FIX: validate UnicodeSubstitution output code points

pyrit/converter/unicode_sub_converter.py

@@ -20,7 +20,12 @@ def __init__(self, *, start_value: int = 0xE0000) -> None:
 
         Args:
             start_value (int): The unicode starting point to use for encoding.
+
+        Raises:
+            ValueError: If ``start_value`` is outside the Unicode code point range.
         """
+        if not 0 <= start_value <= 0x10FFFF:
+            raise ValueError("start_value must be a valid Unicode code point between 0 and 0x10FFFF")
         self.startValue = start_value
 
     def _build_identifier(self) -> ComponentIdentifier:
@@ -49,10 +54,18 @@ async def convert_async(self, *, prompt: str, input_type: PromptDataType = "text
             ConverterResult: The result containing the converted output and its type.
 
         Raises:
-            ValueError: If the input type is not supported.
+            ValueError: If the input type is not supported or a substitution falls outside the Unicode range.
         """
         if not self.input_supported(input_type):
             raise ValueError("Input type not supported")
 
-        ret_text = "".join(chr(self.startValue + ord(ch)) for ch in prompt)
-        return ConverterResult(output_text=ret_text, output_type="text")
+        output_chars: list[str] = []
+        for char in prompt:
+            code_point = self.startValue + ord(char)
+            if code_point > 0x10FFFF:
+                raise ValueError(
+                    f"Unicode substitution for character {char!r} exceeds the maximum code point U+10FFFF"
+                )
+            output_chars.append(chr(code_point))
+
+        return ConverterResult(output_text="".join(output_chars), output_type="text")

tests/unit/converter/test_unicode_sub_converter.py

@@ -22,6 +22,19 @@ async def test_unicode_sub_custom_start():
     assert result.output_type == "text"
 
 
+@pytest.mark.parametrize("start_value", [-1, 0x110000])
+def test_unicode_sub_rejects_invalid_start_value(start_value):
+    with pytest.raises(ValueError, match="valid Unicode code point"):
+        UnicodeSubstitutionConverter(start_value=start_value)
+
+
+async def test_unicode_sub_rejects_derived_code_point_overflow():
+    converter = UnicodeSubstitutionConverter(start_value=0x10FFFF)
+
+    with pytest.raises(ValueError, match="exceeds the maximum code point"):
+        await converter.convert_async(prompt="a", input_type="text")
+
+
 async def test_unicode_sub_empty():
     converter = UnicodeSubstitutionConverter()
     result = await converter.convert_async(prompt="", input_type="text")