doc_processer/test_omml_preprocessing.py

"""Comprehensive test for OMML conversion with preprocessing."""

from app.services.converter import Converter


def test_case_1_array_with_spaces():
    """Test: Array with spaces in column specifier (the original issue)."""
    print("\n" + "=" * 80)
    print("Test 1: Array with spaces in column specifier")
    print("=" * 80)
    
    converter = Converter()
    
    # The problematic LaTeX from the error
    latex = r"""\begin{array}{l} D = \left| \begin{array}{c c c c} a _ {1 1} & a _ {1 2} & \dots & a _ {1 n} \\ \vdots & \vdots & & \vdots \\ a _ {i 1} + 0 + \dots + 0 & 0 + a _ {i 2} + \dots + 0 & \dots & 0 + \dots + 0 + a _ {i n} \\ \vdots & \vdots & & \vdots \\ a _ {n 1} & a _ {n 2} & \dots & a _ {n n} \end{array} \right| \\ = \left| \begin{array}{c c c c} a _ {1 1} & a _ {1 2} & \dots & a _ {1 n} \\ \vdots & \vdots & & \vdots \\ a _ {i 1} & 0 & \dots & 0 \\ \vdots & \vdots & & \vdots \\ a _ {n 1} & a _ {n 2} & \dots & a _ {n n} \end{array} \right| + \left| \begin{array}{c c c c} a _ {1 1} & a _ {1 2} & \dots & a _ {1 n} \\ \vdots & \vdots & & \vdots \\ 0 & a _ {i 2} & \dots & 0 \\ \vdots & \vdots & & \vdots \\ a _ {n 1} & a _ {n 2} & \dots & a _ {n n} \end{array} \right| \\ + \dots + \left| \begin{array}{c c c c} a _ {1 1} & a _ {1 2} & \dots & a _ {1 n} \\ \vdots & \vdots & & \vdots \\ 0 & 0 & \dots & a _ {i n} \\ \vdots & \vdots & & \vdots \\ a _ {n 1} & a _ {n 2} & \dots & a _ {n n} \end{array} \right|, \\ \end{array}"""
    
    print(f"LaTeX length: {len(latex)} chars")
    print(f"Preview: {latex[:100]}...")
    
    try:
        omml = converter.convert_to_omml(latex)
        print(f"\n✓ SUCCESS: Converted to OMML")
        print(f"OMML length: {len(omml)} chars")
        
        if "oMath" in omml:
            print("✓ Valid OMML structure detected")
        
        # Check preprocessing worked
        preprocessed = converter._preprocess_formula_for_omml(latex)
        if "{c c c c}" not in preprocessed and "{cccc}" in preprocessed:
            print("✓ Array column specifiers fixed: '{c c c c}' → '{cccc}'")
        
        return True
        
    except Exception as e:
        print(f"\n✗ FAILED: {e}")
        return False


def test_case_2_vmatrix():
    """Test: vmatrix environment conversion."""
    print("\n" + "=" * 80)
    print("Test 2: vmatrix environment")
    print("=" * 80)
    
    converter = Converter()
    
    latex = r"\begin{vmatrix} a & b \\ c & d \end{vmatrix}"
    print(f"LaTeX: {latex}")
    
    try:
        omml = converter.convert_to_omml(latex)
        print(f"✓ SUCCESS: Converted to OMML ({len(omml)} chars)")
        
        # Check if vmatrix was converted
        preprocessed = converter._preprocess_formula_for_omml(latex)
        if "vmatrix" not in preprocessed and r"\left|" in preprocessed:
            print("✓ vmatrix converted to \\left| ... \\right|")
        
        return True
        
    except Exception as e:
        print(f"✗ FAILED: {e}")
        return False


def test_case_3_cases_environment():
    """Test: cases environment conversion."""
    print("\n" + "=" * 80)
    print("Test 3: cases environment")
    print("=" * 80)
    
    converter = Converter()
    
    latex = r"f(x) = \begin{cases} x^2 & x \geq 0 \\ -x & x < 0 \end{cases}"
    print(f"LaTeX: {latex}")
    
    try:
        omml = converter.convert_to_omml(latex)
        print(f"✓ SUCCESS: Converted to OMML ({len(omml)} chars)")
        
        # Check if cases was converted to array
        preprocessed = converter._preprocess_formula_for_omml(latex)
        if "cases" not in preprocessed and "array" in preprocessed:
            print("✓ cases converted to array environment")
        
        return True
        
    except Exception as e:
        print(f"✗ FAILED: {e}")
        return False


def test_case_4_aligned_environment():
    """Test: aligned environment conversion."""
    print("\n" + "=" * 80)
    print("Test 4: aligned environment")
    print("=" * 80)
    
    converter = Converter()
    
    latex = r"\begin{aligned} x + y &= 5 \\ 2x - y &= 1 \end{aligned}"
    print(f"LaTeX: {latex}")
    
    try:
        omml = converter.convert_to_omml(latex)
        print(f"✓ SUCCESS: Converted to OMML ({len(omml)} chars)")
        
        # Check if aligned was converted
        preprocessed = converter._preprocess_formula_for_omml(latex)
        if "aligned" not in preprocessed and "array" in preprocessed:
            print("✓ aligned converted to array environment")
        if "&" not in preprocessed or preprocessed.count("&") < latex.count("&"):
            print("✓ Alignment markers removed")
        
        return True
        
    except Exception as e:
        print(f"✗ FAILED: {e}")
        return False


def test_case_5_simple_formula():
    """Test: Simple formula (should work without preprocessing)."""
    print("\n" + "=" * 80)
    print("Test 5: Simple formula")
    print("=" * 80)
    
    converter = Converter()
    
    latex = r"x = \frac{-b \pm \sqrt{b^2 - 4ac}}{2a}"
    print(f"LaTeX: {latex}")
    
    try:
        omml = converter.convert_to_omml(latex)
        print(f"✓ SUCCESS: Converted to OMML ({len(omml)} chars)")
        return True
        
    except Exception as e:
        print(f"✗ FAILED: {e}")
        return False


def test_case_6_nested_structures():
    """Test: Nested structures with multiple issues."""
    print("\n" + "=" * 80)
    print("Test 6: Nested structures")
    print("=" * 80)
    
    converter = Converter()
    
    latex = r"\left\{ \begin{array}{l c} \begin{vmatrix} a & b \\ c & d \end{vmatrix} & = ad - bc \\ f(x) = \begin{cases} 1 & x > 0 \\ 0 & x \leq 0 \end{cases} & \text{step function} \end{array} \right."
    print(f"LaTeX: {latex}")
    
    try:
        omml = converter.convert_to_omml(latex)
        print(f"✓ SUCCESS: Converted to OMML ({len(omml)} chars)")
        
        preprocessed = converter._preprocess_formula_for_omml(latex)
        print("\nPreprocessing applied:")
        if "vmatrix" not in preprocessed:
            print("  ✓ vmatrix converted")
        if "cases" not in preprocessed:
            print("  ✓ cases converted")
        if "{l c}" not in preprocessed and "{lc}" in preprocessed:
            print("  ✓ Array specifiers fixed")
        
        return True
        
    except Exception as e:
        print(f"✗ FAILED: {e}")
        return False


if __name__ == "__main__":
    print("=" * 80)
    print("OMML CONVERSION TEST SUITE")
    print("Testing preprocessing and conversion")
    print("=" * 80)
    
    results = []
    
    try:
        results.append(("Simple formula", test_case_5_simple_formula()))
        results.append(("Array with spaces", test_case_1_array_with_spaces()))
        results.append(("vmatrix", test_case_2_vmatrix()))
        results.append(("cases", test_case_3_cases_environment()))
        results.append(("aligned", test_case_4_aligned_environment()))
        results.append(("Nested structures", test_case_6_nested_structures()))
        
        # Summary
        print("\n" + "=" * 80)
        print("TEST SUMMARY")
        print("=" * 80)
        
        passed = sum(1 for _, result in results if result)
        total = len(results)
        
        for name, result in results:
            status = "✓ PASS" if result else "✗ FAIL"
            print(f"{status}: {name}")
        
        print("\n" + "-" * 80)
        print(f"Total: {passed}/{total} tests passed")
        
        if passed == total:
            print("\n✓✓✓ ALL TESTS PASSED ✓✓✓")
        else:
            print(f"\n✗✗✗ {total - passed} TESTS FAILED ✗✗✗")
        
        print("=" * 80)
        
    except KeyboardInterrupt:
        print("\n\nTests interrupted by user")
    except Exception as e:
        print(f"\n\nTest suite error: {e}")
        import traceback
        traceback.print_exc()