Correctly truncate and deduplicate strings (#492)

Author: SR4ven

CHANGELOG.rst

@@ -29,6 +29,9 @@ Fixes
 - Fixed a crash when using the network monitor
 - Session can now be "quiet" by passing an empty list of loggers
 - Process Monitor: fixed Thread.isAlive for Python 3.9 compability
+- Correctly truncate values of the string primitive when max_len or size is set.
+- The string primitive will no longer generate duplicates when max_len or size is set.
+- Greatly improved string to bytes conversion speed.
 
 v0.2.1
 ------

boofuzz/__init__.py

@@ -649,7 +649,7 @@ def s_random(value="", min_length=0, max_length=1, num_mutations=25, fuzzable=Tr
     Generate a random chunk of data while maintaining a copy of the original. A random length range can be specified.
     For a static length, set min/max length to be the same.
 
-    :type  value:         Raw
+    :type  value:         str or bytes
     :param value:         (Optional, def="") Original value
     :type  min_length:    int
     :param min_length:    (Optional, def=0) Minimum length of random block
@@ -705,25 +705,30 @@ def s_mirror(primitive_name=None, name=None):
     blocks.CURRENT.push(Mirror(name=name, primitive_name=primitive_name, request=blocks.CURRENT))
 
 
-def s_string(value="", size=-1, padding=b"\x00", encoding="ascii", fuzzable=True, max_len=-1, name=None):
+def s_string(value="", size=None, padding=b"\x00", encoding="ascii", fuzzable=True, max_len=None, name=None):
     """
     Push a string onto the current block stack.
 
     :type  value:    str
     :param value:    (Optional, def="")Default string value
     :type  size:     int
-    :param size:     (Optional, def=-1) Static size of this field, leave -1 for dynamic.
+    :param size:     (Optional, def=None) Static size of this field, leave None for dynamic.
     :type  padding:  Character
     :param padding:  (Optional, def="\\x00") Value to use as padding to fill static field size.
     :type  encoding: str
-    :param encoding: (Optonal, def="ascii") String encoding, ex: utf_16_le for Microsoft Unicode.
+    :param encoding: (Optional, def="ascii") String encoding, ex: utf_16_le for Microsoft Unicode.
     :type  fuzzable: bool
     :param fuzzable: (Optional, def=True) Enable/disable fuzzing of this primitive
     :type  max_len:  int
-    :param max_len:  (Optional, def=-1) Maximum string length
+    :param max_len:  (Optional, def=None) Maximum string length
     :type  name:     str
     :param name:     (Optional, def=None) Specifying a name gives you direct access to a primitive
     """
+    # support old interface where default was -1 instead of None
+    if size == -1:
+        size = None
+    if max_len == -1:
+        max_len = None
 
     blocks.CURRENT.push(
         String(

boofuzz/fuzzable.py

@@ -25,12 +25,12 @@ class Fuzzable(object):
     :type name: str, optional
     :param name: Name, for referencing later. Names should always be provided, but if not, a default name will be given,
         defaults to None
-    :type default_value: Any
+    :type default_value: Any, optional
     :param default_value: Value used when the element is not being fuzzed - should typically represent a valid value.
         Can be a static value, or a ReferenceValueTestCaseSession, defaults to None
-    :type fuzzable: bool
+    :type fuzzable: bool, optional
     :param fuzzable: Enable fuzzing of this primitive, defaults to True
-    :type fuzz_values: list
+    :type fuzz_values: list, optional
     :param fuzz_values: List of custom fuzz values to add to the normal mutations, defaults to None
     """
 
@@ -62,9 +62,6 @@ def name(self):
 
         :rtype: str
         """
-        # if self._name is None:
-        #     Fuzzable.name_counter += 1
-        #     self._name = "{0}{1}".format(type(self).__name__, Fuzzable.name_counter)
         return self._name
 
     @property

boofuzz/helpers.py

@@ -439,11 +439,5 @@ def get_boofuzz_version(boofuzz_class):
     return "v-.-.-"
 
 
-def str_to_bytes(value):
-    result = value
-    # if python2, str is already bytes compatible
-    if six.PY3:
-        if isinstance(value, six.text_type):
-            temp = [bytes([ord(i)]) for i in value]
-            result = b"".join(temp)
-    return result
+def str_to_bytes(value, encoding="utf-8", errors="replace"):
+    return six.ensure_binary(value, encoding=encoding, errors=errors)

boofuzz/primitives/random_data.py

@@ -18,7 +18,7 @@ class RandomData(Fuzzable):
     :type name: str, optional
     :param default_value: Value used when the element is not being fuzzed - should typically represent a valid value,
         defaults to None
-    :type default_value: Any, optional
+    :type default_value: str or bytes, optional
     :param min_length: Minimum length of random block, defaults to 0
     :type min_length: int, optional
     :param max_length: Maximum length of random block, defaults to 1