Function TestShouldInternColumn

string_interning_test.go:51–92

(t *testing.T)

}

func TestShouldInternColumn(t *testing.T) {
	tests := []struct {
		name      string
		values    []string
		threshold float64
		expected  bool
	}{
		{
			name:      "High cardinality - should not intern",
			values:    generateUniqueStrings(1000),
			threshold: 0.30,
			expected:  false,
		},
		{
			name:      "Low cardinality - should intern",
			values:    generateRepeatedStrings(1000, 10),
			threshold: 0.30,
			expected:  true,
		},
		{
			name:      "Too small - should not intern",
			values:    []string{"a", "b", "c"},
			threshold: 0.30,
			expected:  false,
		},
		{
			name:      "Medium cardinality at threshold",
			values:    generateRepeatedStrings(1000, 300),
			threshold: 0.30,
			expected:  false,
		},
	}

	for _, tt := range tests {
		t.Run(tt.name, func(t *testing.T) {
			result := shouldInternColumn(tt.values, tt.threshold)
			if result != tt.expected {
				t.Errorf("shouldInternColumn() = %v, want %v", result, tt.expected)
			}
		})
	}
}

func TestBufferStringInterning(t *testing.T) {
	// Create test data with repeated categorical values