post-merge tweaks

python · AA-Turner · May 12, 2025 · May 10, 2025 · May 10, 2025 · May 10, 2025
commit 15fc9cee2d76a73bb4fc1f5f2ae4b1698e10235f
diff --git a/Lib/test/test_zstd.py b/Lib/test/test_zstd.py
@@ -507,7 +507,7 @@ def test_decompress_epilogue_flags(self):
         self.assertFalse(d.needs_input)
 
     def test_decompressor_arg(self):
-        zd = ZstdDict(b'12345678', True)
+        zd = ZstdDict(b'12345678', is_raw=True)
 
         with self.assertRaises(TypeError):
             d = ZstdDecompressor(zstd_dict={})
@@ -1021,6 +1021,10 @@ def test_decompressor_skippable(self):
 class ZstdDictTestCase(unittest.TestCase):
 
     def test_is_raw(self):
+        # must be passed as a keyword argument
+        with self.assertRaises(TypeError):
+            ZstdDict(bytes(8), True)
+
         # content < 8
         b = b'1234567'
         with self.assertRaises(ValueError):
@@ -1096,7 +1100,7 @@ def test_train_dict(self):
 
 
         TRAINED_DICT = train_dict(SAMPLES, DICT_SIZE1)
-        ZstdDict(TRAINED_DICT.dict_content, False)
+        ZstdDict(TRAINED_DICT.dict_content, is_raw=False)
 
         self.assertNotEqual(TRAINED_DICT.dict_id, 0)
         self.assertGreater(len(TRAINED_DICT.dict_content), 0)
@@ -1250,7 +1254,7 @@ def _nbytes(dat):
     def test_as_prefix(self):
         # V1
         V1 = THIS_FILE_BYTES
-        zd = ZstdDict(V1, True)
+        zd = ZstdDict(V1, is_raw=True)
 
         # V2
         mid = len(V1) // 2
@@ -1266,7 +1270,7 @@ def test_as_prefix(self):
         self.assertEqual(decompress(dat, zd.as_prefix), V2)
 
         # use wrong prefix
-        zd2 = ZstdDict(SAMPLES[0], True)
+        zd2 = ZstdDict(SAMPLES[0], is_raw=True)
         try:
             decompressed = decompress(dat, zd2.as_prefix)
         except ZstdError: # expected

diff --git a/Modules/_zstd/clinic/compressor.c.h b/Modules/_zstd/clinic/compressor.c.h
diff --git a/Modules/_zstd/clinic/zstddict.c.h b/Modules/_zstd/clinic/zstddict.c.h
diff --git a/Modules/_zstd/compressor.c b/Modules/_zstd/compressor.c
@@ -318,9 +318,8 @@ _zstd_load_c_dict(ZstdCompressor *self, PyObject *dict)
 /*[clinic input]
 @classmethod
 _zstd.ZstdCompressor.__new__ as _zstd_ZstdCompressor_new
-
     level: object = None
-        The compression level to use, defaults to ZSTD_CLEVEL_DEFAULT.
+        The compression level to use. Defaults to COMPRESSION_LEVEL_DEFAULT.
     options: object = None
         A dict object that contains advanced compression parameters.
     zstd_dict: object = None
@@ -335,17 +334,15 @@ function instead.
 static PyObject *
 _zstd_ZstdCompressor_new_impl(PyTypeObject *type, PyObject *level,
                               PyObject *options, PyObject *zstd_dict)
-/*[clinic end generated code: output=cdef61eafecac3d7 input=a9e9d73f246d6588]*/
+/*[clinic end generated code: output=cdef61eafecac3d7 input=b9ea61ecafbb1b1e]*/
 {
     ZstdCompressor* self = PyObject_GC_New(ZstdCompressor, type);
     if (self == NULL) {
         goto error;
     }
 
-    self->dict = NULL;
     self->use_multithread = 0;
 
-
     /* Compression context */
     self->cctx = ZSTD_createCCtx();
     if (self->cctx == NULL) {
@@ -378,18 +375,17 @@ _zstd_ZstdCompressor_new_impl(PyTypeObject *type, PyObject *level,
         }
     }
 
-    /* Load dictionary to compression context */
+    /* Load zstd dictionary to compression context */
+    self->dict = NULL;
     if (zstd_dict != Py_None) {
         if (_zstd_load_c_dict(self, zstd_dict) < 0) {
             goto error;
         }
-
-        /* Py_INCREF the dict */
         Py_INCREF(zstd_dict);
         self->dict = zstd_dict;
     }
 
-    // We can only start tracking self with the GC once self->dict is set.
+    // We can only start GC tracking once self->dict is set.
     PyObject_GC_Track(self);
 
     return (PyObject*)self;

diff --git a/Modules/_zstd/decompressor.c b/Modules/_zstd/decompressor.c
@@ -530,7 +530,6 @@ stream_decompress(ZstdDecompressor *self, Py_buffer *data, Py_ssize_t max_length
 /*[clinic input]
 @classmethod
 _zstd.ZstdDecompressor.__new__ as _zstd_ZstdDecompressor_new
-
     zstd_dict: object = None
         A ZstdDict object, a pre-trained zstd dictionary.
     options: object = None
@@ -545,14 +544,13 @@ function instead.
 static PyObject *
 _zstd_ZstdDecompressor_new_impl(PyTypeObject *type, PyObject *zstd_dict,
                                 PyObject *options)
-/*[clinic end generated code: output=590ca65c1102ff4a input=e73db62a54e25e4b]*/
+/*[clinic end generated code: output=590ca65c1102ff4a input=0f161edb33d83216]*/
 {
     ZstdDecompressor* self = PyObject_GC_New(ZstdDecompressor, type);
     if (self == NULL) {
         goto error;
     }
 
-    self->dict = NULL;
     self->input_buffer = NULL;
     self->input_buffer_size = 0;
     self->in_begin = -1;
@@ -574,13 +572,12 @@ _zstd_ZstdDecompressor_new_impl(PyTypeObject *type, PyObject *zstd_dict,
         goto error;
     }
 
-    /* Load dictionary to decompression context */
+    /* Load zstd dictionary to decompression context */
+    self->dict = NULL;
     if (zstd_dict != Py_None) {
         if (_zstd_load_d_dict(self, zstd_dict) < 0) {
             goto error;
         }
-
-        /* Py_INCREF the dict */
         Py_INCREF(zstd_dict);
         self->dict = zstd_dict;
     }
@@ -592,7 +589,7 @@ _zstd_ZstdDecompressor_new_impl(PyTypeObject *type, PyObject *zstd_dict,
         }
     }
 
-    // We can only start tracking self with the GC once self->dict is set.
+    // We can only start GC tracking once self->dict is set.
     PyObject_GC_Track(self);
 
     return (PyObject*)self;

diff --git a/Modules/_zstd/zstddict.c b/Modules/_zstd/zstddict.c
@@ -29,13 +29,13 @@ class _zstd.ZstdDict "ZstdDict *" "&zstd_dict_type_spec"
 @classmethod
 _zstd.ZstdDict.__new__ as _zstd_ZstdDict_new
     dict_content: object
-        A bytes-like object, dictionary's content.
+        The content of a Zstandard dictionary as a bytes-like object.
+    /
+    *
     is_raw: bool = False
-        This parameter is for advanced user. True means dict_content
-        argument is a "raw content" dictionary, free of any format
-        restriction. False means dict_content argument is an ordinary
-        zstd dictionary, was created by zstd functions, follow a
-        specified format.
+        If true, perform no checks on *dict_content*, useful for some
+        advanced cases. Otherwise, check that the content represents
+        a Zstandard dictionary created by the zstd functions.
 
 Represents a zstd dictionary, which can be used for compression/decompression.
 
@@ -46,7 +46,7 @@ ZstdDecompressor objects.
 static PyObject *
 _zstd_ZstdDict_new_impl(PyTypeObject *type, PyObject *dict_content,
                         int is_raw)
-/*[clinic end generated code: output=3ebff839cb3be6d7 input=a0356fa7336dea1c]*/
+/*[clinic end generated code: output=3ebff839cb3be6d7 input=e9e22fdc68fa04cc]*/
 {
     ZstdDict* self = PyObject_GC_New(ZstdDict, type);
     if (self == NULL) {
@@ -86,10 +86,7 @@ _zstd_ZstdDict_new_impl(PyTypeObject *type, PyObject *dict_content,
     if (!is_raw && self->dict_id == 0) {
         char *msg = "The dict_content argument is not a valid zstd "
                     "dictionary. The first 4 bytes of a valid zstd dictionary "
-                    "should be a magic number: b'\\x37\\xA4\\x30\\xEC'.\n"
-                    "If you are an advanced user, and can be sure that "
-                    "dict_content argument is a \"raw content\" zstd "
-                    "dictionary, set is_raw parameter to True.";
+                    "should be a magic number: b'\\x37\\xA4\\x30\\xEC'.\n";
         PyErr_SetString(PyExc_ValueError, msg);
         goto error;
     }