Add splicing module and related tests

- Add src/splicing/ module with scaffold_prep, fragment_prep, and engine
- Add tylosin_splicer.py entry script
- Add unit tests for splicing components

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

tests/test_fragment_prep.py

@@ -0,0 +1,95 @@
+import pytest
+from rdkit import Chem
+from src.splicing.fragment_prep import activate_fragment
+
+def test_activate_smart_ethanol():
+    """Test 'smart' activation on Ethanol (CCO). Should attach to Oxygen."""
+    smiles = "CCO"
+    mol = activate_fragment(smiles, strategy="smart")
+    
+    # Check if we have a dummy atom
+    assert mol is not None
+    assert mol.GetNumAtoms() == 4  # C, C, O, *
+    
+    # Check if the dummy atom is attached to Oxygen
+    # Find the dummy atom
+    dummy_atom = None
+    for atom in mol.GetAtoms():
+        if atom.GetSymbol() == '*':
+            dummy_atom = atom
+            break
+    
+    assert dummy_atom is not None
+    
+    # Check neighbors of dummy atom
+    neighbors = dummy_atom.GetNeighbors()
+    assert len(neighbors) == 1
+    assert neighbors[0].GetSymbol() == 'O'
+    
+    # Check output SMILES format
+    out_smiles = Chem.MolToSmiles(mol)
+    assert '*' in out_smiles
+
+def test_activate_smart_amine():
+    """Test 'smart' activation on Ethylamine (CCN). Should attach to Nitrogen."""
+    smiles = "CCN"
+    mol = activate_fragment(smiles, strategy="smart")
+    
+    assert mol is not None
+    
+    # Find the dummy atom
+    dummy_atom = None
+    for atom in mol.GetAtoms():
+        if atom.GetSymbol() == '*':
+            dummy_atom = atom
+            break
+            
+    assert dummy_atom is not None
+    neighbors = dummy_atom.GetNeighbors()
+    assert neighbors[0].GetSymbol() == 'N'
+
+def test_activate_random_pentane():
+    """Test 'random' activation on Pentane (CCCCC). Should attach to a Carbon."""
+    smiles = "CCCCC"
+    # Seed is not easily passed to the function unless we add it to the signature or fix it inside.
+    # For this test, any Carbon is fine.
+    mol = activate_fragment(smiles, strategy="random")
+    
+    assert mol is not None
+    assert mol.GetNumAtoms() == 6 # 5 C + 1 *
+    
+    dummy_atom = None
+    for atom in mol.GetAtoms():
+        if atom.GetSymbol() == '*':
+            dummy_atom = atom
+            break
+            
+    assert dummy_atom is not None
+    neighbors = dummy_atom.GetNeighbors()
+    assert neighbors[0].GetSymbol() == 'C'
+
+def test_activate_smart_fallback():
+    """Test 'smart' fallback when no heteroatoms are found (e.g. Propane)."""
+    smiles = "CCC"
+    # Should fall back to finding a terminal carbon or random
+    # The requirement says "fall back to a terminal Carbon" or random. 
+    # Let's assume the implementation picks a terminal carbon if possible, or just behaves like random on C.
+    mol = activate_fragment(smiles, strategy="smart")
+    
+    assert mol is not None
+    dummy_atom = None
+    for atom in mol.GetAtoms():
+        if atom.GetSymbol() == '*':
+            dummy_atom = atom
+            break
+    
+    assert dummy_atom is not None
+    neighbor = dummy_atom.GetNeighbors()[0]
+    assert neighbor.GetSymbol() == 'C'
+    # Verify it's a valid molecule
+    assert Chem.SanitizeMol(mol) == Chem.SanitizeFlags.SANITIZE_NONE
+
+def test_invalid_smiles():
+    with pytest.raises(ValueError):
+        activate_fragment("NotASmiles", strategy="smart")
+