Fix mixed precision detection in resolve.jl tests

Use string matching to detect mixed precision algorithms instead of symbol comparison.
This ensures the tolerance branch is properly taken for algorithms like RF32MixedLUFactorization.

Author: claude

test/resolve.jl

@@ -4,15 +4,11 @@ using LinearSolve: AbstractDenseFactorization, AbstractSparseFactorization,
                    AMDGPUOffloadLUFactorization, AMDGPUOffloadQRFactorization,
                    SparspakFactorization
 
-# Define mixed precision algorithms that need higher tolerance
-const MIXED_PRECISION_ALGS = [
-    :MKL32MixedLUFactorization,
-    :AppleAccelerate32MixedLUFactorization,
-    :OpenBLAS32MixedLUFactorization,
-    :RF32MixedLUFactorization,
-    :CUDAOffload32MixedLUFactorization,
-    :MetalOffload32MixedLUFactorization
-]
+# Function to check if an algorithm is mixed precision
+function is_mixed_precision_alg(alg)
+    alg_name = string(alg)
+    return contains(alg_name, "32Mixed") || contains(alg_name, "Mixed32")
+end
 
 for alg in vcat(InteractiveUtils.subtypes(AbstractDenseFactorization),
     InteractiveUtils.subtypes(AbstractSparseFactorization))
@@ -61,7 +57,7 @@ for alg in vcat(InteractiveUtils.subtypes(AbstractDenseFactorization),
 
         # Use higher tolerance for mixed precision algorithms
         expected = [-2.0, 1.5]
-        if Symbol(alg) in MIXED_PRECISION_ALGS
+        if is_mixed_precision_alg(alg)
             @test solve!(linsolve).u ≈ expected atol=1e-4 rtol=1e-4
             @test !linsolve.isfresh
             @test solve!(linsolve).u ≈ expected atol=1e-4 rtol=1e-4
@@ -82,7 +78,7 @@ for alg in vcat(InteractiveUtils.subtypes(AbstractDenseFactorization),
         @test linsolve.isfresh
 
         # Use higher tolerance for mixed precision algorithms
-        if Symbol(alg) in MIXED_PRECISION_ALGS
+        if is_mixed_precision_alg(alg)
             @test solve!(linsolve).u ≈ expected atol=1e-4 rtol=1e-4
         else
             @test solve!(linsolve).u ≈ expected