first commit

backend/app/dataset/rules/python/boundary_error.py

@@ -0,0 +1,50 @@
+"""AST-level boundary condition injection for Python."""
+
+import ast
+from typing import Optional
+
+from app.dataset.rules.base import Mutation, MutationRule
+
+
+class BoundaryErrorRule(MutationRule):
+    """Mutate a list-index boundary check from `< len(seq)` to `<= len(seq)`.
+
+    Uses `ast` to find comparisons guarding index access and flips the operator
+    so the boundary becomes off-by-one.
+    """
+
+    name = "boundary_error"
+    language = "python"
+    defect_type = "boundary_condition_error"
+
+    def detect_and_mutate(self, source: str, filename: str = "") -> Optional[Mutation]:
+        try:
+            tree = ast.parse(source)
+        except SyntaxError:
+            return None
+
+        for node in ast.walk(tree):
+            if not isinstance(node, ast.If):
+                continue
+            test = node.test
+            if isinstance(test, ast.Compare) and isinstance(test.left, ast.Name):
+                if len(test.ops) == 1 and isinstance(test.ops[0], ast.Lt):
+                    # i < len(x)  ->  i <= len(x)
+                    comparator = test.comparators[0]
+                    if isinstance(comparator, ast.Call) and isinstance(comparator.func, ast.Name) and comparator.func.id == "len":
+                        lines = source.splitlines(keepends=True)
+                        line = lines[node.lineno - 1]
+                        mutated_line = line.replace("< len(", "<= len(", 1)
+                        if mutated_line == line:
+                            continue
+                        mutated = "".join(lines[: node.lineno - 1] + [mutated_line] + lines[node.lineno:])
+                        return Mutation(
+                            defect_type=self.defect_type,
+                            language=self.language,
+                            line_start=node.lineno,
+                            line_end=getattr(node, "end_lineno", node.lineno) or node.lineno,
+                            mutated_source=mutated,
+                            reference_fix="Use strict `< len(seq)` to avoid index-out-of-range.",
+                            description="Changed index boundary check to off-by-one (<= len).",
+                        )
+        return None