Merge pull request #65 from tcalmant/bug-fix

Bug fix

Author: tcalmant

javaobj/v1/beans.py

@@ -173,9 +173,14 @@ def __hash__(self):
         return UNICODE_TYPE.__hash__(self)
 
     def __eq__(self, other):
-        if not isinstance(other, UNICODE_TYPE):
+        # Accept both UNICODE_TYPE and plain str.
+        # In Python 2, UNICODE_TYPE is unicode while str literals are bytes;
+        # including str here lets assertEqual(java_string, "literal") work
+        # in Python 2 as well as Python 3 (where str == UNICODE_TYPE).
+        if not isinstance(other, (UNICODE_TYPE, str)):
             return False
-        return UNICODE_TYPE.__eq__(self, other)
+        result = UNICODE_TYPE.__eq__(self, other)
+        return False if result is NotImplemented else result
 
 
 class JavaEnum(JavaObject):
@@ -199,7 +204,7 @@ def __init__(self, classdesc=None):
         self.classdesc = classdesc
 
     def __hash__(self):
-        return list.__hash__(self)
+        return object.__hash__(self)
 
 
 class JavaByteArray(JavaObject):

javaobj/v1/transformers.py

@@ -136,7 +136,7 @@ def __extra_loading__(self, unmarshaller, ident=0):
             """
             # Ignore the blockdata opid
             (opid,) = unmarshaller._readStruct(">B")
-            if opid != ClassDescFlags.SC_BLOCK_DATA:
+            if opid != TerminalCode.TC_BLOCKDATA:
                 raise ValueError("Start of block data not found")
 
             # Read HashMap fields

javaobj/v1/unmarshaller.py

@@ -83,7 +83,7 @@
 # Convertion of a Java type char to its NumPy equivalent
 NUMPY_TYPE_MAP = {
     TypeCode.TYPE_BYTE: "B",
-    TypeCode.TYPE_CHAR: "b",
+    TypeCode.TYPE_CHAR: ">u2",
     TypeCode.TYPE_DOUBLE: ">d",
     TypeCode.TYPE_FLOAT: ">f",
     TypeCode.TYPE_INTEGER: ">i",
@@ -508,7 +508,8 @@ def do_object(self, parent=None, ident=0):
         # classdata[]
 
         if (
-            classdesc.flags & ClassDescFlags.SC_EXTERNALIZABLE
+            classdesc is not None
+            and classdesc.flags & ClassDescFlags.SC_EXTERNALIZABLE
             and not classdesc.flags & ClassDescFlags.SC_BLOCK_DATA
         ):
             # TODO:
@@ -831,8 +832,9 @@ def _oops_dump_state(self, ignore_remaining_data=False):
             "(2nd line is an actual position!):"
         )
 
-        # Do not use a keyword argument
-        self.object_stream.seek(-16, os.SEEK_CUR)
+        # Do not use a keyword argument; clamp to avoid seeking before start
+        current_pos = self.object_stream.tell()
+        self.object_stream.seek(max(0, current_pos - 16))
         position = self.object_stream.tell()
         the_rest = self.object_stream.read()
 

javaobj/v2/core.py

@@ -146,7 +146,7 @@ def run(self):
         # Read content
         contents = []  # type: List[ParsedJavaContent]
         while True:
-            self._log.info("Reading next content")
+            self._log.debug("Reading next content")
             start = self.__fd.tell()
             try:
                 type_code = self.__reader.read_byte()

javaobj/v2/transformers.py

@@ -506,7 +506,7 @@ class NumpyArrayTransformer(ObjectTransformer):
     # Convertion of a Java type char to its NumPy equivalent
     NUMPY_TYPE_MAP = {
         TypeCode.TYPE_BYTE: "B",
-        TypeCode.TYPE_CHAR: "b",
+        TypeCode.TYPE_CHAR: ">u2",
         TypeCode.TYPE_DOUBLE: ">d",
         TypeCode.TYPE_FLOAT: ">f",
         TypeCode.TYPE_INTEGER: ">i",

pyproject.toml

@@ -38,8 +38,8 @@ classifiers = [
 ]
 
 dependencies = [
-    "enum34; python_version<='3.4'",
-    "typing; python_version<='3.4'"
+    "enum34; python_version<'3.4'",
+    "typing; python_version<'3.5'"
 ]
 
 [project.optional-dependencies]

setup.py

@@ -72,8 +72,8 @@ def read(fname):
     packages=["javaobj", "javaobj.v1", "javaobj.v2"],
     test_suite="tests",
     install_requires=[
-        'enum34;python_version<="3.4"',
-        'typing;python_version<="3.4"',
+        "enum34; python_version<'3.4'",
+        "typing; python_version<'3.5'"
     ],
     long_description=read("README.md"),
     long_description_content_type="text/markdown",
@@ -91,6 +91,8 @@ def read(fname):
         "Programming Language :: Python :: 3.10",
         "Programming Language :: Python :: 3.11",
         "Programming Language :: Python :: 3.12",
+        "Programming Language :: Python :: 3.13",
+        "Programming Language :: Python :: 3.14",
         "Topic :: Software Development :: Libraries :: Python Modules",
     ],
 )

tests/test_v1.py

@@ -337,18 +337,12 @@ def test_char_array(self):
         jobj = self.read_file("testCharArray.ser")
         pobj = javaobj.loads(jobj)
         _logger.debug(pobj)
-        self.assertEqual(
-            pobj,
-            [
-                "\u0000",
-                "\ud800",
-                "\u0001",
-                "\udc00",
-                "\u0002",
-                "\uffff",
-                "\u0003",
-            ],
-        )
+        # Compare by code points to avoid Python 2/3 string literal ambiguity
+        # (ruff strips u"" prefixes; \u-escapes in plain str are literal in Py2)
+        expected_codepoints = [0x0000, 0xD800, 0x0001, 0xDC00, 0x0002, 0xFFFF, 0x0003]
+        self.assertEqual(len(pobj), len(expected_codepoints))
+        for actual, cp in zip(pobj, expected_codepoints):
+            self.assertEqual(ord(actual), cp)
         self._try_marshalling(jobj, pobj)
 
     def test_2d_array(self):

tests/test_v2.py

@@ -417,18 +417,12 @@ def test_char_array(self):
         jobj = self.read_file("testCharArray.ser")
         pobj = javaobj.loads(jobj)
         _logger.debug(pobj)
-        self.assertEqual(
-            pobj,
-            [
-                "\u0000",
-                "\ud800",
-                "\u0001",
-                "\udc00",
-                "\u0002",
-                "\uffff",
-                "\u0003",
-            ],
-        )
+        # Compare by code points to avoid Python 2/3 string literal ambiguity
+        # (ruff strips u"" prefixes; \u-escapes in plain str are literal in Py2)
+        expected_codepoints = [0x0000, 0xD800, 0x0001, 0xDC00, 0x0002, 0xFFFF, 0x0003]
+        self.assertEqual(len(pobj), len(expected_codepoints))
+        for actual, cp in zip(pobj, expected_codepoints):
+            self.assertEqual(ord(actual), cp)
 
     def test_2d_array(self):
         """