pythongh-91349: Expose the crc32 function from the lzma library

Author: rmmancom

Doc/library/lzma.rst

@@ -310,6 +310,23 @@ Compressing and decompressing data in memory
    *preset* and *filters* arguments.
 
 
+.. function:: crc32(data, value=0)
+
+   .. index::
+      single: Cyclic Redundancy Check
+      single: checksum; Cyclic Redundancy Check
+
+   Computes a CRC (Cyclic Redundancy Check) checksum of *data*. The
+   result is a positive integer, less than :math:`2^32`. If *value* is present, it is used
+   as the starting value of the checksum; otherwise, a default value of 0
+   is used.  Passing in *value* allows computing a running checksum over the
+   concatenation of several inputs.  The algorithm is not cryptographically
+   strong, and should not be used for authentication or digital signatures.  Since
+   the algorithm is designed for use as a checksum algorithm, it is not suitable
+   for use as a general hash algorithm.
+
+   .. versionadded:: next
+
 .. function:: decompress(data, format=FORMAT_AUTO, memlimit=None, filters=None)
 
    Decompress *data* (a :class:`bytes` object), returning the uncompressed data

Lib/test/test_lzma.py

@@ -1,4 +1,5 @@
 import array
+import binascii
 from io import BytesIO, UnsupportedOperation, DEFAULT_BUFFER_SIZE
 import os
 import pickle
@@ -8,7 +9,7 @@
 import unittest
 from compression._common import _streams
 
-from test.support import _4G, bigmemtest
+from test.support import _1G, _4G, bigmemtest
 from test.support.import_helper import import_module
 from test.support.os_helper import (
     TESTFN, unlink, FakePath
@@ -17,6 +18,44 @@
 lzma = import_module("lzma")
 from lzma import LZMACompressor, LZMADecompressor, LZMAError, LZMAFile
 
+class ChecksumTestCase(unittest.TestCase):
+    # checksum test cases
+    def test_crc32start(self):
+        self.assertEqual(lzma.crc32(b""), lzma.crc32(b"", 0))
+        self.assertTrue(lzma.crc32(b"abc", 0xffffffff))
+
+    def test_crc32empty(self):
+        self.assertEqual(lzma.crc32(b"", 0), 0)
+        self.assertEqual(lzma.crc32(b"", 1), 1)
+        self.assertEqual(lzma.crc32(b"", 432), 432)
+
+    def test_penguins(self):
+        self.assertEqual(lzma.crc32(b"penguin", 0), 0x0e5c1a120)
+        self.assertEqual(lzma.crc32(b"penguin", 1), 0x43b6aa94)
+        self.assertEqual(lzma.crc32(b"penguin"), lzma.crc32(b"penguin", 0))
+
+    def test_crc32_unsigned(self):
+        foo = b'abcdefghijklmnop'
+        # explicitly test signed behavior
+        self.assertEqual(lzma.crc32(foo), 2486878355)
+        self.assertEqual(lzma.crc32(b'spam'), 1138425661)
+
+    def test_same_as_binascii_crc32(self):
+        foo = b'abcdefghijklmnop'
+        crc = 2486878355
+        self.assertEqual(binascii.crc32(foo), crc)
+        self.assertEqual(lzma.crc32(foo), crc)
+        self.assertEqual(binascii.crc32(b'spam'), lzma.crc32(b'spam'))
+
+
+# GH-54485 - check that inputs >=4 GiB are handled correctly.
+class ChecksumBigBufferTestCase(unittest.TestCase):
+
+    @bigmemtest(size=_4G + 4, memuse=1, dry_run=False)
+    def test_big_buffer(self, size):
+        data = b"nyan" * (_1G + 1)
+        self.assertEqual(lzma.crc32(data), 1044521549)
+
 
 class CompressorDecompressorTestCase(unittest.TestCase):
 

Misc/NEWS.d/next/Library/2025-03-18-14-18-06.gh-issue-91349.Qrnmxt.rst

@@ -0,0 +1 @@
+Expose the crc32 function from the lzma library as :func:`lzma.crc32`.

Modules/_lzmamodule.c

@@ -1587,10 +1587,40 @@ lzma_exec(PyObject *module)
     return 0;
 }
 
+/*[clinic input]
+_lzma.crc32 -> unsigned_int
+
+    data: Py_buffer
+    value: unsigned_int(bitwise=True) = 0
+        Starting value of the checksum.
+    /
+
+Compute a CRC-32 checksum of data.
+
+The returned checksum is an integer.
+[clinic start generated code]*/
+
+static unsigned int
+_lzma_crc32_impl(PyObject *module, Py_buffer *data, unsigned int value)
+/*[clinic end generated code: output=fca7916d796faf8b input=bb623a169c14534f]*/
+{
+    /* Releasing the GIL for very small buffers is inefficient
+       and may lower performance */
+    if (data->len > 1024*5) {
+        Py_BEGIN_ALLOW_THREADS
+        value = lzma_crc32(data->buf, (size_t)data->len, (uint32_t)value);
+        Py_END_ALLOW_THREADS
+    } else {
+        value = lzma_crc32(data->buf, (size_t)data->len, (uint32_t)value);
+    }
+    return value;
+}
+
 static PyMethodDef lzma_methods[] = {
     _LZMA_IS_CHECK_SUPPORTED_METHODDEF
     _LZMA__ENCODE_FILTER_PROPERTIES_METHODDEF
     _LZMA__DECODE_FILTER_PROPERTIES_METHODDEF
+    _LZMA_CRC32_METHODDEF
     {NULL}
 };