Fix pe parsing for section with 8bytes name

tests/test_parse_pe.py

@@ -91,9 +91,6 @@ def test_pe_parsing_strange_optional_header_size(tmp_path, proc32):
     # Also check that section retrieval works (as its position is based on OptionalHeader Size)
     assert set(s.name for s in mod.pe.sections) == {".text", ".data", ".l1"}
 
-
-
-
 #  Make a test from current_process parsing ?
 def test_pe_parsing_dotnet32_process_64(proc64, pe_dotnet32):
     # .NET pe32 loadable in 64bit process -> rewrite of the OptionalHeader
@@ -126,6 +123,27 @@ def test_pe_parsing_dotnet32_current_process_64(proc64, pe_dotnet32):
         assert imported_dlls == ['mscoree.dll']
         assert ".text" in  sections_names
 
+# PE header of Syschpe32\ntdll.dll with a 8 chars sections .hexpthk that broke the parseur
+# This PE as 8 bytes sections name + VirtualSize non aligned on 0x100 so there is a non-null non-ascii byte after the name
+PE_SECTION_8CHARS = b"""
+eJzzjZrAwMzAwMACxP//MzDsYIAABwbC4AMRagYbCHBlYPBhZEYRu8HAxMjNyMnAwMQAwSAgAMUK
+II4DhA2UY4VKw2hwQLFAmBBTFSBqQYQAwig4MECyg4ogAGiuMRZxvQgIuMDIAPcDAxuqGgUGhgS9
+SAjQhgqA1XGgqnNgYDigFwUBYHmYX7gwzHtAmW9GwSgYBfQAnU8b3zAAy4aMr/+BQNTGwIGht0aF
+IwMkF54BKg0yGBQcGAL+izoApTrfHJ4wvMAAB/+AA4/UnJx8hfL8opwUhbSi/FyFRIWAypKM/Dy3
+/KLwzLyU/PJihZLU4hIFUMU5CoYd0DBAsJuAbBsD7OpA7YsEIM4B4goD3GIMDO6pJcElKR6JeSk5
+qSB+eFFmSapbJpgTlJqYAmUyAFOXa0VqMkN2alFeao6xkV5KTg6NPDkKcAIAM7/CEw=="""
+
+def test_pe_parsing_section_8_chars(proc32, tmp_path):
+    pe_path = (tmp_path / "pe_parsing_8chars_section.exe")
+    pe_data = pe_data = zlib.decompress(b64decode(PE_SECTION_8CHARS))
+
+    with pe_path.open("wb") as f:
+        f.write(pe_data)
+
+    mod = proc32.load_library(str(pe_path))
+    # check that section retrieval works
+    assert set(s.name for s in mod.pe.sections) == {u".XXXXXXX", u".YYYYYYY", u".ZZZZZZZ"}
+
 # A "Portable Executable 32 .NET Assembly" DLL
 # Result of compiling a simple hello-world
 # Can be loaded into a 64b process to witness 32 -> 64b PE conversion at load time

windows/pe_parse.py

@@ -109,7 +109,7 @@ class PESection(IMAGE_SECTION_HEADER):
         if self.target is None:
             name = get_string(self.target, ctypes.addressof(self.Name))[:8]
         else:
-            name = get_string(self.target, self._base_addr)[:8]
+            name = self.target.read_memory(self._base_addr, 8).split(b"\x00", 1)[0].decode("ascii")
         # Decode as UTF-8 as the MS doc say ?
         return name
 
@@ -451,4 +451,17 @@ class PEFile(object):
                     iat_entry.name = str(name) if name else ""
             name = get_string(self.target, self.baseaddr + import_descriptor.Name)
             res.setdefault(name.lower(), []).extend(IAT)
-        return res
+        return res
+
+    @utils.fixedpropety
+    def binid(self):
+        """Return the hex-string {TimeStamp}{SizeOfCode} used by PDB to identify a PE.
+
+        I do not know the official name of this value...
+
+        :type: :class:`str`
+        """
+        nth = self.get_NT_HEADER()
+        timestamp = nth.FileHeader.TimeDateStamp
+        image_size = nth.OptionalHeader.SizeOfImage
+        return "{timestamp:08x}{image_size:x}".format(timestamp=timestamp, image_size=image_size)