
• Run the suite on VM and YYC across all supported shader languages and confirm the same set of checks pass as before